3
Empirical Analysis with Explanatory Design of the Role of French Private Equity Firms in the Formation of Alliances

With the theoretical analysis complete, it now remains to test the propositions advanced in the Chapter 2 (see Table 2.7). In this chapter, we make explicit the methodology that we are using (section 3.1). It involves a multimethod study including an econometric study and a multiple case study. We then use it to test our theoretical model (section 3.2).

This empirical section seeks to test our theoretical framework on the French private equity market. It should allow us to respond to the general questions which we have raised. These are as follows:

• – Why do private equity firms intervene in the formation of alliances for their portfolio companies?
• – How do they intervene?
• – What is their impact on the resulting creation of value?

As these questions apply to the whole set of French private equity firms, the question of the general character of our propositions also arises.

Questions of how and why private equity firms intervene in the formation of alliances can be typically approached through case studies. The question of the general character of the possible roles played by private equity firms suggests a methodology which allows statistical generalization of the results, thus justifying the use of econometric techniques.

After justifying in a more detailed fashion the use of these two techniques (econometric and case study), they are, at first, presented separately. Later, their respective results are then reconciled in order to reach a conclusion at the level of the multimethod study as a whole.

3.1. Methodology: a multimethod study

After having presented the overall concept in section 3.1.1, we justify the joint and complementary use of statistical and econometric techniques and a case study in section 3.1.2.

3.1.1. The overall concept

The research question involves exploring the role of French private equity firms in the formation of strategic alliances and evaluating their impact on the resulting value creation. Thus, we must first of all answer the following general questions: what roles do private equity firms play in the formation of alliances, that is, how do they intervene and why? Bear in mind that we explicitly take into account the point of view of the small and medium-sized enterprises (SMEs) forming the alliance, as well as that of the private equity firm (PEF). Second, we examine the importance of the phenomenon. Is it a phenomenon generalizable among all French PEFs? The hypotheses resulting from the theoretical analysis of the issues posed in fact do have a general character and are valid for French PEFs as a whole. This general character must therefore also be put to the test.

Concerning the first point, the two questions of “how” and “why” can typically be treated through the use of a case study methodology [DAV 05, YIN 09, p. 175]. In contrast, the second point dealing with the importance of the phenomenon for the French private equity market requires the use of a methodology allowing us to analyze the possibility of statistically generalizing observed results over the whole of a parent population. This is what statistical and econometric techniques allow. Consequently, the empirical study is based on a confrontation between the facts and the theoretical concept, relying both on statistical and econometric studies and on a multiple case study. This empirical study will produce a summary of the results of the two studies that comprise it. It therefore consists of a multimethod study [YIN 09, p. 175], which aims to test the theoretical framework we have established.

The whole of the empirical study as well as the two methods comprising it are used, in our case, in an explanatory design, that is, with the explanatory, confirmatory objective of testing hypotheses [MIL 03, p. 83 and 84]. The statistical and econometric study and the case study are used in a complementary manner due to their specificities and according to their respective advantages and disadvantages. Although they are reported separately, their conceptualization and their implementation took place in parallel. Each of them therefore results in its distinct analysis and presentation, before we conclude the analysis with a final reconciliation of the results drawn from the different methods of empirical analysis [YIN 09, p. 174]. Before embarking on these analyses, let us begin by discussing the respective advantages of the two different methods to justify their use.

3.1.2. The joint and complementary use of statistical and economic techniques and case studies

We conjointly conduct statistical and economic studies and case studies because of their specificities and in order to use the two methods according to their respective advantages for testing our research hypotheses. The two methods are therefore utilized in a complementary manner. This section aims to detail the advantages of the two methods for answering our research question.

The strong point of statistical and econometric tests is that they allow us to detect correlations between variables and to understand the frequency of occurrence of the phenomena studied [YIN 09, p. 175]. They thus allow us to corroborate or to refute the propositions raised, allowing us to offer with a certain degree of certainty a statistical generalization of the facts based on a sample of data representative of the parent population of that sample [MIL 03, p. 83].

Used in an explanatory design, case studies, in turn, allow us to test the theoretical concept at a finer and more detailed level. Principally, they make it possible to respond to questions of “why” a certain phenomenon occurs and “how” it occurs. They thus complement the results of econometric tests as they can, once the correlations between variables have been detected, go beyond these correlations and verify the plausibility of the mechanisms underlying the links of causality, which we advance in theory to explain the facts [YIN 09, p. 175]. They thus make it possible to refine the degree of the analysis and to see if the hypotheses, of a general character, hold true in a specific context. They thus go beyond statistical abstraction and aggregated data. They enable a generalization of an analytical type. Moreover, in our specific case, they allow us to specifically account for the points of view of PEFs and SMEs in a single empirical result.

Contrary to the criticism often advanced that it is not possible to generalize from results obtained from case studies, we can plausibly claim to do so [SCA 90, p. 276; FLY 06, p. 219, p. 221; FLY 11, pp. 304–305]. However, we must recognize that the type of generalization differs in nature from that of econometric studies. Thus, we cannot claim a “statistical generalization” but rather a so-called analytical generalization, on the basis of plausibility tests [SCA 90, p. 270f.; GIB 08, p. 1468; GIB 10, p. 714]. With econometric generalizations, an inference is made about a population on the basis of data issuing from a representative sample of this population. In the case of an analytical generalization, we generalize the fact that empirical results support a previously established theory. The empirical results of the case studies are thus compared to a previously elaborated theory, which serves as a model or as a pattern for reproduction (a “template”). The result of the study may be considered to be particularly robust when several facts support the same theory (literal replication) while simultaneously not supporting a rival theory, which aims to explain the phenomenon (theoretical replication) [YIN 09, pp. 38–39 and p. 60]. We can thus claim with a certain degree of certainty to have found an analytical generalization.

By combining the two types of generalization (statistical and analytical), one can claim, in the end, a degree of generalization superior to that which one would obtain by using only a single method. Apart from allowing us to proceed to a more complex analysis using the two methods in a complementary manner, the multimethod study allows us to confirm the results by triangulation [YIN 03, p. 150, MIL 03, p. 83]. This is even more important when we need to grasp concepts which are difficult to evaluate and quantify.

3.2. Testing the theoretical framework

Testing the theoretical framework comprises three parts. The first two are devoted to the two components of the multimethod study: the statistical and econometric study (section 3.2.1) and the multiple case study (section 3.2.2). In section 3.2.3, we compare the results of the two analyses.

3.2.1. Econometric analysis

We begin by presenting the collection of data (section 3.2.1.1) before specifying the nature of the variables and the selected model (section 3.2.1.2). We then outline the approach allowing us to test our hypotheses through the selected model, the difficulties encountered in its implementation and the way in which we overcame them (section 3.2.1.3). Finally, we present and discuss the results obtained (section 3.2.1.4).

3.2.1.1. Collection of data

For the reasons cited in the introductory section, we were not able to access satisfactory secondary data (for example through access to databases) to carry out our study. We thus decided to carry out our own survey. The survey was designed by constructing two questionnaires:

• – one intended for French PEFs who were members of AFIC (Association Française des Investisseurs pour la Croissance, now known as France Invest);
• – the other intended for SMEs in which the PEFs held shareholdings.

The design of the survey in two questionnaires arises from our theoretical modeling. Systematically, we analyzed the problem from the respective points of view of SMEs forming alliances and the PEFs involved in these alliances. The questionnaires contained a first part intended to gather information allowing us to describe the phenomenon (number of alliances formed or percentage of companies in which the PEF had holdings having formed alliances; type of goals underlying these alliances; nature of contracts; roles that the PEFs see themselves as playing therein). This part was presented in the introduction. The second part of the questionnaire was designed to gather the information necessary to test our research hypotheses. The respondents were asked to indicate for different issues if they thought that the point mentioned had an impact on the formation of alliances, and if so, to what extent (ranging from totally negative to totally positive). A third part mentioned certain points related to the respondent or the SME/PEF for which they were responding.

The decision to ask the second section of questions as indicated above is based primarily on the fact that both the PEFs and the SMEs involved were somewhat reluctant to deliver information in the necessary direct form. This difficulty was confirmed after a discussion with the director of the statistical center at AFIC. The chosen solution allowed us to circumvent this and to obtain a relatively high response rate so as to be able to adequately perform our investigation. The responses thus obtained have a subjective character which is, however, balanced, at least in part, by the contrast between the points of view of the PEFs and the SMEs via the two questionnaires.

After design, a first version of the questionnaires was submitted, in a first phase, to the criticism of experts from FARGO (French Center for Research in Finance, Organizational Architecture and Corporate Governance) as well as CMBOR (Center for Management Buy-Out and Private Equity Research). This latter has already produced similar questionnaires in conjunction with EVCA (European Private Equity & Venture Capital Association, now known as Invest Europe). The questionnaires were then submitted for pretesting to the investment managers involved in the case studies and certain directors of their portfolio companies.

Once validated, the questionnaires were sent by email to the investment managers of the PEFs, which were members of AFIC, by means of the Surveymonkey website (http://fr.surveymonkey.com/). After many attempts, we were also able to contact Mr. Hervé Schricke, president of AFIC at the time of the survey. He gave us his consent for his association’s department of statistics to support these investigations by agreeing to publish links to the questionnaires, accompanied by an advertisement, in two of its newsletters. The questions were put online at the beginning of May 2012 for a period of 8 months. The investment managers of the PEFs were encouraged, on the one hand, to fill out the questionnaire intended for PEFs, and, on the other, to pass on the questionnaire intended for SMEs to directors of their portfolio companies. As described in more detail in the introductory part, the responses to the questionnaire addressed to PEFs were mainly obtained following telephone reminders. Attempts were also undertaken to meet the manager of the FSI (strategic fund for investment) so he could give us his consent to disseminate an e-mail including links to our questionnaires directly to PEF members. This request was not successful. The response from the questionnaire intended for SMEs was too low to be usable.

3.2.1.2. Specification of the model

We will start by presenting the dependent variable, as it is measured through the questionnaire, because it affects the choice of econometric model. We will then present the independent variables and the control variables. They have an impact on the way in which we will specify, and thus test, the chosen model. Finally, we will present the characteristics of the selected model.

3.2.1.2.1. Dependent variable

The dependent variable is the “formation of alliances”. In the questionnaires, one question was devoted to this. Respondents to the questionnaire addressed to PEFs were asked to indicate the approximate percentage of the companies they supported who had formed an alliance, either of intra- or extraportfolio type. They could choose from among five responses (Figure 3.1).

Figure 3.1. Overview of the first question of the questionnaire addressed to PEFs

Having posed the question in this way, the responses obtained for the dependent variable are shown in the form of intervals, answers of 0% or 100% being assimilated to intervals:

• Interval 1: [0–0%]; code as 0
• Interval 2: [1–10%]; code as 1
• Interval 3: [11–50%]; code as 2
• Interval 4: [51–99%]; code as 3
• Interval 5: [100–100%]; code as 4

The questionnaire as a whole was constructed so as to obtain responses for the formation of both “intra-” and “extra”-portfolio alliances. Based on this, we decided, for the econometric study, to perform two regressions, the first for the formation of “intraportfolio” alliances, the second for that of “extraportfolio” alliances. The dependent variables of the two models are referred to, respectively, as “allin” and “allex”.

3.2.1.2.2. Independent variables

Following our theoretical analysis, we distinguished three types of independent variables:

• – those arising from contractual hypotheses;
• – those based on a knowledge-based analysis;
• – those which simultaneously cover a contractual variant and a knowledge-based variant.

Here, we discuss only the designation of the independent variables as they appear in the regression, their meaning and the number of the hypothesis that they represent (for a summary of our hypotheses, see Table 2.7).

Contractual variables

• – The variable “Rp” represents the reputation of the PEFs (Hypothesis 1).
• – The variables “Ple” and “State” designate the presence of a competitiveness cluster or the State, respectively (Hypothesis 3).
• – The variable “maj” indicates a majority stake held by the PEF (Hypothesis 4).
• – The variable “Cot” indicates the publicly traded or otherwise status of the SMEs involved (Hypothesis 6).

Knowledge-based variables

• – The variables “Spsec” and “rgInt” designate, respectively, the sectoral specialization and the regional/international specialization of the PEF (Hypotheses 7a and 7b).
• – The variable “Diff” indicates whether the formation of alliances is for the PEF a way of differentiating itself on the private equity market (Hypothesis 9).
• – The variable “Ca” indicates the presence of the PEF on the board of directors of the companies they support (Hypothesis 11a).

Variables of mixed nature (contractual and knowledge-based)

The variable “Nbrepart” takes into account the number of shareholdings managed by the PEF (Hypotheses 8a and 8b).

Hypothesis 10, predicting a more prominent role for the PEF at the formation of “intra” than of “extra” alliances, might be tested by comparing the regressions on allin and allex.

Because of a lack of data, particularly due to the fact that the second questionnaire (intended for SMEs) could not be used, we could not use the econometric study to test Hypotheses 2, 5 or 11b, or parts of Hypotheses 8a and 8b. As for Hypotheses 7a and 7b, they could not be completely tested.

Although it was possible for us to test the impact of the reputation of a PEF on the formation of alliances (Hypothesis 1), we were not in a position to test Hypothesis 2, predicting that a PEF with weak reputational capital would have a greater tendency to insert its portfolio companies into alliances with partners with good reputations on the market.

Hypothesis 5 is associated with the impact of alliances previously formed by the SMEs supported by private equity on the formation of new alliances. The information necessary to test this hypothesis was included in the questionnaire intended for SMEs, which did not succeed.

Hypothesis 11b concerns the participation of the PEF in organizations such as AFIC, EVCA or others. The variable allowing us to take account of the participation or otherwise of PEFs in such associations is binary by nature: coded 1 if they are part of at least one association and 0 in the alternative case. The problem which finally led me to not retain this variable is that the questionnaire was only sent to PEFs who were members of AFIC. Thus, all the PEFs who replied received code 1 for this variable, which did not allow us to properly analyze the impact of participation/non-participation of PEFs in associations on the formation of alliances.

For Hypotheses 8a and 8b, we were not able to gather information concerning the geographical distance between the PEF and its portfolio companies and could not verify its impact on the formation of alliances.

For Hypotheses 7a and 7b, we predicted a positive impact of a regional or sectoral focus in a PEF’s investments on the formation of regional or sectoral alliances. We also predicted that a PEF’s investments in different countries would have a positive impact on the formation of alliances allowing international development of the companies involved. However, we only obtained information related to the general impact of the PEF’s investment focus on the formation of intra-/extra-alliances. Consequently, we could not measure the impact on the formation of certain types of alliance (regional/sectoral/international).

All these hypotheses were, however, tested by the case study method, which will be presented in its turn.

3.2.1.2.3. Control variables

The control variables are those variables which are likely to explain the formation of alliances (intra or extra) for companies supported by private equity, linked to determinants different from those associated with the independent variables. They allow us to ensure that our hypotheses are valid “all else being equal”. We selected four of these. They mainly consist of variables representing certain characteristics of PEFs or relating to the portfolio manager.

The first, “Spfin”, takes account of the PEF’s investment specialization according to the stage of financing (startup, venture, growth, turnaround, etc.). We may suppose that the more the PEF supports companies who are going through precarious phases, the more these companies will possibly lack resources and will thus be interested in forming alliances. In addition, the younger these companies, the more they will lack the visibility to attract alliance partners. The role of a PEF should thus be strengthened.

The variable “indus” indicates the membership of the PEF in an industrial group or company, and the variable “finan”, the membership of the PEF in a financial group (for example a bank). In these two cases, we are dealing with a captive PEF, for example a subsidiary of an industrial group or of a bank. We may thus suppose that a PEF, which is a subsidiary of a financial group, will adopt a financial rather than an industrial approach, which may have a negative impact on the formation of alliances. For PEFs, which are subsidiaries of an industrial group, we may suppose that the group or company investing in a start-up will be reluctant for the latter to form too many alliances, for reasons of self-interest. We may thus suppose that if the company invests in a start-up to support and subsequently profit from its know-how, the company may limit the formation of alliances for fear of losing exclusive access to this know-how [DUS 10].

Finally, the variable “Exp” takes into account the experience of the investment manager who supports those companies likely to form alliances. We may suppose that the more experienced this person, the more skills they will have to create relationships with the companies they support. Having had experience with a number of such cases, they will be likely to know more companies who may constitute potential alliance partners for the supported companies. The experience of the investment manager should thus have a positive impact on the formation of alliances.

By the same logic, we have included the variable “Nbresyn” representing the number of partners in syndication with the PEF. The fact that several PEFs co-invest in a single SME may have a positive effect on the formation of alliances by that company. A greater number of PEFs increase the skills and experience available, and thus the probability that the idea of forming an alliance will emerge. Moreover, the SME may also benefit from access to a larger network of potential alliance partners. We thus expect a positive link.

3.2.1.2.4. The nature of the independent and control variables

Except for the variable “Diff”, which is binary and thus takes the values (0; 1), all the other independent variables are of the type of responses to a Likert scale with six terms. The questions asked of the respondents relate to the impact of variables on the formation of either intra- or extraportfolio alliances. They could choose between six responses (not relevant; totally negative impact; rather negative; neutral; rather positive; totally positive). “Not relevant” indicates, for example, that if we ask a PEF if the fact of taking majority stakes has an impact on the formation of alliances for its holding, they could tick “not relevant” in the case that they do not take majority stakes. The “not relevant” information thus allows us to distinguish the PEFs who cannot answer the question from those who do not wish to respond to it.

Measured thus, the independent and control variables are ordered variables. In other words, the categories of responses can be ordered in a hierarchy (except, in principle, for the “not relevant” response), although we cannot measure the distances between the categories [JAM 04, CAR 07, p. 111]. The variables are thus of an ordered, qualitative nature. They have been coded as follows:

• 0 = not relevant
• 1 = totally negative
• 2 = rather negative
• 3 = neutral
• 4 = rather positive
• 5 = totally positive

However, in coding these variables from 0 to 5 without further precision, we make these variables quantitative for the software, although they are qualitative and categorical. So as to better take account of their nature, we transform these variables (except for the variable “Diff”) into binary variables, taking the values (0; 1). For example, if the respondent indicates that the variable Rp has a “neutral” impact on the formation of alliances, the values for this variable will be as follows:

To do this, we indicate to the software that the variables are in binary modalities (dummy variables). In Stata, the software being used, this is done by using, for example, not the independent variable “Rp”, but the variable with the prefix “ib3.”, thus, “ib3.Rp”. The “i” (for “indicator”) indicates that it is to be treated as binary, the “b3” is an option that allows us to tell the software to base its calculations on the modality of the reference coded 3 (“neutral” in our case)¹. Each modality of the variable (Rp in our example) is thus considered as binary.

This transformation, in turn, leads to eliminating the hierarchical nature of the variables. Excluding the “not relevant” category, a gradation does exist between the modalities, and we lose this information. However, as indicated above, if we treat the variables as discrete, the software uses a measurement scale of 0–5, which is also inaccurate. Moreover, in this case, we would have to exclude the category “not relevant” (coded 0) for the regression, as its inclusion in this continuous scale (ranging from 0 to 5) would be completely erroneous. We thus also lose this information, which, in our case, may be harmful given the small size of the overall sample and the sometimes non-negligible number of observations involving the category “not relevant”.

We continue by treating our variables as composed of several binary modalities. We thus lose the ranking information. Nevertheless, given that we are aware of the presence of a scale, we can “rectify” this loss of information when we interpret the results. The “loss” of information in the end therefore does not appear to be of great importance.

3.2.1.2.5. Models selected

With a dependent variable in the form of intervals, we have the possibility of using three econometric models:

1. 1) regression by intervals intreg, which thus treats our dependent variable as a continuous variable, but one whose observations are only found within certain intervals;
2. 2) probit/logit ordered regressions which treat our dependent variable as made up of several binary modalities, while keeping a scale between the categories;
3. 3) linear regression by substituting for the intervals of our dependent variable the central values of each interval.

Let us return to point (1). Regression by intervals is a generalization of the tobit model. The function was designed, in particular, for data of the interval type. That is, data where we know the interval in which the observation is located without knowing its exact value². This corresponds exactly to our dependent variable: respondents indicate a percentage interval of the number of its portfolio companies who have formed alliances, without yet indicating an exact number. They have a choice between five intervals. Through regression by intervals, our dependent variable (allin or allex) is thus treated as an a priori quantitative variable (going from 0% to 100%), but of which we only know the intervals and not the exact values. Each interval is thus conceived as a range of data with censoring to both left and right, called “interval censoring”.

For point (2), a priori nothing prevents us using ordered logit or probit, which are the nonlinear models, considering in our case the dependent variable as made up of several binary categories with an order between them. These models are based on the introduction of a latent, unobservable variable y, which is assumed to be linear and which takes the values of our dependent variable which are observed each time that it crosses defined thresholds. As the latent variable is linear, we can reduce the problem to an analysis of the variance on this latent variable [LEB 00, pp. 10–11]. The two models (ordered probit and logit) thus take the following form [LON 01, pp. 138–139]:

where:

• – = latent, unobserved variable, which represents the observed dependent variable of an ordered polytomous nature (y_i);
• – x_i = the independent variables that, in our case, are qualitative and ordered;
• – i = the different observations or, in our case, the set of PEFs who are members of AFIC indicated by i and which allow us to observe the different values of y_i;
• – ε_i = the error term.

In our case, the relationship between and y_i is specified as follows:

The values taken by the independent variables thus determine that of the latent unobservable variable, which may be interpreted as a propensity to generate an event (of the type y_i = 0, 1, 2, 3 or 4 in our case). We thus observe y_i = 0, 1, 2, 3 or 4 when finds itself in one of the intervals defined above. We deduce from this that (with “probability” = Prob):

We can then replace by (x_i * β + ε_i)

The logit and probit regression models then differ as to the distribution function of the rule chosen for the error term ε_i. This is either a Gaussian rule (probit) with an expected value of 0 and a variance equal to 1, or a logistic rule (logit) with a variance of π²/3.

Concerning point (3), the third solution mentioned consists of dealing with our data through a linear regression on the central values of the intervals of our dependent variable. However, this would entail a quite coarse simplification of the nature of the data for our independent variable. Thus, we will continue by only using intreg and ordered logit/probit.

3.2.1.3. Implementation of the regression: difficulties encountered and their resolution

The application was conducted by means of the Stata software, version 12. Before proceeding with the regression, a look at the frequencies for the different categories of our dependent variables might prove useful in view of the problems associated with categories represented by low frequencies.

For allin as for allex, we notice that category 4, indicating that 100% of companies supported in the portfolio form alliances, is very sparsely populated (respectively, 1 and 4 observations). We therefore decided to combine it with category 3³.

We also decided to not include the category coded as 0, which indicates that for the PEF in question, no company in its portfolio has formed an alliance of this type (intra or extra). This is interesting information at the descriptive level, as it signifies that some PEFs do not engage in these two types of alliance (intra/extra) for their portfolio companies. This information allows us to distinguish between these PEFs and those who simply did not wish to respond. On the other hand, the information is not useful for the regression itself, which is used to determine the role of the PEF when it forms such a type of alliance. We therefore only include those PEFs which in fact do so.

For the same reason, our regression will not take into account modality 0 (not relevant) for the independent variables when it co-occurs with category 0 for allin/allex. Note that category 0 for the independent variables covers two different kinds of information. Either the respondent has indicated 0, thus “not relevant”, because the independent variable is not relevant to them, or because, while that variable is relevant to them, it is not so in a general sense because they do not form that kind of alliance. For example, if we raise the question of the impact of taking a majority stake on the formation of “intra” alliances, the respondent may indicate “not relevant” because they do not take majority stakes. They may also indicate “not relevant” because they only form “extraportfolio” alliances.

3.2.1.3.1. Ideal scenario and difficulties encountered

The ideal scenario is being able to test, by means of the intreg function or the ordered logit/probit function, the impact of the whole set of independent variables on the dependent variable, considering their character as qualitative over several modalities, except for the variable “Diff”, which is binary.

Due to our small sample size (50 questions were usable for testing the set of variables for allin, and 59 for allex) and the high number of independent variables and control variables over five modalities, the number of iterations for the software to calculate is so high that it may produce unexpected results. A first, quick solution might be to use different software to Stata, since software may use different algorithms. However, the problem could not be resolved in this way. Different solutions might thus be envisaged that we will explain in the following section.

3.2.1.3.2. Solutions envisaged

There are four main paths to resolving the problem posed. They are presented in order of priority. That is, if solution 1 is satisfactory, then it will not be necessary to test the other solutions. The order of priority is set so as to privilege the solutions which allow us to lose as little information as possible, while still treating the different variables in an appropriate fashion according to their nature. The possible solutions are as follows:

1. 1) reducing the number of modalities to calculate per independent and control variable, by regrouping the modalities when it makes sense based on their frequencies;
2. 2) performing multiple regressions; each only involving one part of the variables instead of integrating all the variables into a single regression. In view of our theoretical approach, we might then envisage carrying out three regressions: one with the contractual variables, one with the knowledge-based variables and a third regrouping the variables stemming from hypotheses comprising two subhypotheses (one based on a contractual argument, the other on a knowledge-based argument);
3. 3) a combination of (1) and (2);
4. 4) finally, if the problem still persists, a solution might be to test, for some independent variables or control variables, whether treating them as quantitative instead of qualitative with several modalities does not significantly affect the final result. If this is the case, one might then envisage making tests dealing with the whole set of independent and control variables as quantitative and only retaining the significant variables. We would then re-conduct the regression, treating the remaining variables as qualitative over several modalities so as to be able to interpret them more correctly. This solution is, however, only acceptable if (a) the quantitative version gives results similar to those of the qualitative version over several modalities and if (b) the number of non-significant variables to be removed from the model is high enough that the software can perform the calculations treating the remaining independent and control variables as qualitative over several modalities.

3.2.1.3.3. Final scenario chosen

We begin by testing the first solution and stopping there if it allows us to surmount the difficulties encountered. On the basis of the frequency tables for the modalities of the independent variables and control variables, we began by regrouping into a single category the negative modalities (in terms of the evaluation of impact) which were less frequent over the whole set of variables (1 and 2), as well as the positive modalities (4 and 5), where it made sense to do so. However, this was not enough for Stata to be able to carry out the regression for intreg as for the ordered logit/probit. Despite the loss of information, we had to strictly regroup all the positive modalities on one side and negatives on the other side for all our variables for the software to be able to carry out the calculations⁴. This was necessary for all our independent and control variables except for the “Diff” variable which is binary.

These regroupings were sufficient to find a solution for intreg, the regression by intervals. In contrast, for ordered logit/probit, we had to move on to supplementary regroupings or proceed with three regressions.

Moreover, even if it is possible in principle to econometrically model our problem by means of an ordered logit or an ordered probit, we must nevertheless ensure that our data do not reject the assumption of equality of slopes (“parallel regression assumption”, or “proportional odds assumption” in the case of logit). In fact, the two models are based on the strong assumption that the coefficients of the variables are identical, independent of the level of the dependent variable, while the constants differ [LON 01, pp. 150–152; ROU 09, p. 54]. To test whether our data satisfy this hypothesis, two tests can be performed [LON 01, p. 151]. These are the Wolfe and Gould test [WOL 98] and the Brant test [BRA 90]. The first verifies the hypothesis through the model considered as a whole. The second, Brant’s, goes further, verifying the hypothesis for each variable taken individually. The Brant test cannot be carried out on our data since the modalities of our variables are too sparsely populated and thus the test does not apply.

The first test shows us that when we chose to consolidate even more modalities so as to carry out a single regression on our data, the assumption of equality of slopes is not violated. In the case where we use three regressions while preserving several modalities, two of these regressions do not satisfy the assumption of equality of slopes. In which case, we cannot then use ordered logit or probit for these regressions [LON 01, p. 152]. We thus decided to not retain the possibility of resorting to three regressions, even though it is possible to apply in the case where the results lead us to reject the hypothesis of equality of slopes, some models of ordered logit and probit being designed to not have to meet this hypothesis. These are the models [LON 01, pp. 168–170]:

• – generalized ordered logit;
• – stereotype ordered regression model;
• – continuation ratio model.

This solution is however excluded, as decomposing our principal regression into three regressions does not allow us to make a direct comparison of our independent variables over the dependent variable. Further, that would require in our case a decomposition into a “normal” ordered logit/probit for the regression satisfying the assumption of equality of slopes, and two adapted logit/probits so as not to have to satisfy the hypothesis of equality of slopes, which would make it difficult to compare results. Moreover, since the application of ordered probit on the whole set of our independent variables, with a more concentrated regrouping of the modalities, leads to violating the assumption of normality of residuals, we will continue only with regression by intervals. It should however be noted that, independently of the difficulties encountered with implementation of ordered logit/probit, regression by intervals seems the most appropriate for the dependent variable⁵, given its characteristics.

To use the intreg function, it is necessary to define left- and right-censoring for the intervals of our dependent variables (allin and allex)⁶. Having done this, we can in principle proceed to the regression. Before continuing, however, it is necessary to check certain assumptions:

• – the absence of multicollinearity between the independent and control variables;
• – the assumption of normality of residuals;
• – the assumption of homoskedasticity of residuals.

Note that, because of the small size of the sample and the fact that these observations are divided into several modalities for the independent variables, we will see that even after proceeding to consolidation of the positive and negative modalities for our independent variables, some categories are still too sparsely populated (sometimes with only one or two observations). Inevitably, this skews our results and led to the rejection of the three hypotheses. Accordingly, before implementing the final regression, we took the decision to only compare the modality corresponding to the hypothesis being tested to the consolidated set of all the other modalities. The results of the three tests of hypotheses which follow are based on these regroupings.

Verification of the absence of multicollinearity

“Collinearity” means that two variables are, to a certain extent, linear representations of each other. We speak of multicollinearity when more than two variables are involved. The higher the degree of multicollinearity, the more it may disrupt the estimation of the coefficients of the model and lead to biased conclusions. It is thus necessary to ensure that the problem is not too great.

The test which allows us to verify the absence of multicollinearity (or, rather, whether it is at an acceptable level) between the variables consists of calculating the variance inflation factors (VIFs) [HOE 70] of the different independent and control variables. It measures the increase in the error term of the model due to the correlation of one variable with the others. The higher the VIF, the stronger the multicollinearity. The idea of calculating the VIF is as follows: each independent and control variable is regressed on the others starting with a linear regression. The coefficient of determination (R²) of each of these regressions then indicates whether there is a linear relationship; in that case, R² will be equal to 1. The tolerance of these models is expressed by (1 – R²) [GUJ 04, p. 356]. For example, when R² = 1, there thus being a linear relationship between two or more variables, the degree of tolerance is 0 (1-1). The VIF is equal to 1/(1 – R²). The higher the VIF, the greater the amount of multicollinearity. If we go back to the example of an R² equal to 1, the VIF in fact tends toward infinity (1/0). It is useful to declare the presence of multicollinearity when one VIF is greater than or equal to 10 [CHA 06, p. 238] and/or the average of the VIFs is greater or equal to 2 [DE 12, p. 6]. Alternatively, it is also possible to detect multicollinearity via the matrix of correlations [DE 12, p. 6]. Here, we present the calculation of the VIFs.

Figure 3.2. Test of multicollinearity over the variables treated as qualitative based on the calculation of the VIFs for allin

In calculating the VIFs for our independent and control variables, by definition, the VIFs are generated by modality for each variable⁷. Figure 3.2 gives an overview of the VIFs per modality per variable, as an example, for our data from the allin regression. Note that this test requires us to perform a linear regression on our data. The intreg function does not allow the software to calculate the VIFs.

According to the table, we do not find any VIFs greater than or equal to 10 (the highest value is 4.55). On the other hand, the average of the VIFs is slightly more than 2 (for the VIFs of our model, it equals 2.83). Even if, often, only the threshold of 10 is used, which leads to conclude the absence of significant multicollinearity for our variables, the average VIF indicates a slight presence of multicollinearity. The same applies for the data of the allex regression. We thus conclude that the coefficients of our variables and of their modalities run the risk of being biased and that they should be interpreted cautiously.

Verification of the assumption of normality of residuals

Intreg (just like the MCO linear regression) assumes that the residuals follow a normal distribution. Thus, we must verify that our data satisfy this hypothesis. Normality of the residuals ensures that the p-values of the t-test and the F-test are not biased⁸. Figure 3.3 shows the results of the test for the allin regression.

Figure 3.3. Testing the normality of residuals for allin

The null hypothesis to be tested is that of the normality of the residuals. The results of the test lead us to not reject the hypothesis of normality of the residuals for allin. The same applies for the data of the allex regression.

Verification of the absence of heteroskedasticity

Classical linear regression, like regression by intervals, supposes that the variance of the error term, of which we have just verified the normality, is constant, in which case we speak of homoskedasticity (homogeneous dispersal). Otherwise, we are in the presence of heteroskedasticity (“heterogeneous dispersal”). The presence of heteroskedasticity consequently means that the estimators of the regression may be biased. Interpreting the data without taking this problem into account may therefore lead to erroneous conclusions.

Note that the test for heteroskedasticity⁹ can only be carried out on a linear regression, and not on intreg or a tobit in general. The test is thus carried out based on a linear regression over our categorical data. For allin, it gives the following result (Figure 3.4).

Figure 3.4. Test of heteroskedasticity for the linear regression over allin

This involves the Breusch–Pagan test that tests the probability of rejecting the null hypothesis of homoskedasticity. Given the result obtained, we cannot reject this hypothesis and we conclude the absence of heteroskedasticity for our data. The same applies for the data of the allex regression.

We will now proceed to the interpretation of the results. However, we must be aware of the relative fragility of these results due to:

• – the small size of our sample, of the distribution of these observations over different modalities for our independent variables and of a high number of variables to test;
• – The possible presence of multicollinearity, even if it seems close to the acceptable limits;
• – the approximation made for the test for heteroskedasticity.

3.2.1.4. Results of the study and interpretation

We first present the results of the regression on the intra alliances data (section 3.2.1.4.1), then those of the regression on the extra alliances data (section 3.2.1.4.2). The last section is devoted to a comparison of the results of the two regressions (section 3.2.1.4.3).

3.2.1.4.1. Regression on the intra alliances data

To recap, out of the 77 PEFs that replied to the questionnaire, 60 indicated that they were involved with the formation of intraportfolio alliances. This phenomenon involved, in a little less than half the cases, 10% or less of the companies that they supported. It mainly involved alliances motivated by the establishment of customer–supplier relations, exchanges of organizational practices or the international development of companies. Nearly a quarter of these alliances remained informal. Apart from that, they mostly took the form of sales agreements. The PEFs indicated that they mainly intervened by facilitating the first exchanges between the partners in alliances, raising the idea of the alliance and providing contacts. Out of the 60 questionnaires completed by the PEFs involved in the formation of intraportfolio alliances, 50 turned out to be usable for the application of the regression.

We apply the regression by intervals for the variable allin. This regression takes the following form¹⁰:

where:

• – : the values observed by intervals for the allin data (intraportfolio alliances);
• – Rp: reputation of the PEF;
• – Ple: presence of a competitiveness cluster;
• – État: presence of the State;
• – Cot: publicly traded status of the companies taking part in the alliance;
• – Spse: sectoral specialization of the PEF;
• – rgInt: regional/international specialization of the PEF’s investments;
• – Diff: the fact that the formation of alliances constitutes a point of differentiation for the PEF on the private equity market;
• – Nbrepart: number of companies managed by the investment manager;
• – Ca: the number of seats held on strategic boards (board of directors, supervisory board or other strategic boards) by the PEF;
• – Spfin: specialization of the PEF’s investments in terms of the stage of development of the SMEs being supported (startup, venture, growth, turnaround, etc.);
• – indus: membership of the PEF in an industrial group or company;
• – finan: membership of the PEF in a financial group (for example a bank);
• – Exp: the experience of the director of investment;
• – Nbresyn: the number of partners in syndication with the PEF;
• – ε_i: the error term.

Note that the reference base¹¹ from which the coefficients of the other modalities of the different independent variables are calculated is not always the same. Sometimes it is category 3, sometimes category 2 or even category 0¹². The reference base is chosen variable-by-variable, in order to be able to read the coefficient of the modality in a direct link with the research hypothesis. For example, for the variable Rp, we predict a positive impact of the PEF’s reputation on the formation of alliances. We thus wish to read at least the coefficient of modality 4, corresponding to this positive impact. So we oppose it to the other modalities which, after consolidation, are included in modality 3.

Figure 3.5 gives an overview of the results of the regression by intervals.

The χ² probability of 0.15 indicates that this model is not statistically significant at the usual minimum level of 10%. At lower right, we see that none of our observations were censored, either to right or to left. The whole set of observations were “censored” by intervals.

Figure 3.5. Application of the regression by intervals for allin

The table itself shows, for each independent variable, the coefficients obtained for their different modalities. The interpretation of these coefficients is performed in the same way as for a linear regression. However, the table does not indicate the overall coefficient for the set of different modalities of the variables¹³. To do this, we must proceed to complementary tests, to see which are the variables that are globally significant¹⁴. We will then be able to assess the impact of the different modalities for the significant variables.

The overall coefficients for the independent variables (after application of the contrast command) are shown in Figure 3.6.

Figure 3.6. Overall significance of the independent variables for allin

In this figure, we see that only the variables Etat and maj appear to be significant at the 5% level. The Diff and rgInt variables are significant at the 10% level.

As the overall significance of the model is greater than the level of 10% (χ² of 0.15), we tried to improve it by removing the less significant variables. Removing the variable Cot (the least significant variable) had virtually no impact on the coefficients of the other variables and the overall significance of the model remained higher than the level of 10% (0.12). Removing the variable Rp (the least significant after Cot) had no more influence on the significance of the other variables and allowed the model to be significant at the level of 10% (χ² of 0.1). Pursuing this approach led to then eliminating the variable Nbrepart, which brought the overall significance of the model to under the 10% level (χ² of 0.07). However, in this case, the coefficients of the remaining variables were affected.

With the 10% level significance having been reached for the model shorn of the two variables Cot and Rp, it was left for us to interpret the results from this model. However, in light of the weak impact on the overall model of the removal of these two variables, we decided to interpret the results given in Figure 3.3, including all the variables. The bias resulting from this choice was likely low.

The only variables appearing significant at the 5% level were the variables Etat and maj. We start with the variable maj (taking majority stakes). We had predicted a positive relationship between taking majority stake and forming intraportfolio alliances. For the variable maj, two modalities are reported in Figure 3.3: modalities 3 (“neutral impact”) and 4 (“positive impact”). The coefficients associated with these two modalities are calculated in relation to the modalities “not relevant” and “negative impact”, which are de facto consolidated. Modalities 3 and 4 both appear to be significant at the 5% level.

The interpretation of these modalities is as follows: compared to those PEFs who were not involved in or indicated a negative impact of taking majority stakes on the formation of intraportfolio alliances, companies involved with PEFs who mentioned a neutral impact (modality 3) or a positive impact (modality 4) formed a lesser percentage of intraportfolio alliances (the coefficients of modalities 3 and 4 are negative). This result is contrary to the hypothesis proposed. This latter is therefore rejected.

For the variable Etat, we also predicted a positive impact of the presence of the State on the formation of alliances. Figure 3.5 again shows modalities 3 and 4 for the variable Etat. The coefficients of these modalities are calculated in opposition to the remaining modalities, “not relevant” and “negative impact”. Modality 3 appears insignificant at the 5% level. On the other hand, modality 4 appears significant at the 1% level. In comparison with the PEFs who replied “not relevant” or “negative impact” for the presence of the State on the formation of intraportfolio alliances, PEFs who indicated a positive impact tended to have a lower percentage of companies in their portfolio forming alliances of the intra type. Again, this leads us to reject the proposed hypothesis.

The variable Diff (differentiation), significant at the 10% level, has a coefficient which conforms with what was predicted by the theory: there is a positive link between the fact that a PEF differentiates itself on the private equity market by the formation of alliances, and the formation of alliances for its portfolio companies, which conforms to intuition.

The same is true for the variable rgInt, which measures the impact of regional/international specialization of the PEF’s investments on the formation of alliances. The variable is significant at the 10% level. The interpretation of the coefficient of modality 4 (significant at the 10% level) apparently supports the impact predicted by the theory. Thus, compared to the PEFs who responded “not relevant”, “negative impact” or “neutral impact”, the PEFs who responded “positive impact” (modality 4) show a higher percentage of companies in their portfolio forming intra alliances. However, we cannot conclude from this that the coefficient of the variable rgInt supports the predicted theoretical link. According to the latter, remember, on the one hand, a PEF investing primarily at the regional level would have a positive impact on the formation of sectoral and often regional alliances; while on the other, a PEF investing in various countries would tend to be involved in the formation of a greater number of alliances with the goal of international development of the supported companies. Now, we only have information concerning the impact on the formation of intra/extra alliances, without being able to distinguish the sectoral or international character of the alliance.

We can see that modality 4 of the variable finan seems equally significant in the table in Figure 3.5. However, the variable as a whole is not significant, as can be seen in Figure 3.6.

Table 3.1 sums up the significant results obtained, with confirmation or rejection of the hypothesis.

Table 3.1. Summary of the significant results obtained for allin with confirmation or rejection of the hypothesis

Allin regression		Level of significance	Sign of coefficient	Expected sign	Acceptance/ rejection of the hypothesis
Model as a whole		10% (after removal of the insignificant variables Cot and Rp)
Hypothesis	Independent variables
H3	Etat	5%		+	Rejection of hypothesis
	modality 3 (“neutral impact”)	5%	–
	modality 4 (“positive impact”)	5%	–
H4	maj	5%		+	Rejection of hypothesis
	modality 3 (“neutral impact”)	5%	–
	modality 4 (“positive impact”)	1%	–
H7b	rgInt	10%		+	Tends to support the hypothesis
	modality 4 (“positive impact”)	10%	+
H9	Diff	10%	+	+	Acceptance of the hypothesis
Etat: the presence of the State in the PEF’s capital. Maj: taking majority stakes. rgInt: regional/international specialization of the PEF. Diff: the motivation of differentiation on the private equity market by the PEF.

3.2.1.4.2. Regression on the extra alliances data

71 out of the 77 PEFs who responded to the questionnaire indicated that their portfolio companies formed extraportfolio alliances. For almost a third of these PEFs, this comprised between 11% and 50% of the companies they supported. A quarter of these companies indicated that more than half of the companies in their portfolio formed alliances of the extra type. These alliances thus seem more frequent than in the case of intraportfolio alliances.

Extraportfolio alliances mainly have the goal of developing customer–supplier relations. From this comes the joint development of new products/services and of the company at international level. These alliances may take all kinds of forms. In more than a quarter of cases, it involves sales agreements. The PEFs reported being involved mainly by contributing contacts, facilitating the first exchanges between future partners, or by introducing the idea of the alliance. Occasionally, they may determine the choice of partner in the alliance or intervene in negotiating the terms of the alliance contract. Out of the 71 questionnaires that dealt with the formation of extraportfolio alliances, 59 were usable.

We apply the regression by intervals for the dependent variable allex. This regression takes the following form¹⁵:

where:

• – : the values observed by intervals on the data of allex (extraportfolio alliances);
• – Rp: reputation of the PEF;.
• – Ple: presence of a competitiveness cluster;
• – État: presence of the State;
• – Cot: publicly traded status of the companies taking part in the alliance;
• – Spse: sectoral specialization of the PEF;
• – rgInt: regional/international specialization of the PEF’s investments;
• – Diff: the fact that the formation of alliances constitutes a point of differentiation for the PEF on the private equity market;
• – Nbrepart: number of companies managed by the investment director;
• – Ca: the number of seats held on strategic boards (board of directors, supervisory board or other strategic boards) by the PEF;
• – Spfin: specialization of the PEF’s investments in terms of the stage of development of the SMEs being supported (tartup, risk, development, turnaround, etc.);
• – indus: membership of the PEF in an industrial group or company;
• – finan: membership of the PEF in a financial group (for example a bank);
• – Exp: the experience of the director of investment;
• – Nbresyn: the number of partners in syndication with the PEF;
• – ε_i: the error term.

Figure 3.7 gives an overview of the values of the coefficients of the modalities of the independent variables.

The regression as a whole is this time significant at the 5% level with a χ² probability of 0.05. To the lower right, we can read that none of our observations is censored to left or to right. The whole set of observations is essentially “censored” by intervals. Before interpreting the values of the table, we again concern ourselves with the overall significance of the independent variables (Figure 3.8).

Four variables appear to be significant overall (without distinguishing the modalities). These are variables Rp (at 1% level), Spsec (at 1% level), Etat (at 5% level) and Diff (at 5% level).

Figure 3.7. Application of the regression by intervals for allex

Figure 3.8. Overall significance of the independent variables for allex

Although the model is significant at the 5% level, we have also attempted to see if this significance could be substantially improved by removing the less significant variables. Thus, by removing successively the variables Cot (publicly traded status of the companies forming the alliance), Ca (the presence of the PEF in the strategic boards such as the board of directors or the supervisory board), Ple (the presence of a competitiveness cluster), Exp (the experience of the director of investment), Nbresyn (the number of partners in syndication of the PEF) and finan (the specialization of the PEF’s investments according to the stage of development of the supported companies), the global model becomes significant at the 1% level (χ² of 0.0048), without these removals significantly affecting the other variables.

As for the intraregression, given these results, we base our interpretation on Figure 3.7, which presents the values of the coefficients associated with the modalities of the different variables without removing any variable.

For the variable Rp, the coefficient of modality 4 appears significant at the 1% level. This allows us to conclude that the PEFs who reported a positive impact for reputation on the formation of extraportfolio alliances create a higher percentage of these for the companies in their portfolio, in comparison to the PEFs who replied “not relevant”, “negative impact” or “neutral impact”, which confirms the hypothesis.

For the variable Etat, modality 4 appears significant (5% level). According to its coefficient, we can conclude that, in comparison to the PEFs who responded “not concerned” or “negative impact” to the presence of the State in the formation of extraportfolio alliances, the PEFs who indicated a positive impact tend toward a lower percentage of companies in their portfolio forming this type of alliance. This result contradicts the hypothesis advanced.

Modalities 3 and 4 of the Spsec variable seem significant (at levels of 1% and 5%, respectively). Compared to the PEFs who indicated “not relevant” or “negative impact” to sectoral specialization of the PEF on the formation of extra alliances, those who mentioned a neutral or positive impact report more companies in their portfolio forming extraportfolio alliances. However, we must make the same reservation as for the rgInt variable when interpreting the results for the intra data. Overall, the positive relationship between sectoral specialization of the PEF and the formation of intra alliances goes in the direction expected at the theoretical level. However, though we predicted a positive link between the sectoral specialization of the PEF and the formation of a certain type of alliance (sectoral alliances), we do not possess the information allowing us to really test this hypothesis.

The variable Diff is significant at the 5% level and, according to its coefficient, the PEFs for which the formation of alliances is a means of differentiation on the private equity market show a higher level of supported companies forming extra alliances, compared to other PEFs. This result conforms to the predicted link.

Note that modality 4 of the variable Nbrepart also appears significant in the table in Figure 3.7. However, the variable as a whole is not significant, as can be seen in Figure 3.8.

Table 3.2 sums up the significant results obtained, with confirmation or rejection of the hypothesis.

Table 3.2. Summary of the significant results obtained for allex with confirmation or rejection of the hypothesis

Allex regression		Level of significance	Sign of coefficient	Expected sign	Acceptance/ rejection of the hypothesis
Model as a whole		5% (or 1% after removal of insignificant variables)
Hypothesis	Independent variables
H1	Rp	1%	+	+	Acceptance of the hypothesis
	modality 4 (“positive impact”)	1%
H3	Etat	5%	–	+	Rejection of hypothesis
	modality 4 (“positive impact”)	5%
H7a	Spsec	1%	+	+	Tends to support the hypothesis
	modality 3 (“neutral impact”)	1%
	modality 4 (“positive impact”)	5%	+
H9	Diff	5%	+	+	Acceptance of the hypothesis
Rp: the reputation of the PEF. Etat: the presence of the State in the PEF’s capital. Spsec: the sectoral specialization of the PEF. Diff: the motivation of differentiation on the private equity market by the PEF.

3.2.1.4.3. Comparison of the roles of PEFs between intra and extraportfolio alliances

The comparison of the roles of PEFs according to the type of alliances formed, intra or extra, confronts several limitations:

• – a difference of degree of significance between the allin and allex regression (respectively, significant at the 10% level after removing two variables and at the 5% level, even 1% after removing several variables);
• – the potential presence of multicollinearity for our data that may bias the results, even if it seems that this multicollinearity is relatively weak;
• – the small size of our samples and the large number of variables, which suggests a weak robustness of the results obtained.

We will, nevertheless, try to draw up a conclusion and to reconcile the results obtained. Remember that, as the questionnaire intended for SMEs never took place, the results that we are interpreting compare, on the one hand, the perception that the PEFs have of the impact of the different independent variables on the activity of forming alliances among their portfolio-companies, and, on the other hand, the real percentage of their portfolio-companies who have formed an intra- or extraportfolio alliance, as indicated by the PEF.

For intraportfolio alliances, only two variables appear to be significant: Etat (state involvement) and maj (the PEF taking majority stakes). The variable maj appeared significant for the regression on allin, and not significant for the allex regression. The first results conform to the hypothesis advanced; an assumed positive effect of taking majority stakes on the formation of intraportfolio alliances. On the contrary, the expected sign of maj on the formation of intra/extra alliances is contrary to what we expected.

The variable Etat is significant in both allin and allex regressions and shows the same effect for both regressions – contrary to what was predicted by the theory. This result thus reverses the hypothesis presented. In addition, note that the hypothesis predicted a positive impact of the presence of the State or of a competitiveness cluster on the formation of alliances. Now, only the variable Etat seemed significant in the two regressions, but with a reverse relationship to what was expected.

As for the variable Ple (presence of a competitiveness cluster), it was not significant for either regression. The negative impact of the presence of the State on the formation of alliances, nevertheless, seems curious. Even if it does not have the positive impact expected, we would have then imagined a neutral rather than a negative impact.

The variable Diff (differentiation of the PEF on the private equity market), significant in the allin regression at the 10% level and in the allex regression at 5%, shows, in both cases, the expected sign implying a positive link between the strategy of differentiation through the formation of alliances by the PEF on the private equity market and the percentage of supported companies forming extraportfolio alliances.

One last variable seems significant for the allin regression: the variable rgInt (regional/international specialization of the PEFs) at the 10% level. Its sign (positive) agrees with the hypothesis presented, even if we must be very careful in interpreting this variable in the light of the hypothesis presented, as it does not really allow us to test that hypothesis. The latter actually predicts a positive link between regional or international specialization in the PEF’s investments and the formation of alliances of, respectively, regional or international scope. Now, we can only interpret the impact of the variable on the formation of intra- or extraportfolio alliances. Moreover, we did not predict differences in the case of formation of intra or extraportfolio alliances. Now, the variable appears significant for the allin regression with a positive sign, and not significant for the allex regression with a negative sign.

The variable Rp (the reputation of the PEF) appears significant in the allex regression. It is quite interesting to compare this with its significance for the intra regression. There it did not appear as significant, whereas in the extra regression, Rp is significant and indicates a positive link with the formation of extraportfolio alliances. However, in the intra regression, the sign of Rp is also positive. This reinforces, at least in part, the hypothesis that predicted a positive link between the reputation of the PEF and the formation of alliances, as well as the arguments advanced that the link between the reputation of the SCI and the formation of alliances would be more direct in the case of forming extraportfolio alliances.

Finally, the variable Spsec (sectoral specialization of the PEF) indicates a positive link between sectoral specialization of the PEF and the formation of extraportfolio alliances. Positive coefficients are found both for intra and for extraportfolio alliances, although the variable does not appear to be significant. We may thus conclude a favorable tendency toward the hypothesis even if, for reasons similar to those expressed for the rgInt variable discussed above, we cannot entirely establish the acceptance or the rejection of the hypothesis presented, given that we have not really been able to test it. This latter predicted a positive link between sectoral specialization of the PEF and the formation of sectoral alliances. However, we do not possess information about the formation of sectoral alliances.

Table 3.3 summarizes these different points. It presents the results for the significant variables of the allin and allex regressions, compares these results in terms of significance and the sign of the variables, and indicates whether the theoretical hypothesis is rejected or not.

Table 3.3. Summary of the results for the significant variables of the two regressions allin and extra

Hypothesis	Regression	Level of significance		Sign of coefficient		Expected sign	Acceptance/ rejection of hypothesis
	Independent variables	Allin	Allex	Allin	Allex
H4	maj	5%				+	Rejection of hypothesis
	modality 3 (“neutral impact”)	5%	–	–	–
	modality 4 (“positive impact”)	1%	–	–	–
H3	Etat	5%	5%			+	Rejection of hypothesis
	modality 3 (“neutral impact”)	5%	–	–	+
	modality 4 (“positive impact”)	5%	5%	–	–
H7b	rgInt	10%	–			+	Unable to reach a conclusion
	modality 4 (“positive impact”)	10%	–	+	–
H9	Diff	10%	5%	+	+	+	Acceptance of the hypothesis
H1	Rp	–	1%			+	Acceptance of the hypothesis
	modality 4 (“positive impact”)	–	1%	+	+
	Spsec	–	1%			+	Tends to support the hypothesis
	modality 3 (“neutral impact”)	–	1 %	+	+
	modality 4 (“positive impact”)	–	5 %	+	+
Maj: taking majority stakes. Etat: presence of the State in the PEF’s capital. rgInt: regional/international specialization of the PEF. Diff: the motivation of differentiation on the private equity market of the PEF. Rp: the reputation of the PEF. Spsec: sectoral specialization of the PEF.

It remains to compare the overall results of the allin and allex regressions to test Hypothesis 10. According to the latter, PEFs play more important roles in the case of formation of intraportfolio alliances than extraportfolio alliances. Our results do not allow us to confirm this hypothesis.

Finally, taking account of the fragility of the results obtained, it seems that we must be very careful as to the conclusions to be drawn. This fragility further justifies the use of multiple case studies to reinforce or not the conclusions of the econometric study.

The entirely provisional conclusions, which it seems possible to draw, however, seem to be as follows:

• – PEFs who differentiate themselves on the private equity market by the formation of alliances have a higher level of supported companies who form extraportfolio alliances and, possibly, also intraportfolio alliances;
• – the reputation of PEFs has a positive impact on the formation of extraportfolio alliances;
• – the presence of the State seems to have a negative impact on the formation of alliances;
• – based on only the results of the econometric study, we cannot conclude that greater roles are played by PEFs in the case of the formation of intraportfolio alliances (as opposed to extraportfolio alliances).

3.2.2. Multiple case study

We begin by explaining the purpose and nature of the study (section 3.2.2.1) as well as the procedure for choosing the fields for the cases (section 3.2.2.2), before presenting them (the fields and the cases). We will then describe the data sources to which we had access (section 3.2.2.3), then explain the way in which we carried out our analysis (section 3.2.2.4). The analysis itself is found in section 3.2.2.5. We finish with the conclusions that we were able to draw from this study (section 3.2.2.6).

3.2.2.1. The purpose and nature of the case study

In the statistical and econometric analysis, we tested the general nature of the hypotheses derived from our theoretical concept, the parent population being French PEFs. This case study will now allow us to proceed to an analytical test of generalization, that is, to see if empirical results support a previously established theory, to determine whether the mechanisms of causality, advanced in theory, are plausible. In other words, the statistical and econometric study allowed us to verify empirically that there is indeed a correlation in the predicted direction between the independent variables and the dependent variable. It now remains to verify the plausibility of the causal mechanisms which explains the links between these variables. Using a case study methodology brings two additional benefits. First, it allows us to verify the explanations advanced from the point of view of the SMEs forming an alliance, as well as the point of view of the PEFs. This is not possible using a statistical approach, because it involves two different visions for the same, statistically identifiable empirical result. Taking into account different points of view allows a triangulation of data [GRA 07, p. 28; GIB 10, pp. 712–713]. Second, it offers the possibility, by combining it with the statistical and econometric study, of triangulating the methods used.

The study consists of a multiple case study with explanatory design¹⁶ which is also “embedded” [YIN 09, p. 59]. “Multiple” means it includes more than one case (eight in total), in order to increase both its internal and external validity, through the possibility of replication. This allows us, in the end, to be able to claim with greater certainty an analytical generalization of the results of the study. The term “embedded” signifies that we are taking into account different perspectives for the same unit of analysis. This is justified for two reasons. On the one hand, we take into account two points of view in our theoretical analysis: that of the SMEs forming the alliance, and that of the PEF. As opposed to statistical and econometric studies, case studies allow us to take account of such an aspect, and this is one of the reasons why we use them. In our study, therefore, we question the managers of companies supported by private equity as well as the PEFs for the same alliance. On the other hand, collecting different points of view allows us, again, to reinforce the robustness of our results if the results from the different points of view converge.

Multiple and embedded case studies, according to Yin [YIN 09, p. 59], are the most complex to design, as the selection of cases must be made in such a way that all the criteria arising from the theoretical concept are represented. To qualify as a literal replication, it is necessary to choose “fields of study/analysis” which may contain several similar cases. In our case, the field of analysis consists of a PEF and its portfolio of investments. It is necessary, moreover, that the PEF can provide at least two examples of alliances being formed (or in the process of formation) by its portfolio companies. This constitutes a first difficulty, as it means that it is not only necessary to find French PEFs presenting some of the sought-after characteristics arising from our research hypotheses, but also to ensure in advance that each field contains two similar cases, in order to enable literal replication not only between the fields, but also within the same field. The examples of alliance formation with which these PEFs provide us therefore constitute the cases, properly speaking. For example, let us take the variable “taking majority stakes” which arises from our research Hypothesis 4. We must then choose two PEFs taking majority stakes, to be able to proceed to replication between the fields. Moreover, each of these PEFs must be able to provide us with two examples of formation of alliances, in order to be able to proceed to literal replication within a single field. This all becomes more complicated when we realize that, while using as few fields (PEFs) as possible, the whole set of variables arising from our hypotheses must be represented in this way.

The study is dynamic in nature. The period of study varies from case to case, as it extends over the time of the formation of the alliance which may vary in different cases. Some of the alliances studied were, furthermore, still in the course of formation at the end of our study.

3.2.2.2. Selection process and presentation of the cases and fields of study

This section is devoted to presenting the selection process of the cases and fields of study, before describing them.

The selection process of the cases and fields of study

The selection of cases is a theoretical and intentional question, guided by the theoretical concept [SCA 90, p. 273; DAV 05]. The latter includes either the variables relevant to the companies forming the alliance (for example the publicly listed or otherwise status of the company), or the variables relevant to the PEFs, given that we are interested in the role of PEFs in the formation of alliances for their portfolio companies. This means, for example, variables such as their specialization, whether they are established PEFs on the market or young and building their reputation, the number of portfolio companies that they support, whether they take majority or minority stakes, or again, if they advertise their relationship building activities. A “case”, however, represents the formation of an alliance between at least two companies in the presence of a PEF. It is therefore necessary, in our study, to distinguish the “field” from the “case” itself. The chosen PEFs thus constitute the fields of study, and the examples of alliances (for their portfolio companies) that they provide us with, the cases.

Practically, this means that the “fields” for the case studies, that is, the PEFs, are chosen in such a way that the independent variables concerning the PEF’s role in the formation of alliances that stem from our research hypotheses are represented. Ideally, the fields are chosen in such a way as to contrast with each other with respect to one or more of the criteria linked with the research variables, while being comparable apart from these criteria. This then allows us to ensure that the observed results are not distorted by the presence of other variables which may have an impact upon the phenomenon to be studied. Thus, all the criteria relating to PEFs arising from our hypotheses are represented, with their various modalities. This means that, for example for the criterion of “taking majority stakes”, we have chosen at least one PEF which takes majority stakes and at least one PEF which takes minority stakes. Moreover, each criterion is represented twice, in order to allow literal replication. Again, with regard to the variable “taking majority stakes”, this implies that we need two examples of alliances where the PEF that supports the enterprises takes majority stakes.

As we will explain, we have chosen four PEFs among those which we studied, and for each of them two cases of the formation of alliances, in order to be able to carry out a literal replication, not only across the study in general but also for each field. This resulted overall in eight case studies. The actual number of case studies is not very important (contrary to statistical and econometric studies where the sample size of the study is decisive for its significance), even if, in general, the more cases for replication, the greater the degree of certainty for the results observed. Nevertheless, particularly due to the considerable time which the analysis of a case study may take up, a balance must be struck between the time the study requires and the gain associated with the analysis of the case in question [BAX 08, pp. 546–547]. According to Yin [YIN 09, p. 58], if we wish for a higher degree of certainty, a multiple case study comprising five to six cases for replication is considered as offering a good degree of confidence in the results obtained.

The process of contacting French PEFs in the hope of finding adequate fields, and who were ready to collaborate, began from the first year of the research project and ended during the third year. This process proved to be quite long and difficult, but eventually came to fruition. Contact was generally made through an initial email, followed by a first reminder and several telephone calls.

Returning to Table 2.7, the criteria for the PEFs are presented in Table 3.4.

Table 3.4. Variables according to the research hypotheses

Hypothesis	Criterion
H1	PEF possesses a reputation due to a certain age and is being established on the private equity market
H2	A weak reputation as the PEF is still young and is becoming established on the private equity market
H3	The presence of the State, of a regional authority, of a competitive cluster in the capital of the PEF and the legal status of a joint-stock company of venture capitalists or a VCMF(venture capital mutual fund)
H4	The PEF takes majority stakes
H7a	Regional or sectoral PEFs
H7b	PEFs with an international presence
H8a	A large number of portfolio companies to be managed by the investment manager and a large geographic distance separating the PEF from the supported companies
H8b	A low number of portfolio companies to be managed by the investment manager and a small geographic distance separating the PEF from the supported companies
H9	The PEF issues advertising relationship building, creating clubs/forums for meetings
H11a	Holding seats on strategic boards
H11b	The PEF participates in associations

Hypotheses 5, 6 and 10 are not listed in Table 3.4, since they do not include criteria specific to a type of PEF. Hypothesis 5 concerns the impact of alliances previously formed by the SMEs on the formation of a new alliance. Hypothesis 6 focuses attention on the publicly listed or otherwise status of the supported companies. According to Hypothesis 10, the roles played by the PEFs in the formation of alliances should be more important in the case of the formation of intraportfolio alliances, compared to the formation of extraportfolio alliances.

The selected PEFs who agreed to collaborate and act for us as fields for the study of two cases of formation of alliances are (1) Siparex Group, (2) Demeter Partners, (3) Industries et Finances Partenaires (I&FP) and (4) an anonymous PEF. They are presented in the following section.

Table 3.5 gives an overview of the criteria that the four fields of analysis/PEFs satisfy in relation to our research variables.

Table 3.5. Fields of analysis according to the research variables

Hypothesis	Variables/Fields for cases	Anonymous PEF	IFP	Siparex	Demeter
	Cases for literal replication	2	2	2	2
H1/H2	Reputation	Weak	Average	Strong	Strong
	Age of fund	Young	Average	First French PEF	Old
H3	State/region/comp. cluster	Yes	No	No	No
	Legal status	Venture capital company	Fund	Fund	Fund
H4	Stakes	Minority	Majority	Major/minor	Major/minor
H7a	Focus of investments	Sectoral	General	General	Sectoral
H7b	Scope of investments	Regional	National	International	International
H9	Advertising	No	No	Yes	Yes
	Clubs/forums	No	No	Yes	Yes
H11a	Seats on BoD	Yes	Yes	Yes	Yes
H11b	Geographic distance	No	Average	No	Yes

The presentation of the fields of study and the cases

This section is devoted to presenting the four fields of study, that is the four PEFs chosen according to our research variables that were willing to collaborate with us. As previously mentioned, these were Siparex Group (section 3.2.2.2.1), Demeter Partners (section 3.2.2.2.2), I&FP (section 3.2.2.2.3) and an anonymous PEF (section 3.2.2.2.4). Their presentation includes three parts:

• – the description of the PEF in terms of the criteria on the basis of which they were chosen;
• – the PEF’s activity in the formation of alliances in general;
• – the presentation of the cases for study which they provided us with.

3.2.2.2.1. Siparex Group

General context: the presentation of the field (the PEF)

The Siparex Group¹⁷ was chosen for our study because it combines the following distinct criteria: founded in 1977 in Lyon, it was one of the first French PEFs on the French private equity market. Since its creation, it has engaged in more than 700 investments for a total sum of more than a billion euros. It therefore has a certain amount of reputational capital. Moreover, it was the first PEF to set up a social club for its portfolio companies. Club Siparex has existed since the creation of Siparex Group, more than 36 years ago. From the outset, the service of creating links between its portfolio companies through alliances has constituted a way for the PEF to differentiate itself on the private equity market. Siparex Group is a generalist PEF, investing in all sectors, at all stages of development. It is of a national as well as international scale, with a regional structure that allows it to support its portfolio companies through a regional plan. The PEF mainly takes large minority stakes in the companies it supports. Currently, Siparex Group manages more than 240 companies.

Siparex Group does business mainly in France with multiregional locations (six in total in France). It also possesses foreign offices and partners throughout the Euromed zone (Morocco, Tunisia, Egypt) as well as in Italy and Spain.

Legally, Siparex Group takes the form of a venture capital mutual fund (VCMF). There is a management company: Sigefi Private Equity, which has two shareholders: Sigefi Partners, which includes the six directors and seven associate directors, and Siparex Associates, which includes the main partners and underwriters. The management company manages two funds. The first is dedicated to growth capital and leveraged buy-outs. The second, called “proximity-innovation capital”, is targeted at regional support, on the one hand, and supports innovation on the other hand, covering all needs for funding ranging from start-up to buy-outs, while prioritizing investment in growth companies and mature regional companies in the sectors of IT, Internet, mobile devices, software, networks/telecoms and clean tech. Siparex Group therefore covers a wide spectrum of private equity, ranging from start-ups to growth capital and buy-outs, taking minority or large minority stakes. However, it excludes turnaround capital.

There are two logics within Siparex. A logic of specialization of teams with respect to the size of the companies being supported. This is extremely important because, as the portfolio manager put it when interviewed at Siparex Group: “you do not support an SME with a 10 million turnover in the same way as an SME with a 150 million turnover”. Then, there is a logic of territory or proximity. The portfolio manager we talked to said: “Our role is that of an active and non-interventionist shareholder, and we have to be close to the manager. The further away we are, the less well we perform our supporting role, which in general lasts between five and seven years. Within the Siparex group, this logic is shown through our regional offices. These are found in Lyon, Paris, Strasbourg, Nantes, Lille and Limoges”.

Specific context: activity of forming alliances

The following information comes from interviews with one of the investment managers of Siparex Group who participated in our study. Siparex Group systematically intervenes in making links between its portfolio companies through alliances, given that Club Siparex was established specifically for this purpose. Building links between their portfolio companies through alliances is thus one of this PEF’s specificities. Two full-time salaried employees are charged with discovering possible synergies between the PEF’s portfolio companies, through Club Siparex. The latter takes the form of an association chaired by the company manager of one of the portfolio companies. At the same time, the investment manager who supports the companies intervenes in the search for appropriate external partners for alliances, when no adequate partner can be found within the PEF’s investments. Nevertheless, the desire to form an alliance must come from the company manager or, at least, the investment manager must perceive a latent interest from the company manager, which they will then be able to assist and concretize. The concrete idea of the alliance thus often comes from the manager of the company being supported but may equally come from the investment manager. Club Siparex thus allows the building of links between companies which would otherwise have no reason to be in contact.

The vast majority of the alliances formed are informal and are created among all the portfolio companies who do not necessarily operate in the same sectors. These consist of exchanges of organizational practices between companies, the logic of sharing information, sharing experiences and business development. When alliances are formalized, in principle they take the form of customer–supplier relations. Alliances involve companies at different stages of development and in all sectors of activity. In general, Siparex Group intervenes in facilitating the first exchanges and meetings between potential partners. The PEF provides its network through Club Siparex, which also organizes events on major themes, but generally oriented on a macro-economic vision. Company managers of very large groups are also involved, in a more or less symbolic manner.

Alliances are mainly formed between French companies, even though (as previously stated) Siparex Group has bureaus in the Euromed zone, Italy and Spain. The portfolio manager we questioned adds, however: “But, for example, if there is a company that wants to find a Moroccan supplier in a very specific sector, we will ask our counterparts in Morocco if they have any names or contacts, and if so, we’ll put the companies in contact”.

The presentation of cases (the alliances)

Siparex Group offered us the chance to analyze the building of links between their portfolio companies, within a working group organized around e-commerce. The goal of the working group is for the companies to share their vision of business and of expected developments in the market, to exchange technical tools they have been able to put in place within their companies, in order to constitute a benchmark. The role of Siparex Group was a co-ordination role, an animation role and, prior to that, worked to build relationships in order to help the companies have a somewhat similar presentation, so they can provide the same information concerning what they do, the information systems and software that they use, the questions that these pose, the experiences that they would like to share, the problems that they encounter, etc. Siparex Group is a facilitator of alliances of any type. Once links between the companies are built, it is up to the managers to use them. Siparex Group does not intervene in the alliance itself.

Three company managers participating in the working group agreed to cooperate with our study. These were the managers of the companies Sophem, Meseo and Mooviin. The three companies met each other via Siparex Group and the working group over e-commerce. Alliances were formed between Mooviin and Sophem, and between Mooviin and Meseo.

The Mooviin-Sophem alliance: Sophem is an e-commerce sales site for ready-to-wear clothes. Mooviin, originally, was a manufacturer and distributor of furniture. It later shifted to the Internet and has achieved great success. The two company managers met through Siparex Group, which initiated the relationship. The original goal was for the Sophem comany to discuss with the Mooviin company its experience setting up an Enterprise Resource Planning system using the Magento platform, which the two companies shared. The subsequent exchanges, however, gave rise to an alliance between the two companies at the marketing level. They thus planned to set up a gift-book operation, with discount vouchers to link their respective orders. In the long run, the two companies plan to include other regional e-commerce sites, in the area around Annecy, in these joint operations. The manager of Sophem stated: “… they are in the process of setting up an ERP… we know that we absolutely have to go to meet them and see them, because they are on the same technological platform as our e-commerce site. So, they’re on Magento and we think it would be interesting to see how they’ve gone about setting up their ERP, why they did things that way. We want to visit them a little, to look around the place, to help us to find good service providers and have a good approach and, also, to discuss their strategies with them. Because right now, they’re in the furniture area and we’re in ready-to-wear, and thus we have to see if we can combine our expertise or, also, see if we can carry out joint marketing operations. So, we want to meet them. We already met the company manager and the person who looks after the IT project and then, on a second visit, we went to see their web service, for e-commerce and marketing, to see the experiences that they had and see if we could do joint gift-book operations… a gift-book with discount vouchers, and even also including other e-commerce sites outside Siparex, but sites in the region, to engage in joint operations”.

The Mooviin-Meseo alliance: Links between the two companies were initiated by Siparex Group. They were based around the creation of an Internet site. Meseo is a site specializing in the sale of technological products for the home, the garden, the swimming-pool and for well-being. The company was looking to move into furniture, to make it available on its own Internet site. Mooviin, is a specialist in furnishings and home decor, and was also in the process of setting up an Internet site. The alliance thus consisted, on the one hand, of exchanging experiences and, on the other hand, of offering certain Mooviin products on Meseo’s website.

3.2.2.2.2. Demeter Partners

General context: the presentation of the field (the PEF)

Demeter Partners takes minority or large minority stakes in the capital of generally non-listed companies in the eco-industry and eco-energy sectors, with the goal of financing and supporting their development. It was chosen for our study because it combined the distinct criteria of a sectoral PEF, operating a “Demeter Partners Entrepreneurs” club which promotes exchanges and the establishment of synergies between their portfolio companies. This service of building relationships constitutes for Demeter Partners a strategy of differentiation on the private equity market. Demeter Partners takes the form of a VCMF, with investments all over France as well as overseas. It is a well-established PEF on the private equity market, possessing some reputational capital. Although it mainly takes minority stakes, it has also taken some majority stakes (20% of their investments). This usually involves leveraged buy-outs. Their general activity, however, consists of supporting companies in the growth phase with mainly a minority stake (60% of investments). Some stakes in companies in the venture-capital phase (20% of investments) are also involved. Demeter Partners is thus placed as a major player in European Private Equity in the field of eco-industry and eco-energy. Apart from its presence in France (more than 50% of its support goes to French companies), the PEF has a team in Spain and Germany, and has forged partnerships with German, Spanish and English overseas funds involved in the clean-tech field.

The Demeter Partners team consists of four executives, eleven directors of investment and six back-office employees. Demeter Partners manages three funds (Demeter, Demeter 2 and Demeter 3), with about twenty lines of investment for the first two. Demeter 3 is dedicated to supporting companies in the start-up phase in the field of clean-tech.

With regard to the formation of alliances, Demeter Partners, on the one hand, promotes itself in particular by mentioning on its site that it brings “added values through its relationships in the sector and the development of synergies with other portfolio companies”. On the other hand, it offers its portfolio companies the opportunity to join its Demeter Entrepreneurs Club. Created in 2007, it brings together managers of the supported companies and aims to provide for the latter a convivial place for debate and reflection on the environment and sustainable development, to develop industrial and commercial synergies between the portfolio companies of the Demeter funds’ different portfolios, and to give their portfolio companies access to a network of experts and institutional relationships.

Specific context: the activity of forming alliances

Demeter Partners systematically intervenes in the formation of alliances, in particular, through the Demeter Entrepreneurs Club. This mainly involves relationships of the customer–supplier type, in the form of procurement contracts. These contracts generally last a year but may be subsequently renewed. Cooperative ventures between companies have also been set up, with the goal to respond to common and/or foreign calls for tenders, thus allowing reduced travel costs. But in that case, it involves jointly replying to a one-off offer.

Since 2007, the Demeter Entrepreneurs Club has regularly brought together – generally three or four times a year – managers of companies in the Demeter portfolio. This club aims to be a convivial place for debate and reflection on the themes of the environment and sustainable development, to develop industrial and commercial synergies between the portfolio companies of the Demeter funds’ different portfolios, and to give their portfolio companies access to a network of experts and institutional relationships. Club meetings typically revolve around sectoral themes, or around export financing (receivables). It is a chance for new entrants to introduce themselves and then for all the participants to discuss. The goal is to create synergies, by establishing commercial links and sharing knowledge through exchanging good practices or contacts.

According to the investment manager we talked to, the principal role of Demeter Partners in the formation of alliances is that of an intermediary. They explained as follows: “We play an intermediary role. The idea of building a relationship generally comes from Demeter Partners. Then, Demeter Partners puts the company managers into contact and leaves them to sort their alliance relationship out, except in the case of difficulties… Already, before taking a stake in a company, we systematically examine who else in the portfolio they could work with. We then discuss this with the target companies at the first meeting, when we present the companies in our portfolio. This is part of the marketing. Next, this may lead to a meeting between the companies, either privately, or at the salon that we hold once a year, or at the Demeter Partners Entrepreneurs Club, for example. After that, we play two roles. The first is more important; it involves establishing trust. Being a perennial fund, we can reinvest money into companies in a fragile situation. The second, less important role, is that we can intervene in the negotiation of the contract. That is, generally, after a first contact our intervention ends there, except if there is a problem. We can then intervene to smooth over the edges. For example, if one party tries to raise the price before breaking the contract, we talk to them to try to put things back in order”.

The presentation of cases (the alliances)

Demeter Partners gives us two examples of alliances. These are the alliances between IES and Comarth, and between Panosol and Eurener. These are two examples of the types of alliances typically formed under Demeter Partners. They take the form of a customer–supplier type relationship. One came to fruition, while the other was dissolved once Demeter Partners withdrew from the capital of the supported companies.

The Panosol–Eurener alliance: currently, Demeter Partners is no longer involved in the capital of these companies. When the alliance was formed, Demeter Partners held shares in both Eurener and Panosol. The PEF built links between them, which resulted in the formation of a customer–supplier type of alliance. Panosol (a French company) makes modules for solar panels. They receive photovoltaic cells, rearrange and re-sell them to, among others, Eurener (a Spanish company). The companies were linked through a sales contract, renewed every year. They requested the intervention of Demeter Partners in their negotiations due to communication problems, but the alliance persisted. It persisted up until the time that Demeter Partners withdrew from the capital of these companies. The acquisition of Panosol by GDF Suez put an end to Panosol’s relationship with Eurener, given that relations with clients other than Eurener were preferred and that GDF Suez has a majority stake in Panosol. The Panosol–Eurener alliance, although cross-border (France–Spain), was purely of the customer–supplier type and not an alliance with the goal for one of the companies to establish itself overseas.

The Comarth–IES alliance: this alliance also consists of a customer–supplier type relationship, preceded by a phase of common development. Comarth is a Spanish manufacturer of electric cars which was looking for battery chargers. IES makes battery chargers. Demeter Partners was originally a shareholder in IES, which had, on its own initiative, surveyed different companies on the electric car market, including Comarth. The idea of the alliance thus came from the manager of IES. Demeter Partners learned about Comarth through its stake in IES and decided to support it. The fact that Demeter Partners was then present in the capital of both companies facilitated the alliance. The investment manager we questioned at the PEF said that this: “allowed us to grease the wheels so that they continued their commercial relationship. They had sold one or two test chargers, and Comarth was a really small company before we took a stake in them. So I think that the fact that we got involved in Comarth allowed them to solidify a bit as a client so that this would work”. The manager of IES confirmed this: “The fact that Demeter had taken a stake in them accelerated things. Demeter Partners learned about Comarth thanks to IES. But it wasn’t the contact itself that Demeter Partners provided, but rather the necessary support for the relationship to develop. The fact that Demeter was a common shareholder facilitated communication a lot”. The IES–Comarth alliance, though also cross-border (France–Spain), is, again, purely of the customer–supplier type, and not an alliance with the goal of one of the companies to establish itself overseas. The relationship continues today.

3.2.2.2.3. Industries et Finances Partenaires

General context: the presentation of the field (the PEF)

I&FP¹⁸ was founded in 1999 by Emmanuel Harlé and Pierre Mestchersky. The two founders each brought two decades of experience in the field of industry and finance. They quickly hired Franck Boulland, experienced in the field of sectoral consolidation (“build-up”). I&FP is a Parisian PEF, established on the private equity market. Legally, I&FP takes the form of a mutual fund. The I&FP management company manages two funds with a total sum of €200 million in the form of a VCMF. A third fund is in the making. As a generalist PEF specializing in the operations of sectoral consolidation, I&FP takes majority stakes in the companies it supports over a territory of national scope. Today, I&FP boasts a team comprising six persons.

Sectoral consolidation (“build-up”) consists, for I&FP, of creating a holding company in which it has a majority share. This holding company then acquires, usually, four to six companies from the same sector in order to form an autonomous, more competitive group, which can – after I&FP exits – join an industrial group and constitute a division or business unit within it. Via the holding company in which I&FP holds a majority stake, I&FP also holds a majority in the companies acquired by this holding company. The portfolio companies of I&FP are thus the build-ups, or the holding companies, which themselves acquire companies. The latter are for the most part non-listed family businesses put up for resale, or subsidiaries of groups that wish to relinquish them because they are no longer part of their core business or strategic priorities. These companies are generally established in their market niches and have already proved themselves. However, they still possess strong potential for development.

Since its creation, I&FP has carried out around fifteen build-ups, and thirty complementary investments. These have taken place in every kind of sector, such as, for example, call centers, non-alcoholic beverages, e-commerce, household appliances, financial services, cosmetics, the manufacture and distribution of consumable products for the construction industry, services to communities, industrial joinery, archiving, spas and fitness, etc.

Specific context: activity of forming alliances

I&FP is specialized in sectoral consolidation. To present its activity of forming alliances for the companies which it supports, it is necessary to clarify what we mean by “alliances” in the context of build-ups.

I&FP intervenes above all by detecting exploitable synergies between the SMEs consolidated under the same holding company. This therefore involves exchanges of organizational practices between companies which are not formalized. In the context of alliances, companies combine their resources and know-how to attain goals which would have been out of their reach if they had gone it alone. Apart from this cooperation, they remain autonomous. Among the most cited goals we find, in particular: access to complementary resources, creation of synergies, realizing economies of scale or scope (for fields such as R&D), transfer or learning of knowledge, risk-sharing, the conquest of new markets (geographic or sectoral), and achieving critical size. An alliance can be formalized or not. Although a group has been formed (companies regrouped under a holding company) in the case of a sectoral consolidation, the firms are still distinct entities which remain autonomous at the operational level. It is therefore possible for us to include the formation of synergies between companies in a build-up in our study. In the context of sectoral consolidation, what particularly interests us therefore is not especially the acquisitions of companies carried out through a leveraged buy-out, but rather the exchanges that exist between the companies in a group after its formation and the synergies which they may realize. These synergies may be found at the level of costs. Companies combine, for example, the purchase of certain products to increase their bargaining power in the face of suppliers. Other synergies are possible, for example, at the marketing level or at the level of commercial activity.

I&FP might also put into contact, through an alliance, companies which belong to different sectoral consolidations, or companies inside or outside their portfolio. In general, ideas for external partners come to I&FP through its research carried out in the aim of finding companies as purchasing targets for its build-ups. Thus, the associate director whom we questioned says: “For example, I’m involved with a factory which makes melted cheese for pre-packaged foods. In the context of my research for buying other companies in this field – the field of intermediary food products – well, I stumbled upon companies which really had know-how, but whose bosses didn’t want to sell. In that case, it’s a great pity to say: ‘well, okay, we can’t do anything on the level of capital’, but it’s also a great pity to say: ‘This company which could offer us a commercial relationship to develop, etc.’. So, let’s talk to the managers anyway and get these managers talking to each other”. We must nevertheless mention that the activity of building links between companies through alliances remains very marginal outside of sectoral consolidation.

The presentation of cases (the alliances)

I&FP has enabled us to analyze a typical sectoral consolidation from the perspective of alliance formation. This involves the case of Caterine Restauration¹⁹. The Caterine Restauration group was formed between 2006 and 2010, as I&FP successively brought together five companies in the field of mass catering. These companies were Océane de Restauration, Sogirest, Culinaire des Pays de l’Adour, Centrale de Restauration Martel and Occitanie Restauration. The group was later acquired by the Compass group, a global leader in the sector, in May 2010.

Although in the French mass catering sector, on-site management represents the majority of activity compared to the preparation of dishes away from the places of consumption, the particularity of Caterine Restauration lies in specializing in the latter activity. At its formation, Caterine Restauration was therefore positioned in a diametrically opposite way to the dominant trend on the mass-catering market.

The consolidation of companies under Caterine Restauration was accomplished successively. The group adopted a flat management structure. This form of organization allowed it to leave a wide freedom of management to each entity, while putting into action a strategy of sharing “best practices”, and the creation by I&FP of a number of common values, founded on sharing certain support functions and coordinating commercial and communications activities. Reporting is also carried out collectively. The associate director whom we questioned said: “Our experience also tells us that to ensure that all this works well, we must preserve operational and commercial autonomy, we must preserve brand autonomy and respect the identity and culinary specificity of each business and, if possible, we must maintain most of the good practices already in place”.

To put its build-up strategy into action, I&FP hired a director-general for the holding company, the facilitator of the group. The idea of offering this role to one of the managers of the acquired companies was quickly dismissed as, from experience, “in the end, bosses do not particularly like managing their colleagues”, said the associate director whom we questioned. Thus, the director of the holding company does not come from the trade but constitutes the medium through which to share good practices which are identified together with the managers of the companies in the Caterine Restauration holding company during their collegial discussions. Next, one of the five companies plays a particular role: that of the reference company for the four others. This is Océane de Restauration. It was the group’s first acquisition and is also the largest structure among the five companies forming the build-up. Its manager’s role consists of sharing their experiences, their good practices and to make them available to the other companies, while respecting their local specificities and, in turn, profiting from the good practices of others. Its manager describes their particular role as follows: “To give you an example, at the level of Human Resource management, we were the only company to have a Human Resources director. The others didn’t have one. And it was Océane de Restauration’s Human Resources manager who was then hired to be the Human Resources director for the Caterine Restauration group. Similarly, when purchasing was centralized, it was Océane de Restauration’s director of purchasing who became Caterine Restauration group’s director of purchasing”. We can thus determine that the director-general of the holding company plays the role of management, while the manager of the reference company plays a role at the operational level of the business.

Sharing experiences, identifying good practices and duplicating them in the other companies, while maintaining their autonomy of implementation, resulted in synergies at several levels. In particular, this involved all the functions that were not part of the core business of the companies; so everything under the non-food heading, in the field of insurance, paper products, trays. This latter involved above all the realization of economies of scale by pooling purchases, which in practice meant, for example, the choice of common suppliers, which would thus allow the companies to band together in the face of suppliers. Next, resources were pooled between the companies at the legislative level, at the level of personnel and human resources in general (for example, classification of employees, pooling of salaries). As previously indicated, good practices at these three levels (purchases, legislation, human resources) were taken from the Océane de Restauration company and duplicated in the four others.

Apart from the synergies that the companies under Caterine Restauration were able to realize between them, links were built (involving the Océane de Restauration and Occitanie de Restauration companies) with a business belonging to another build-up (Européenne Food). The associate director whom we talked to said: “Yes, we actually tried to put in place exchanges between the world of packaging and Caterine Restauration, which were not very conclusive simply because it was all based on one thing: we wanted Caterine Restauration to be able to offer recyclable packaging”. A similar attempt at building links was made between another company under Caterine Restauration (Occitanie de Restauration) and Européenne Food. Thus, the person we talked to said: “Then, I identified that Occitanie de Restauration needed to have a reliable, regular supplier for all things involving drinks. Since Occitanie de Restauration was in Castres and Européenne Food had a warehouse in Toulouse, it made sense that the two companies should talk to try to put in place a partnership, that is, a classic customer–supplier partnership, and so that was put in place […]. It’s really foolish for Caterine Restauration to buy recyclable trays other than in the world of packaging, which is capable of providing a solution. By contrast, afterwards, if it’s not justified at the economic level, they are free to contract elsewhere […]. I have other examples like that in my head”.

3.2.2.2.4. Anonymous PEF

General context: the presentation of the field (the PEF)

Anonymous PEF is a regional and sectoral PEF with the legal status of a venture capital firm, which positions itself as a specialist in industrial venture capital. Headquartered at Metz in the Lorraine region, it takes minority stakes in the companies it supports. It is still a young PEF, created in 2009, in the process of becoming established on the French private equity market. The status of SCR a venture capital firm has the specificity, from the point of view of our problematic, that Anonymous PEF has the status of a joint-stock company (JSC). This signifies, in theory, that the shareholders of Anonymous PEF may have an influence on the investment decisions and the strategies it pursues.

Anonymous PEF is headed by a single investment manager, in charge of the strategic deployment of investments. With €18 million in capital, Anonymous PEF invests in innovative industrial projects in the field of materials. The PEF is administered by a board of directors composed of representatives of the shareholders, as well as independent personalities from the field of finance and industry. An investment committee analyzes the investment proposals drawn up by the management team including, in particular, the investment manager. Based on the recommendations of this committee, it is up to the board of directors to approve the investment strategies. This latter is managed and chaired by the president of the Materialia competitiveness cluster and by the Lorraine region’s economic and social advisor, Jean-Louis Pierquin.

Today, Anonymous PEF supports eight portfolio companies. Two of them were still prospects at the beginning of our analysis near the end of 2010. These companies are mostly situated in Moselle or allow cooperation with companies from the region. In addition, their activity is linked with the field of materials.

Specific context: activity of forming alliances

Concerning the activity of forming alliances, Anonymous PEF systematically investigates the possibility of realizing synergies, whether between its portfolio companies or between a prospect and a company already in the portfolio (so-called “horizontal” alliances), or between at least one portfolio company (or a prospect) and one of Anonymous PEF’s shareholders (so-called “vertical” alliances). In fact, its status as a joint-stock company enables it to not only actively form alliances between its portfolio companies but, also, between portfolio companies and shareholders. The formed alliances, or those in the process of formation, are mainly intraportfolio alliances, generally between two portfolio companies, often on the initiative of Anonymous PEF.

Thus, a first example of an alliance could be that existing between two companies in the investment portfolio. The first is a start-up which manufactures ecological composite plastic resins, which are 30% composed of plastic resins and 70% of natural resins. These two manufacturers are specialized, in particular, in everything concerning the manufacture of profiled slats. The second makes stretch ceilings and uses a lot of slats as support for these ceilings. The company is thus interested in acquiring slats to be able to offer more ecological products.

A second example is represented by the alliance between the first company and a third. This latter manufactures solar panels with an aluminum frame structure. The first company’s product offers the possibility of replacing the aluminum with plastic and a natural component. For the third company, this allows it to offer solar panels which are cheaper and more environmentally friendly with an aspect of sustainable development, and thus to offer clients a different product allowing it to differentiate itself on the market.

A third arrangement to form an alliance is under discussion between two other companies. The first designs innovative implants for patients suffering from disorders of the larynx, while the second specializes in the design and manufacture of surgical materials and implants including surgical screws. The first company manufactures titanium implants for surgery. The idea is to put the two companies into contact, so that they can share their strategic vision and distribution channels.

Finally, a fourth example illustrates the formation of an alliance between a portfolio company of Anonymous PEF and a portfolio company of a related fund. One of the two companies forming the alliance manufactures roof panels. The idea of the alliance is that this company could make roof panels within which are integrated the solar panels that the other company provides. This would allow the two companies to provide a market offering with greater added value. Thus, the second company can not only sell solar panels, but also a roofing solution; and the first company can sell a roofing solution which is also an energy-producing solution. Finally, there is a fifth example similar to the preceding, between two other portfolio companies.

Under the heading of vertical alliances, two companies have been put into contact by the investment manager with one of the shareholders of Anonymous PEF. This shareholder is greatly involved in the construction of buildings, which might be interesting, first, for one of Anonymous PEF’s portfolio companies which is seeking roofing solutions. The portfolio company could either sell solar panels to the Anonymous PEF shareholder, or buy roofing solutions for them. It is thus a customer–supplier relationship. Second, the shareholder might also be interesting for another portfolio company. This latter was put into contact with the R&D section of the Anonymous PEF shareholder. The portfolio company is specialized in the coating of metal surfaces. It could therefore, in the long term, offer solutions for surface coating, varnishing for the shareholder’s products, covering layers. Nevertheless, this involves longer term projects whose outcome in terms of alliances is uncertain.

The previously described alliances were often put in place at Anonymous PEF’s initiative. The formation of alliances is encouraged by the presence of the Lorraine regional authority in Anonymous PEF’s capital. It promotes synergies between companies so that there might be benefits for the region. For vertical alliances, the role of Anonymous PEF is generally limited to creating the relationship, while for horizontal alliances, Anonymous PEF’s intervention continues. All through the process which leads to the formation of the alliance and, occasionally, to the signing of a contract, the investment manager of Anonymous PEF is present and will facilitate exchanges, follow the (future) partners of the alliance and try to ensure that it comes to fruition, if there is an economic logic. This may mean either simply being present and intervening to establish trust, playing the role of guarantor, or intervening at the level of communication or exchange.

Extraportfolio alliances putting an Anonymous PEF portfolio company in contact with an external company to form relationship are also possible but, currently, rarer. The ideas for such alliances come from collegial discussions between the manager of the supported company and the investment manager of Anonymous PEF and may be at the initiative as much of the PEF as of the supported company.

The presentation of cases (the alliances)

The two cases of alliances offered to us by Anonymous PEF are the first two cases of alliances that were briefly presented previously.

The three portfolio companies involved each manufacture their own products. The first designs a material which can be integrated into many products, in particular, those of the two other companies. Anonymous PEF’s investment manager quickly realized this and spoke to the managers of the other companies involved. Being interested, the managers were brought together for a lunch to discuss it, to see if it is technically feasible and conceivable to use the first company’s materials to improve the existing products of the two others.

The first interfirm relationship is interesting as it may offer, eventually, a more environmentally friendly product on the market. One of the companies has a product based on stretched canvasses. These canvasses are held by rails. The product is interesting for its environmental performance. Given that the canvas is stretched, for example, there is no gluing or paper in comparison to PVC canvasses. Thus, besides the technical aspects, there are advantages on the environmental level. In addition, today, sustainable development represents a strategic aspect on the marketing level. The second company’s environmentally friendly profiled slats can thus be used to support the stretch ceilings of the first, replacing the PVC-based solution that they use and thus providing a more ecological component. However, in order to achieve this, it is necessary for the second company to adapt their product so that is can be used by the first. Finally, if this succeeds, this will give rise to a customer–supplier relationship where the second company would deliver materials to the first for their products. Currently, the companies find themselves in a common development phase in order to adapt the supplier company’s product to the needs of the customer–company.

The second alliance is similar to the first. Anonymous PEF’s investment manager detected an opportunity to combine the two companies and spoke to their managers. They played the role of intermediary and organized a meet-and-greet on the site of one of the companies forming the alliance. The managers discussed the usefulness of the putative relationship and its feasibility. Again, this involved a longer term project which, at best, might result in a customer–supplier relationship where one of the two companies would deliver materials to the other. Again, it was necessary for the company furnishing the materials to launch a development phase in order to offer materials compatible with the products of the other enterprise. The exchanges continue to this day.

In both cases, the exchanges were not formalized as a contract. The alliances may result in the drawing up of a contract if, eventually, the supplier company succeeds in developing materials which are directly usable by the two customer companies.

3.2.2.3. Sources of data

This section describes the sources that we had access to in the context of our case study. We begin by explaining the general approach to data collection (section 3.2.2.3.1), continuing with the sources that we had access to according to the field of analysis (section 3.2.2.3.2).

3.2.2.3.1. General data collection approach

Our data are mostly qualitative and were gathered from structured interviews for the most part. They were complemented by quantitative data collected from two questionnaires. The questionnaires were addressed, on the one hand, to PEFs and, on the other hand, to SMEs supported by private equity in the framework of the statistical and econometric study. Some of them were addressed to and filled out by the persons involved in the case studies. This allowed us to triangulate the data obtained from these persons through two different methods, questionnaires and interviews.

Our data sources are thus primary data sources. Access to sources of secondary data such as, for example, written documents (reports, press articles, contracts for alliances) was not available to us. In any case, databases rarely contain the type of information that was needed. This is mainly due to two factors. The first is that we were studying the formation of alliances between mostly non-listed companies, who publish relatively little information and even less information of a strategic nature. The second factor is that the great majority of alliances between firms supported by French PEFs are not formalized.

In what follows, we describe our primary source for the case studies: the interviews. The questionnaires were presented in the section on the statistical and econometric study. The great majority of the interviews we conducted were structured in nature. Some semi-structured interviews were carried out at the beginning of the research, in order to better identify the subject and to see, together with our interlocutors on the PEF side, if they would be able to serve as an adequate field for our study in its empirical phase of testing hypotheses. Thus, we undertook two waves of interviews. The first wave, smaller than the second, was preparatory in design and included semi-directive questions. The second, main wave, was explanatory in design and structured in nature. Overall, about fifteen people were involved in the multiple case study.

First wave of interviews

The first phase aimed to better identify the subject and to let the people we interviewed express themselves freely, in their own words, concerning the research question “the role of PEFs in strategic alliances”, and according to their own experiences. The interview guide was used by the researcher only to restart discussion after the interviewee finished their comments. For the majority of cases, these introductory interviews were conducted in parallel to the emergence of the theoretical framework, in order to have the best conditions to prepare the phase of empirical testing of the hypotheses resulting from our mobilized theoretical frameworks. The interviews varied in length – in general, from thirty minutes to one hour and thirty minutes. In one particular case, the interview lasted more than two hours.

Second wave of interviews

The second wave of interviews aimed to test the theoretical framework. It constituted the main stage of interviewing. Discussion with our interlocutors was oriented in such a way that the respondent expressed themselves in relation to the research hypothesis, while still having the opportunity to speak freely; that is, without us involuntarily influencing their responses in such a way that they would necessarily support the hypotheses. Thus, the questions contained in the interview guide were formulated in order to try to lead the respondents to take a position with relation to the hypotheses, and thus to mitigate the risk that they would not understand the question (if it was posed in too academic a fashion) or did not wish to respond clearly (if this embarrassed them). After the formulation of the different hypotheses, some of them were rephrased directly in the form of a question. For other hypotheses, more subtle formulations were put forward. In the interview, discussion remained free to the extent that the questions were phrased and adjusted by the researcher based on the responses and the information delivered by the respondent. The length of these interviews varied between one to two hours.

The interview guides were previously submitted to the judgment of researchers, so they might be able to give advice, as suggested by, for example Yin [YIN 03, p. 62].

The unit of analysis of our study consists of an alliance formed between at least two companies in the presence of a PEF. This means that a case consists of a formed alliance (or one in the process of being formed) between at least two companies. The analytical perspectives we took into account were the points of view of the managers of the SMEs forming the alliance, as well as the point of view of at least one investment manager working for the PEF involved, who is in charge of supporting the companies forming the alliance. Depending on our fields of analysis, the number of interviews conducted might slightly vary according to the availability of people. This is explained in the following section.

3.2.2.3.2. Data sources according to fields of analysis

We will now describe the collection of data from each field of analysis (PEF) as well as from auxiliary sources, which we were able to access.

Siparex Group

We first made contact with Siparex Group by sending an email to different people at the PEF, including the executives and the investment managers. A positive response was obtained after several attempts. We made contact with one of the deputy directors of the group. There then followed an initial telephone discussion of exploratory nature and, then, a discussion of explanatory design. This same person was responsible for putting us in contact with portfolio companies of the group which had formed alliances. We thus were able to make contact with the managers of the companies Sophem, Meseo and Mooviin, who participated in the discussion group on e-commerce which was initiated by Siparex Group. These three company managers also granted us a telephone discussion in order to test our theoretical framework. Our questionnaires were then sent to all the interviewees. We were able to obtain complementary information via Siparex Group’s website as well as its supported companies.

Demeter Partners

Contact was made with Demeter Partners in a similar way to Siparex Group, by sending e-mails to different people working for the PEF. After follow-up and a telephone call, one of the secretaries at the PEF directed us to one of the investment managers, who agreed to help us in our research project. We carried out the two waves of interviews. The investment manager also agreed to put us in touch with the managers of portfolio companies of Demeter Partners who had formed the two alliances that the PEF suggested we study. Three of these managers accepted. These were the managers of the companies Panosol (for the Panosol–Eurener alliance) and the managers of IES and Comarth for the alliance that these latter two companies had formed. All three agreed to conduct a telephone interview with us, which began with a phase of free expression and thereafter continued with an explanatory design. Finally, we sent them our questionnaires. We were able to obtain complementary information on the Internet sites of the companies and of the PEF.

Industries et Finances Partenaires

Contact with I&FP was made through the intermediary of a person from CDC-Enterprises who, knowing about our research topic, put us in touch with the PEF. Specialists in sectoral consolidation, I&FP allowed us to study an exemplary case which it had been able to carry out: the Caterine Restauration build-up, and the synergies which the firms taking part had been able to achieve. Two interviews were carried out at I&FP’s offices with the associate director in charge of the Caterine Restauration build-up. The first, of exploratory nature, took place in the presence of a second interlocutor: the director-general of the Caterine Restauration holding company, who was present at the formation of the build-up under I&FP. Their role was to act as an interface between the managers of the SMEs who took part in the consolidation and the PEF I&FP. Their principle role was to detect and achieve possible synergies between the companies in the consolidation. The second interview was of explanatory design and aimed to test the theoretical framework.

The associate director of I&FP who agreed to work with us in our study then put us in touch with the manager of the “mother” company of the Caterine Restauration Group, Océane de Restauration. They allowed us a direct interview, in Paris. This was conducted starting with a brief exploratory phase, followed by the testing of our theoretical framework. Next, we were put in touch with the manager of the company Centrale de Restaurataion Martel, a subsidiary of Caterine Restauration, which granted us a telephone discussion that took place in the same way as for the manager of Océane de Restauration. Three other companies make up the Caterine Restauration consolidation. However, it was not possible for us to talk to their managers, given that the who were managers in addition to present at the time of the consolidation are no longer there. Finally, the points of view of the PEF, the SMEs forming the alliance and that of the director-general of the Caterine Restauration holding company, we were able to obtain, exceptionally, a fourth point of view. This was that of the financial director of Compass, which bought Caterine Restauration after I&FP’s exit. During a telephone discussion, he was able to explain to us, according to his point of view, the added value of the alliance formed and the role that I&FP played in it. Finally, our questionnaires were sent to the PEF and the managers of the SMEs who were interviewed. We were able to obtain complementary information from the Internet sites of I&FP and the Caterine Restauration group.

Anonymous PEF

After sending an e-mail, followed by a telephone conversation, the investment manager of Anonymous PEF offered us a first meeting at Metz. We carried out an interview of exploratory nature. At that stage, one of the companies whose manager had agreed to participate in the study was still a prospect. It had become a portfolio company at the time of the interview of explanatory design. A second interview – this time of explanatory design – was thus subsequently carried out with the investment manager of Anonymous PEF, again at Metz. The latter also put us in contact with the managers of the portfolio companies who had formed alliances which he suggested we study. A first contact with the managers of all the portfolio companies of Anonymous PEF at the time took place at the first anniversary of Anonymous PEF, to which they were all invited, and we were also. Subsequently, the investment manager of the PEF followed up on the contact with the company managers who had formed alliances which he had suggested we study. The managers replied, and the interviews of explanatory design were conducted over the telephone.

The questionnaires were sent to the persons interviewed. Finally, we were able to obtain complementary information from Anonymous PEF’s websites, which had been set up during our study, and those of its portfolio companies.

3.2.2.4. Analysis procedure

The analysis is constructed in such a way as to take account of and address the fact that we are proceeding to a multiple case study which is embedded. To do this, we worked on two levels. The first level consisted of an intrafield analysis. This is embedded in the second, which consists of an interfield analysis. Practically, the first level consists of summarizing by hypothesis (which we regroup into themes) the different points of view of the PEFs and the SMEs forming the alliance and contrasting them. This produces for each field of analysis – that is, each PEF – a table comparing the point of view of the PEF with that of the surveyed SMEs, to see if they overlap. Once that is done for each of the four fields of analysis by hypothesis, we pass onto the second level. This latter no longer differentiates between the points of view of the PEF and the SMEs forming the alliances. It takes up the intrafield summaries and contrasts their results to those obtained for the other fields. The hypothesis to be tested is confirmed or rejected if all the fields of analysis (including the first levels of analysis which compare field by field the points of view of the SMEs forming the alliances and the points of view of the PEFs) result in a similar response. The two following points focus on the detail of the analysis tables for the two levels.

The summaries of points of view and their grouping into tables were produced in an analytical way on the basis of the interviews. We made no use of coding or lexical analysis, as these methods did not seem relevant in the specific framework of our research. In a case where the researcher had access to different sources of information (beyond the interviews, for example to reports and/or press articles), a lexical study of the different documents might be useful to highlight the themes addressed in these texts in relation to the research variables. However, in the precise context of our research, we only possessed the information coming from the interviews we conducted. Given the structured character of the discussions, the respondents were directly led to reply to questions linked to the research hypotheses. Superfluous information was accordingly minor or could be eliminated directly by the researcher, without needing to use coding software. Another tool of this type of analysis consists, in particular, of counting the number of times certain words or themes appear, in order to highlight these items to grasp a theoretical concept which is difficult to evaluate. In the context of our research, only the concept of reputation is involved, and it was associated with a precise question in the interview. It was therefore sufficient for us to refer to the response and summarize it.

3.2.2.4.1. First level of analysis

The first level of analysis gave rise to Table 3.6 which follows. There are thus four tables for each hypothesis, corresponding to the four fields. For each hypothesis, we begin with the Siparex Group field of analysis, then move onto Demeter Partners, then to I&FP, ending with Anonymous PEF.

Table 3.6. Example of an intrafield analysis table

Variable	PEF	Alliance 1	Alliance 2
	PEF name	SME 1 name	SME 2 name	SME 3 name
General opinion
Intraportfolio alliances
Extraportfolio alliances

Table 3.6 presents the case of the most complete table possible. If we take as an example Hypothesis 1: “The reputational capital of PEFs strengthens the mechanism of trust and, consequently, has a positive impact on the formation of alliances”. The independent variable to analyze is the reputational capital of the PEF, and we study its impact on the dependent variable “the formation of alliances”. We therefore begin with the analysis of the first field, Siparex Group. To do this, we fill in the table with the points of view of the PEF and the SMEs forming the two alliances, from the interviews conducted. In most cases, one of the SMEs participating in the study is involved in both examples of alliances. This is why there are often three company managers being quoted and not four (an alliance had been defined, for the empirical section, as a relationship between two companies, even if in principle, an alliance is a relationship between at least two companies). We begin by identifying the general opinion of the PEF and the SMEs as to the reputation of the PEF. Next, we contrast their points of view on the link between the reputation of the PEF and the formation of alliances by its portfolio companies. A distinction between the two types of alliance – intra- and extraportfolio – is made if necessary. In the case where points of view diverge, we analyze the reason for this. Is it due to the context, or is there a real divergence? This is done successively for the four fields.

3.2.2.4.2. Second level of analysis

At the second level of analysis, the four intrafield tables of level 1 are in turn summarized into a final table (see the example of Table 3.7), which opposes the summary of the overall point of view for each field to that of the others. Thus, it is possible to see if all the fields confirm the hypothesis or not and, above all, confirm or refute the mechanism of causality that the theory presumed.

Table 3.7. Example of an interfield analysis table

Variable	Siparex Group	Demeter Partners	Industries et Finances Partenaires	Anonymous PEF
Supports the hypothesis	Yes/no	Yes/no	Yes/no	Yes/no
Mechanism of causality

Table 3.7 takes up the variable, which is to be tested. In the first line, we no longer see the PEF and the SMEs forming the alliance in order to compare their points of view, but rather the four fields with, on the second line, the global summary of the first level of analysis, to know if the field in question supports the hypothesis or not. On the third line, finally, we see the summary of the study of the presumed mechanism of causality by the different persons interviewed at level 1. Finally, the hypothesis is confirmed or rejected if all the fields lead to the same conclusion. If this is not the case, a discussion follows to understand the reason. Is the absence of the convergence of different points of view due to the specific context of the field of analysis? Or is the hypothesis invalid in certain specific contexts? Finally, even if the independent variable being tested is indeed shown as able to explain the studied phenomenon, it remains to see if the presumed mechanism of causality is confirmed or not.

3.2.2.5. Case analysis

The case analysis is carried out with reference to the theories presented in our theoretical section, the goal being to test our research hypotheses. The analysis is divided into two main parts:

• – testing the contractual hypotheses along with the complementary hypotheses arising from the sociological network theories related to them;
• – testing the knowledge-based hypotheses and the complementary hypotheses arising from the sociological network theories.

3.2.2.5.1. Contractual hypotheses

We start the analysis of the contractual hypotheses from the point of view of the SMEs forming the alliance. We continue with those of the PEF. The approach of the analysis follows the two-level plan presented in Section 3.2.2.4.

The SMEs’ point of view

As we saw in the theoretical section, the literature whose argument is based on contractual theories presents two main levers allowing the reduction of losses of value linked to the presence of transaction and agency costs in a situation of informational asymmetry and uncertainty. This involves trust, an informal mechanism of governance, allowing the self-execution of contracts, reinforced by the mechanism of reputation and the disciplinary lever. We begin with the hypotheses relating to the first lever, from the point of view of the SMEs, and carry onto the second. A distinction between intra- and extraportfolio alliances is made where necessary.

The mechanism of reputation

Concerning the mechanism of trust from the point of view of the SMEs forming the alliance, we offer the following hypothesis:

HYPOTHESIS 1.– The reputational capital of PEFs reinforces the mechanism of trust and, consequently, has a positive impact on the formation of alliances.

The lack of visibility of companies supported by private equity, due to their non-publicly listed status and limited track record, generally leads to strong informational asymmetry. Potential alliance partners may therefore feel distrust as to the quality of the company and refuse to enter into cooperation. On the one hand, PEFs may contribute to the establishment of trust between prospective partners by reducing informational asymmetry. Given that they finance the companies in question, they have usually minutely analyzed the files before choosing a candidate for financing and have already collected information for their own decisions. Moreover, their presence on the board of directors also allows them direct access to information of a strategic nature. On the other hand, when the quality of a young SME cannot be directly observed, external actors base their decisions on the quality of the actors working with the latter, in order to evaluate that of the young SME. PEFs may play a role in certification of the quality and the financial stability of the companies they support, which reduces the mistrust of external actors. This role should be all the more important when the PEF has a reputation on the markets. We must now verify the plausibility of this mechanism across our different cases.

• A) First level of analysis
  • a) Siparex Group

The points of view of the PEF and the SMEs have been analyzed based on the interview transcripts and collected in Table 3.8. We begin by identifying, based on the interviews carried out, the general opinion of the PEF and the SMEs concerning the reputation of Siparex Group. Next, we contrast their points of view on the link between the reputation of Siparex Group and the formation of alliances by its portfolio companies. In general, they overlap. The responses are given in detail just below the table.

Table 3.8. The Siparex Group field in relation to Hypothesis 1

Reputation	PEF	Alliance 1	Alliance 2
	Siparex Group	Sophem	Mooviin	Meseo
General opinion	Dependence on history. Siparex is one of the first French PEFs. Good image and good professionalism conveyed through Club Siparex.	Regional PEF/good reputation/visibility	Regional PEF known for its professionalism	Regional PEF/strong reputation, in particular in Lyon
Intraportfolio alliances	Sign of trust/careful analysis of the files/allows assurance of the quality of the SME and its project.	Reputation does not matter, it is trust, the actors know each other.	Sign of seriousness/gaining visibility/strengthens credibility.	Label of quality/role of a guarantee assuring the quality of the SME and its financial stability/ensures an image of seriousness.
Extraportfolio alliances	Sign of trust/careful analysis of the files/allows assurance of the quality of the SME and its project.	No extraportfolio alliances yet, but positive role as builds credibility at the level of the company’s financial stability and quality.	Sign of seriousness/gaining visibility/strengthens credibility because a PEF performs analysis.	Reassures and facilitates the meeting of companies.

1. 1) General perception of reputation
   1. I) The PEF’s point of view

From the point of view of the PEF, Siparex Group, reputation is dependent on history. The French private equity market is quite new and more or less corresponds to the birth of Siparex Group, 36 years ago. The PEF therefore possesses wide visibility based on its history and is known for its professionalism. Siparex Group adds: “It’s a little like if you asked what the Bank of France’s image is”. Another of its assets is the factual aspect including, in particular, the number of operations performed, sustainability over time, and the satisfaction of supported entrepreneurs.

• II) The SMEs’ point of view

This perception of Siparex Group’s reputation is reinforced by the managers of the portfolio companies interviewed. According to the latter, Siparex Group is a PEF with a good reputation, a good image and known for its professionalism and its regional orientation, in particular, on the region around Lyon where it was founded.

• 2) Impact of the PEF’s reputation on the formation of alliances
  • I) The PEF’s point of view

According to the PEF, there is a positive correlation between its reputation and its portfolio companies forming alliances. Siparex Group capitalizes on the group’s good image and professionalism through its network and Club Siparex, which allow the group’s portfolio companies to form alliances. For a prospective alliance partner, the presence of a PEF such as Siparex Group is a sign of trust, since the company knows that a PEF, before investing in an SME, carries out diagnostics on its financial aspects, on the level of development and on the company’s strategy. Trust is built over time, but the presence of a PEF can strengthen this trust. This is all the more important when the company being supported is young, as trust is then more difficult for a third party to gain, given that the company does not have a lot of background or track record. This allows a sort of testimonial to the company’s quality, since according to the Investment Manager we interviewed: “… as a rule, out of a hundred files which are sent to us, we are only going to study about twenty and only finance about a dozen. That means that for the companies in which we’ve taken capital, there is a relatively important filter”. As for financial instability, this is inherent to SMEs. The presence of a PEF may offer some reassurance of financial stability. This is worth as much for partners in intraportfolio as in extraportfolio alliances.

• II) The SMEs’ point of view

From the point of view of the SMEs, there is also a positive correlation between the reputation of Siparex Group and the formation of alliances. The three company managers we interviewed were able to form an intraportfolio alliance thanks to the PEF. There were also proposals for extraportfolio alliances. For the SME managers, the presence of Siparex Group and its reputation is above all a sign of seriousness for prospective partners. The PEF plays the role of a guarantee, of a quality label. Its presence allows some third parties to be reassured, in particular prospective alliance partners, by giving an image of quality, of seriousness, on the one hand, and reassurance as to the company’s level of financial stability on the other hand. For the manager of Meseo, the presence of Siparex Group and its reputation clearly facilitates the formation of alliances, intra- as much as extraportfolio. Without Siparex Group, building links with other companies within alliances would have been much more difficult. For the manager of the company Sophem, the presence of Siparex Group is particularly important in the case of forming intraportfolio alliances. Belonging to the same PEF allows the building of trust between prospective alliance partners. For intraportfolio as much as extraportfolio alliances, Siparex Group’s reputation lends credibility to the quality of the company supported by private equity, whether at the level of its financial stability or its business model.

Interviewees added, however, that the effect of Siparex Group’s presence and its reputation on a third party depends on the latter. Thus, certain suppliers or customers might equally associate a negative image with the presence of a PEF in a company’s capital. This also goes for external partners, and thus in the case of formation of extraportfolio alliances. For the alliances formed, though, the role was positive.

In summary, all the people interviewed testified to a positive correlation between the reputation of Siparex Group and the formation of alliances for its portfolio companies, whether intra- or extraportfolio alliances. The presence of the PEF is above all a sign of trust, of the seriousness of the company being supported, and allows prospective partners to be reassured as to the merits of the company or its projects, and also gives reassurance at the level of its financial stability.

• b) Demeter Partners

The approach for this field of analysis is the same as for the previous one. The points of view of the PEF and those of the SMEs have been analyzed on the basis of interview transcripts and collected in Table 3.9. We begin by identifying from the interviews conducted the general opinion of the PEF and of the SMEs concerning the reputation of Demeter Partners. Next, we contrast their viewpoints on the link between Demeter Partners’ reputation and the formation of alliances by its portfolio companies. Generally, they overlap. The responses are given in detail just below the table.

Table 3.9. The Demeter Partners field in relation to Hypothesis 1

Reputation	PEF	Alliance 1		Alliance 2
	Demeter Partners	Comarth	IES	Panosol	Eurener
Generl opinion	Reputation of a PEF (assures the SMEs’ financial stability); good industrial reputation and a sectoral PEF, well known on the markets (in particular in its sector of activity).	Very good industrial image; clear impact on the formation of alliances.	Sectoral PEF with an engineering culture. Good reputation; clear impact on the formation of alliances.	Demeter has a very good reputation as it is a sectoral PEF. Its reputation plays a large role in the formation of alliances as Demeter is well known in its sector of specialization and has a large network there.
Intraportfolio alliances	Intermediary role in establishing trust between SMEs because of fear of going bankrupt (compared to a large group) can bail out the SME from its own funds in case of difficulties. Particularly important for international alliances; reputation of supporting good projects (so confidence in the merits of the alliance project).	Only one relationship built with IES (limited potential for building relationships). Injection of trust which facilitates exchanges; reassurance as to level of financial stability.	One relationship built with Comarth.	One internal alliance with Eurener thanks to Demeter Entrepreneurs Club. Good reputation of Demeter gives assurance that alliance partner is serious.
Extraportfolio alliances	Like for intraportfolio alliances, guarantees extraportfolio alliances because sectoral PEF, known by SMEs outside investment portfolio in its sector of specialization. Demeter Partners knows these companies and inversely this is what inspires confidence and trust.	Higher potential for building relationships than for internal. Discussion underway. The fund’s presence gives an image of strength and seriousness to Comarth that inspires confidence and trust; reassures at the level of financial stability.	Higher potential for building relationships than for internal. The presence of Demeter is a very good sign, especially when entering into relationships with large groups. Demeter gives the company credibility and is a badge of quality with suppliers. It has an impact on trust and is reassuring about the quality and financial stability of the company. Often meetings have been held at Demeter premises.	There have been proposals, but nothing has materialized. But Demeter’s reputation has an important impact, particularly, in its area of specialization.

1. 1) General perception of reputation
   1. I) The PEF’s point of view

From the point of view of the PEF, Demeter Partners has a good reputation among companies, known more as an industrial than a financial PEF, which allows it to initiate projects. Demeter Partners’ reputation is not only that of a PEF in general, but also a personal reputation among entrepreneurs. The latter are familiar with it due to its sectoral nature. Entrepreneurs hear about Demeter Partners by word of mouth from other entrepreneurs who are themselves supported by the PEF.

1. II) The SMEs’ point of view

This perception of the reputation of Demeter Partners is reinforced by the managers of portfolio companies interviewed. According to them, Demeter Partners has a very good image and reputation due to it being a sectoral PEF with a culture of both engineering and finance.

1. 2) Impact of the PEF’s reputation on the formation of alliances
   1. I) The PEF’s point of view

From the PEF’s point of view, there is a correlation between Demeter Partners’ reputation and the formation of alliances. In the case of Demeter Partners, this correlation is positive because their reputation is good. It could be negative if the PEF has a bad press. Demeter Partners’ reputation allows the establishment of a situation of trust between prospective alliance partners. In the case of forming intraportfolio alliances, the fact of having a common shareholder is reassuring, all the more so if the shareholder is a PEF, which represents a guarantee at the level of the companies’ financial stability. This aspect is important, as the alliances are often of the customer–supplier type, and the supplier wants to be sure of being paid. In addition, the presence of the PEF is reassuring at the level of the quality of the companies being supported. However, apart from the reputation of the PEF, what is most important is the personal relationship between the PEF and its portfolio companies. The person interviewed at Demeter Partners states: “I think that above all else it is the personal relationship. If the entrepreneur feels comfortable with the person at our firm who is supporting them, they tell themselves that we are trustworthy. All the same we’re dealing with SMEs, business which are a bit “old-fashioned” where people have to trust each other in spite of everything. If there’s someone who reassures you and tells you that this company is good and, even better, if they are a shareholder of it, you have more of a tendency to work with them than with anyone else”.

Apart from intraportfolio alliances which, in the case of Demeter Partners, are often alliances between companies in different countries, the most common configuration is the extraportfolio alliance, linking mainly French companies. In these cases, the PEF’s reputation has a positive impact on the formation of alliances, but more important, again, is the personal relationship, as Demeter Partners indicates: “Yes. But, what will count for more, again, is the person who has been in contact. So yes, there’s Demeter, but there’s also the personal relationship between people”. This fact is mainly linked to the context. Demeter Partners is a sectoral PEF, well known by entrepreneurs in that sector, whether they work with Demeter Partners or not.

1. II) The SMEs’ point of view

According to the managers of Demeter Partners’ portfolio companies that we interviewed, there is a positive correlation between the PEF’s reputation and the formation of alliances. Demeter Partners’ reputation allows the establishment of a situation of trust, in the cases of intra- and extraportfolio alliances equally. The presence of the PEF is reassuring at the level of the company’s financial stability as much as its quality. Apart from the reputation of Demeter Partners, it is mainly a history of people. Thus, the manager of Comarth said the following when asked whether the reputation of Demeter Partners, in itself, might play a role in the formation of alliances for their company: “The fact of having Demeter behind us as well? Yes, and well, of course, we use it. For example, last week the meeting at Brussels with the Pierre et Vacances group for an order, it was clear that having Demeter behind us was like the contract was already won. We use it, yes indeed. Today we have a contract with the Post Office to sell vehicles, the French Post Office that is in effect today. It’s true that Comarth worked for it, but there’s also the support and the assurance given by having an investment fund like Demeter behind you. So, thanks to that, we also get contracts. Of course we use it, yes. What’s more, Demeter asks us to use it as well. They tell us “don’t be shy, use it”, then they assure us that we have this economic power behind us anyway to take on the contract… I use this resource, for example, if I have a meeting with a big European company, in France, at Paris or elsewhere, I try to take these people onto firm ground, that is taking the company to Demeter’s premises so I can use it, anyway… the image that Demeter offers with premises in Paris. And they’re behind me at these meetings, they offer support during these meetings. … We use the fact of having a partner like Demeter to show that we’re strong and we’re capable of fulfilling the contract, whatever it is. Logically, groups like that assess our financial strength and we have financial strength thanks to Demeter… And I have their support, as a rule, when it’s a contract of significant scale, it’s obvious that we’ll use Demeter’s name, it’s obvious”.

To summarize, we can identify that the interviewees confirm the existence of a positive correlation between the reputation of Demeter Partners and the formation of alliances by its portfolio companies, whether intra- or extraportfolio alliances. The presence of the PEF is mainly a sign of trust, of the seriousness of the company being supported, and reassures prospective partners as to the merits of the company or of its projects, but especially and above all gives reassurance about its financial stability.

1. c) Industries et Finances Partenaires

The points of view of the PEF and that of the SMEs have again been analyzed on the basis of interview transcripts and grouped into Table 3.10. We proceed in the same way as we did for the other fields. Generally, the points of view analyzed overlap. The responses of the interviewees are detailed immediately afterwards.

Table 3.10. The Industries et Finances Partenaires field in relation to Hypothesis 1

Reputation	PEF	Caterine Restauration
	Industries et Finances Partenaires	Océane de Restauration	Centrale de Restauration Martel
General opinion	One of the rare PEFs specializing in build-ups on the small and middle-caps market. It is the perception of their investors and of commercial banks, as well as more and more of managers who want to sell their company. The latter hear about I&FP by word of mouth.	I&FP has very good reputation and visibility among buyers. On the other hand, managers who want to sell their company cannot usually make judgments on this because they do not know the different PEFs. In this case, reputation plays no role except that of PEFs in general. PEFs are known among themselves.	–
Intraportfolio alliances	Reputation is a factor in the attraction of potential targets, but not afterwards. After acquisition, when companies make alliances, the trust between I&FP and their portfolio companies is the determining factor.	It is more the presence of the manager of Océane de Restauration which played a role in the establishment of trust among the managers of other companies bought by I&FP.	Company managers are not usually familiar with PEFs. Their reputation is therefore judged on the basis of investment funds in general. In their experience, it is quite a constructive PEF.
Extraportfolio alliances	Same as for intraportfolio alliances. External alliance partners were identified through market screenings by I&FP’s collaborators, looking for acquisition targets. They are thus generally familiar with I&FP and vice versa. At the stage of building links within alliances, it is no longer reputation but trust which plays a role.	–	The reputation of I&FP is important in alliances between companies from different build-ups, but has a negative effect on commercial relations, as the presence of an investment fund is generally perceived negatively.

1. 1) General perception of reputation
   1. I) The PEF’s point of view

According to I&FP, they are perceived as one of the rare French PEFs specializing in build-ups on the small and middle cap market, which has contributed to the success of several projects. At least, this is the image of them that is held by their own investors, the commercial banks who provide them with files and, more and more, managers who wish to sell their company. Notwithstanding this, however, many company managers are not familiar with the different PEFs and can only judge the reputation of PEFs in general and not of each PEF in particular.

1. II) The SMEs’ point of view

From the point of view of the company managers interviewed about I&FP, the reputation of I&FP is that of a PEF in general, since company heads are not familiar with the different PEFs. The manager of Océane de Restauration stated: “I think that investment funds, if they have a reputation, it’s better known among private equity companies and funds. But not by their prospective portfolio companies”. This vision is shared by the manager of Centrale de Restauration Martel. For them, the reputation of the PEF, if it has an impact, is more at the level of buyers, that is the groups who, in general, buy the build-up when I&FP exits. From their point of view, the reputation of I&FP should be good.

1. 2) Impact of the PEF’s reputation on the formation of alliances
   1. I) The PEF’s point of view

According to I&FP, its reputation only plays a very indirect role in the formation of alliances. It mainly plays a role in attracting potential targets. Alliances are formed mainly within a build-up. They may also take place between companies from different build-ups under I&FP. In the case of alliances involving companies outside I&FP, this mainly involves previous potential targets. Thus, these companies already know I&FP and vice versa. There is already a small link of trust, which plays a positive role in the formation of alliances. Reputation is no longer a direct factor at this stage. Thus, I&FP responds to the question of whether its reputation plays a role in attracting external partners to alliances: “Yes, potential targets which we did not end up buying, etc., know me, know us, but this isn’t reputation, you see, it’s really because the boss is a boss, an entrepreneur, so he has in mind that there’s a path of development, so they engage, but the reputation of I&FP does not come into play at all”. Posing the question of whether the reputation of I&FP is more important in attracting potential targets rather than forming alliances, the response is affirmative.

1. II) The SMEs’ point of view

According to the manager of Océane de Restauration, the reputation of I&FP has had no impact on the formation of alliances, nor in the establishment of trust. He says: “No, that had no impact on the establishment of trust. You must know that, in general, company managers are even rather skeptical and reluctant toward investment funds when they want to sell their company. A fund, at least initially, does not induce a situation of trust”. What allowed a situation of trust to be established was the presence of the manager of Océane de Restauration – the first company acquired under the Caterine Restauration build-up – with I&FP when they approached other potential targets. He says: “It was mainly Franck, thus the fund which did market research and who suggested potential targets. Next, we discussed it together and then we went to see them… The reluctance was less than if the fund had been there on its own. They have more confidence if they see that a company which has already been acquired by the fund is doing well… anyway, companies often don’t know the various funds. I think that investment funds can only be known by the other funds”. The manager of Centrale de Restauration Martel adds that the reputation of the PEF may have an impact on the formation of intraportfolio alliances between companies from different build-ups, given that the managers of the SMEs forming the alliances are all supported by the PEF, know it and can judge its specific reputation. As to relations with external companies, in particular commercial relations, the presence of a PEF and its reputation as a PEF in general even have a negative effect. Thus, as they put it: “This [reputation] may play a role, but at the level of commercial relations, it has a negative effect. It may play a role, actually, in synergies with other build-ups, but apart from that it doesn’t give a good brand image at the commercial level… Let’s say that competitors are quick to say that if you’re part of an investment fund, that’s a bad sign and it may have a bad impact on the brand and thus on the company”.

In summary, interviewees testified to only an indirect correlation between the reputation of I&FP and the formation of alliances, whether intra- or extraportfolio. The reputation of I&FP plays a positive role in the attraction of potential targets for build-ups. Partners for alliances which I&FP might suggest to their supported companies are generally companies which had been, at one time, potential targets for a build-up. At that moment, therefore, the actors know each other and do not rely on reputation. We can nevertheless conclude that this plays an indirectly positive role in the formation of alliances, as it facilitates the establishment of a link of trust when approaching companies as potential targets. For managers of companies supported by I&FP, it is difficult for them to judge the reputation of I&FP because they are not familiar with the different PEFs. If I&FP’s reputation plays a role, it is mainly in the case of intraportfolio or inter-build-up alliances, as these companies are then familiar with the PEF. In the case of extraportfolio alliances, it is hard for them to judge.

1. d) Anonymous PEF

The points of view of the PEF and the SMEs have been analyzed on the basis of interview transcripts. After presenting the general opinion of Anonymous PEF’s reputation, we contrast their points of view. In general, they converge.

Table 3.11. The Anonymous PEF field in relation to Hypothesis 1

Reputation	PEF	Alliance 1	Alliance 2
	Anonymous PEF	YYY	WWW	XXX
General opinion	Emerging reputation because Anonymous PEF has existed since the end of 2009. Known at the regional level, in Lorraine, as an actor specializing in venture capital which is beginning to make itself known in France. Reputation may attract files or attract external companies to alliances.	Does not know PEFs and can thus only judge the reputation of PEFs in general. But the presence of Anonymous PEF very clearly has a positive impact on the formation of alliances.	Reputation of an industrial PEF.
Intraportfolio alliances	Reputation is favorable to establishment of trust. The process of selecting candidates for funding is quite long which reflects, in the end, the quality of the companies chosen.	Facilitates establishment of trust and reassures at the level of the quality of the company and the product and at the level of its financial stability.	The fact of sharing a common shareholder facilitates the establishment of trust. This allows reassurance about the quality of the company and the product and at the level of its financial stability.
Extraportfolio alliances	Anonymous PEF through its reputation and presence plays a stabilizing role, allowing the reduction of distrust from external partners. It gives visibility to the SMEs it supports which are themselves too small. Apart from that, the presence of Anonymous PEF demonstrates the financial stability and quality of the companies it supports.	The presence of Anonymous PEF grants visibility and reassures external actors at the level of the company’s quality and financial stability since PEFs are known to undertake in-depth analyses before investing. In summary, it facilitates the establishment of trust.	The presence of Anonymous PEF is reassuring at the level of the quality of the company and its product – a material – because the main shareholder of Anonymous PEF is the world leader in the field of materials. In addition, this is reassuring at the level of the company’s stability and the merits of its vision, as this has already been audited and validated by Anonymous PEF. This therefore contributes to the establishment of trust.

1. 1) General perception of reputation
   1. I) The PEF’s point of view

According to Anonymous PEF, its reputation is emerging as the fund has existed since 2009. It is that of a regional PEF, known especially in Lorraine as an actor specializing in venture capital which is beginning to make itself known a little throughout France.

1. II) The SMEs’ point of view

The company managers interviewed found it difficult to judge the reputation of Anonymous PEF. Generally, they were not familiar with the different PEFs. In their experience, it was a regional PEF with a financial side, but which put no unhealthy pressure on them. Anonymous PEF seeks to develop and grow its portfolio companies and to profit from them but in a reasoned fashion. For one of the company managers, what was important when the PEF made its approach was the understanding between him and the PEF: “It’s a question of personalities”, he said.

1. 2) Impact of the PEF’s reputation on the formation of alliances
   1. I) The PEF’s point of view

Anonymous PEF’s reputation allows its investment manager to receive files, and may play a role in attracting external partners in the case of extraportfolio alliances. The presence of Anonymous PEF, due to its reputation, gives reassurance as much at the level of the financial stability of the companies it supports as at the level of their quality. This therefore helps in creating a situation of trust. The PEF helps give visibility to the companies it supports, which are too small and have too little visibility to make themselves known to other companies. Thus, Anonymous PEF’s investment manager comments on the fact that Anonymous PEF, by its mere presence, may play a role in the formation of alliances, even extraportfolio alliances for its portfolio companies. He gives as an example: “Yes, in giving credibility to XXX [one of Anonymous PEF’s portfolio companies], knowing that there was a fund involved in the company’s capital reassured the two partners who were big companies. CSUN is quoted on the Nasdaq. So, this helped reassure them about XXX’s financial capacity”. He gives another example: “For an example of alliances, WWW made an alliance with a company called Modulife in Lyon. Modulife is a builder of individual houses. Modulife asked to meet with us. I went to Lyon with WWW to meet them. And it’s the same, Modulife was reassured by the fact that a fund was part of WWW’s capital. They were reassured about WWW’s financial base. This was an important factor and a catalyst in the relationship”.

As to certification of quality, the responses given were not very clear. The presence of the PEF helped reassure partners about the quality of the companies it supports in the sense that it believes in the supported company’s project. On the other hand, this does not give reassurance about the quality of the product, for example. Thus, Anonymous PEF’s investment manager says: “We’re not consulted about the quality of their product… The companies know more than me about their products/services. So, if they ask for my opinion, it’s on their potential for development and their financial base. We offer a guarantee that the company will not go bankrupt from one day to the next and that, if it gets into trouble, we will be able to help, there will be something to cope with temporary cash-flow difficulties. So, we’re a factor giving credibility to the alliance”.

1. II) The SMEs’ point of view

As to the formation of extraportfolio alliances, none have been realized yet but, according to the managers of its portfolio SMEs, Anonymous PEF’s reputation clearly has a positive impact on the judgment of third parties concerning them. The presence of Anonymous PEF is reassuring because third parties know that a PEF conducts careful analysis before investing, and ensures serious follow-up after investment. On the one hand, this is reassuring at the level of the financial stability of the companies. On the other, it is reassuring about the quality of their projects and products, as stated by the manager of one of the portfolio companies: “It’s reassuring, yes, of course. Our product is a material. And the majority shareholder of Anonymous PEF… is actually one of the world leaders in the materials sector”. Finally, the presence of Anonymous PEF contributes to the establishment of trust which is favorable to the formation of alliances. In comparison to a situation without Anonymous PEF, its presence facilitates the building of relationships.

In the case of forming intraportfolio alliances, the fact of sharing the same PEF gives reassurance as to the quality of the other company and its financial stability. The establishment of trust is clearly facilitated.

To summarize, interviewees testified that the reputation of Anonymous PEF clearly has a positive impact on the formation of alliances for its portfolio companies. The presence of Anonymous PEF is reassuring because third parties know that a PEF carries out careful analysis before investing and closely follows up after investment. This is reassuring, on the one hand, at the level of the companies’ financial stability. On the other hand, this gives reassurance as to the quality of their projects.

1. B) The second level of analysis

After having analyzed, field by field, the points of view of the PEFs, on the one hand, and of the SMEs forming alliances on the other hand, we now contrast the summary obtained for each field to the others. This comprises a second level of analysis. Table 3.12 presents the summaries of the points of view from the different fields analyzed above.

Table 3.12. Interfield summary in relation to Hypothesis 1

Reputation	Siparex Group	Demeter Partners	Industries et Finances Partenaires	Anonymous PEF
Supports the hypothesis	Yes	Yes	Yes, but weak	Yes
Mechanism of causality	The presence of the PEF is above all a sign of trust, of the seriousness of the company being supported and reassures prospective partners as to the merits of the company or its projects and is also reassuring at the level of its financial stability.	The presence of the PEF is above all a sign of trust, of the seriousness of the company being supported and reassures prospective partners as to the merits of the company or its projects, but is reassuring mainly and equally at the level of its financial stability.	I&FP’s reputation plays a positive role in attracting potential targets for build-ups, and partners for alliances that I&FP may suggest to its supported companies are generally companies who were, at one time, potential targets for a build-up.	The presence of Anonymous PEF is reassuring because third parties know that a PEF carries out careful analysis before investing and closely follows up after investment. This is reassuring, on the one hand, at the level of the companies’ financial stability. On the other hand, this gives reassurance about the quality of their projects.

Across the four fields of analysis, we can confirm a positive link between the reputation of the PEF and its portfolio companies forming alliances. However, it is clear that the reputation of the PEF has two components: (1) the reputation of PEFs in general; (2) the PEF’s specific reputation compared to its peers. The component which is always involved is the first. Depending on the actors on which it acts, it may be positive or negative. In the first case, actors see a PEF as an agent which conducts scrupulous and detailed analyses before deciding to support a prospect. In addition, there is often quite a large selection. This may help to establish trust in the company supported by private equity on the part of potential partners to an alliance. In the second case, some actors associate a negative impact to the presence of a PEF. This is often clients who, in a customer–supplier relationship with a company supported by private equity (playing the supplier’s role), may fear too intrusive an intervention from the PEF. The latter could, according to the actors, suddenly raise the price. In general, at least according to the cases analyzed in this study, although the two effects are present, the first is more important for the formation of alliances.

As to the second component of reputation (the specific reputation of a PEF in regard to its peers), it is mostly perceived by the PEFs among themselves. Company managers often cannot make a judgment unless they already know the PEF in question. This may be the case when it involves a sectoral or regional PEF or, again, when companies, although they may be outside the investment portfolio of the PEF in question, have already been in contact with it because they were, at one time, a prospect for the PEF. This personal reputation can also be good or bad.

In summary, for three of the four fields of analysis, all the interviewees saw a positive link between the PEF’s reputation and the formation of alliances. The PEF’s presence allows faster establishment of trust between prospective alliance partners. It gives reassurance about the financial stability and the quality of the company supported. In the case of I&FP, the link between the PEF’s reputation and the formation of alliances is only indirect. Reputation plays a role in attracting potential targets which, at a later date, are the source of ideas for constituting potential alliance partners. Altogether, the presumed mechanism of causality is supported.

The mechanism of discipline

Concerning the mechanism of discipline and from the point of view of the SMEs forming the alliance, we offer the following hypothesis:

HYPOTHESIS 4.– The PEF taking a majority stake in the companies which it supports has a positive impact on the formation of alliances.

Prospective alliance partners may refuse to join it for fear that their partners may adopt uncooperative behaviors once the alliance is formed. PEFs who sit on the board of directors of at least one of the companies forming the alliance are in a position to observe and monitor their contribution. They may facilitate the exchange of information, and are able to control the behavior of alliance partners and to discipline them in cases where they adopt uncooperative behaviors. This may reduce the mistrust of prospective partners in advance of the transaction and, thus, have a positive impact on the formation of the alliance. As PEFs may hold either a minority or a majority stake in the companies they support, this disciplinary role for PEFs is assumed to be greater when they hold a majority stake.

1. A) First level of analysis
   1. a) Siparex Group

The points of view of the PEF and of the SMEs have been analyzed based on the interview transcripts and grouped into Table 3.13. Their points of view on the possible link between taking a majority stake and the formation of alliances are contrasted there. Generally, they overlap. The responses are detailed below the table.

Table 3.13. The Siparex Group field in relation to Hypothesis 4

Stakes	PEF	Alliance 1		Alliance 2
	Siparex Group	Sophem	Mooviin	Meseo
General opinion	In general, intervention as a large minority which implies active support, but which is not interventionist or intrusive.	Large minority stake.	Minority stake.	Minority stake.
Intra- and extraportfolio alliances	Our portfolio companies listen to us, we bring them ideas and reflections and they are keen for them. But we are only a source of proposals. We might intervene as a conciliator between alliance partners, but the situation has never arisen.	We discuss, we listen to them, but we decide. We listen to them as much as if they held a majority stake.	We listen to the proposals, but we decide on implementation. If we do not want to discuss with someone or if there is a problem, we do not get Siparex to intervene, that doesn’t work.	It changes nothing in building relationships whether it is minority or majority. We listen to them just the same. They might intervene in a structuring way in case of worries, but that has never happened. They introduce us, they initiate the relationship, after that, it is for us to work it out. They are not supposed to intervene in the case of conflict, that is not their role.

1. 1) General situation

Siparex Group takes minority stakes, on the order of 15% to 30% of capital, in the companies it supports.

1. 2) Impact of the PEF taking majority stakes on the formation of alliances
   1. I) The PEF’s point of view

According to the point of view of the portfolio manager we interviewed, the fact of holding a minority or a majority stake has an impact on the formation of alliances. By taking minority stakes, Siparex can only suggest ideas. Their adoption and implementation depend totally on the supported companies. In contrast, in the case of taking majority stakes, this would suggest a much stronger commitment, allowing the PEF to ensure that its ideas are implemented.

A situation of disagreement or conflict between alliance partners has not yet arisen in the presence of Siparex Group. However, one might imagine Siparex Group intervening as a conciliator. Companies taking part in an alliance might also request the PEF to do this. This potential intervention does not depend on the percentage of equity held by Siparex Group in the supported companies.

1. II) The SMEs’ point of view

In the case of the company Sophem, Siparex Group holds a minority stake with preferential rights. At the level of alliance formation, the fact that Siparex Group holds a minority stake changes nothing for the manager of Sophem, who listens to the PEF and its proposals in all cases. The manager of Meseo shares this point of view. In his case, Siparex Group takes a minority stake in their company. The fact that the PEF does not hold a majority stake changes nothing for the formation of alliances. He listens to Siparex Group in all cases and in the same way as if it had taken a majority stake. A situation of conflict or disagreement has not yet arisen in his case within the alliances which he has been able to form under Siparex Group. Nevertheless, in such a case, the manager of Meseo can envisage a structuring intervention on the part of the PEF, yet without intervening too much in the relationship. The conduct of the alliance must be led by the company managers, not by the PEF. On the other hand, Siparex Group initiates the relationship. The PEF acts as the trigger which facilitates the mutual introduction of the companies wishing to form an alliance. Its presence allows the companies to be heard by each other in a more attentive fashion.

In the case of the company Mooviin, Siparex Group also took a minority stake. For the company’s manager, he listens to Siparex Group and its ideas are welcome if they are good. If they do not interest him, he does not listen to them. For him, it was out of the question to work with a PEF which took a majority stake in his company. He wishes to retain decision-making power. In case of conflicts between partners in alliances, it is out of the question for Siparex Group to intervene. The alliance relationship must be managed by the company managers taking part in it, and not by the PEF. On the other hand, Siparex Group does intervene to facilitate the initiation of exchanges between prospective alliance partners.

In summary, interviewees confirmed the existence of a positive correlation between taking a majority stake and the formation of alliances, whether intra- or extraportfolio alliances. However, the existence of a minority stake does not change much. The PEF’s ideas as to the formation of alliances are listened to in the same way, and implemented by company managers if the idea seems good to them. As to the PEF intervening in case of conflicts, the situation has not arisen, but all participants in the study can imagine the PEF intervening to sort out conflicts, which would be independent of whether it took majority or minority stakes.

1. b) Demeter Partners

The different points of view have been analyzed on basis of the interview transcripts and grouped in Table 3.14. The opinion of the PEF and the SMEs on the link between the PEF taking a majority stake and the formation of alliances by its portfolio companies are compared there. Generally, they overlap. The responses are given in detail below the table.

Table 3.14. The Demeter Partners field in relation to Hypothesis 4

Stakes	PEF	Alliance 1		Alliance 2
	Demeter Partners	Comarth	IES	Panosol	Eurener
General opinion	Mainly large minority stakes and very few majority. But in the latter case, Demeter does not force anything. They therefore do not intervene differently if they hold a majority or minority stake. Demeter does the same work of building relationships, no matter the percentage of capital held. This has no impact on the formation of alliances.	Half stake (50%) (large minority).	Majority stake.	Minority stake.
Intra- and extraportfolio alliances	Demeter suggests ideas independently of the percentage of capital held in its portfolio companies. In contrast, its opinion may weigh more if Demeter holds a majority stake. The work of building relationships remains exactly the same. It may be less taken up when Demeter intervenes as a “small”-minority. Demeter may intervene in conflicts between alliance partners as an arbitrator acting as a “go-between”. Demeter does not force anything, but may prevent things going wrong. This is even requested by companies. The service is provided independently of the proportion of capital held (most often, Demeter holds a minority stake).	The percentage of capital held changes absolutely nothing for the formation of alliances. Demeter’s ideas and proposals are listened to and received in the same way as if they held a majority stake. Demeter is really there to provide follow-up and support, regardless of the level of investment. Demeter has already intervened a few times in discussions in cases of technical problems, by greasing the wheels or as an arbitrator.	It does not matter if Demeter holds a majority or minority stake for the formation of alliances. Demeter’s propositions are listened to in any case.	Panosol listens to and is attentive to Demeter’s proposals and ideas, but does it at its own pace. What might change is that, with a majority stake, the weight of the advice on what the managers do is stronger.

1. 1) General situation

Demeter Partners mainly intervenes by taking minority stakes on the order of 30% to 45% of capital. Two portfolio companies involve minority stakes on the order of 5%. Taking a majority stake is not common.

1. 2) Impact of the PEF taking majority stakes on the formation of alliances
   1. I) The PEF’s point of view

According to Demeter Partners, taking minority or majority stakes has no impact on the formation of alliances. Demeter Partners stated as follows: “There are even companies in which we only hold 5%, but we do the same work of relationship building, etc. as for the others”. On the other hand, where one might disagree is whether Demeter Partner’s suggestions concerning alliance formations are taken up. A PEF has more control on the take-up of its suggestions in the case of a majority stake compared to a minority stake. The investment manager we interviewed agreed: “Afterwards, is it followed up, actually, the fact that we hold 95% or 5%, that will change… it’s true that when you’re a small minority, it’s less often followed up”.

Asked whether Demeter Partners might intervene in case of conflicts between alliance partners, the response is affirmative. The interviewee stated as follows: “Yes. And I’d say that it would be requested by the alliance partners. What they also like is, when things are not necessarily going well, when negotiations aren’t succeeding, they can go through us. After that, we’re not going to say, “you do it like this”, because we don’t decide for them, but we act as a go-between, as an arbitrator. Like that, it allows them to have a parallel track. We’re the first-line negotiator between, for example, the buyer and the seller [it involves customer–supplier relationships, often made concrete in the form of commercial agreements] So they have a path through the shareholder which facilitates things. We won’t force things, but we can avoid things going wrong if they can’t make a deal. And so, yes, we sometimes do that… In the end, in a customer–supplier relationship, it evolves over time and there will always be moments when you have to come and grease the wheels”. Such intervention by Demeter Partners is not systematic. The PEF only intervenes when negotiations between partners in a customer–supplier type alliance has gone badly. It is moreover often the manager of the company in question who calls Demeter Partners to ask them to intervene.

1. II) The SMEs’ point of view

Demeter Partners holds a minority stake in the interviewed companies Panosol and Eurener, a half share (50–50%) in the company Comarth and a majority stake in IES.

For IES, the fact that Demeter Partners holds a majority stake has no impact on the formation of alliances. If Demeter Partners held a minority stake, they would listen to them in the same way. This opinion is shared by the manager of Comarth, in which Demeter Partners holds 50%. According to him, Demeter Partners is really there to bring support, independent of the level of shareholding. This changes nothing.

Demeter Partners also intervened to provide support in exchanges between Comarth and its alliance partner IES. “We could also envisage Demeter Partners intervening as an arbitrator,” explains Comarth’s manager. But a situation which would require such an intervention has not yet arisen. For Panosol, in which Demeter Partners held a minority stake during the time of its support, Demeter Partner’s level of shareholding played a role in the sense that Panosol could consider Demeter Partners’ ideas and implement them at their own pace, without pressure. On the other hand, Panosol paid attention to Demeter Partners’ suggestions independently of its percentage stake.

In summary, interviewees confirmed a positive correlation between taking a majority stake and the formation of alliances, whether intra- or extraportfolio alliances. However, taking a minority stake changed little, unless it was that the PEF had no impact over whether its ideas were implemented. However, for company managers, the PEF’s ideas on the formation of alliances were heard in the same way as if they had a majority stake. As to intervention by the PEF in case of conflicts, this situation arose in one of the alliances studied. In that case, Demeter Partners intervened to grease the wheels to act as a go-between, play the role of an arbitrator. For partners in alliances where such a situation did not arise, they could imagine the PEF intervening to resolve conflicts, which would be independent of whether they held a minority or a majority stake.

1. c) Industries et Finances Partenaires

The points of view of the PEF and the SMEs have been analyzed on the basis of the interview transcripts and grouped into Table 3.15. Their points of view on the link between taking a majority stake and the formation of alliances are contrasted, and their responses are detailed immediately below the table.

Table 3.15. The Industries et Finances Partenaires field in relation to Hypothesis 4

Stakes	PEF	Caterine Restauration
	Industries et Finances Partenaires	Océane de Restauration	Centrale de Restauration Martel
General opinion	Majority stakes.	Majority stake.	Majority stake.
Intra- and extraportfolio alliances	I&FP inspires the strategy, it should be up to the companies to implement it. A majority stake ensures that company managers will take a greater part in certain debates, especially concerning strategic decisions in order to find solutions to certain problems or issues in a concerted manner. As a majority owner, I&FP’s ideas are received more easily. Aside from that, other PEFs may get involved with minority but still significant stakes, of “large” minority size (40–45%). There may have been a few misunderstandings and I&FP intervened to resolve them. I&FP’s majority stake clearly facilitated the regulation of managers’ behavior, greasing the wheels. I&FP is the promoter, but the managers agreed to this principle at the moment they sold their company.	For Océane de Restauration, the level of shareholding had no impact on the formation of alliances. As majority shareholders, I&FP have not been at all intrusive. They have not forced anything. There have not been any disagreements and so they have not had to intervene at that level. But if they had been required to, it would have made no difference whether they were majority or minority owners. I&FP is listened to in any case. What is important, equally and in contrast, is that even if I&FP bought the companies, the managers were able to keep a minority stake.	No response.

1. 1) General situation
2. I&FP only takes majority stakes.

1. 2) Impact of the PEF taking majority stakes on the formation of alliances
   1. I) The PEF’s point of view

For a build-up operation, it is necessary to take majority stakes. Given that I&FP strongly centers its activity on a few companies, the means of portfolio diversification are limited. On the other hand, the PEF has to hold a majority and be the project leader. This is important to be able to infuse companies with the strategy to follow that comes from I&FP. On the other hand, the companies will put it into operation and implement it. The associate director of I&FP adds: “… The idea of why we hold so firmly to this fact of being a majority, it’s to have a balance of powers. I return a little to what you’re saying, it’s that we don’t want to, we’re not looking to ‘oblige’ people to do things. We want to oblige them to participate in the debate”. Concerning the realization of synergies between the companies under a build-up, ideas are received and accepted more easily due to the fact that I&FP holds a majority stake. This is also possible with the PEF in a minority but, often, it entails a nevertheless significant stake of up to 45%, accompanied by minority rights granting them decision-making rights on certain points. Apart from that, everything also depends on the relationship between the PEF and the manager of the company being supported.

For the build-up we studied, there had been no conflicts of interest, but a number of mild disagreements between the companies in the beginning. The fact that I&FP held a majority stake allowed it to be clearly identified as the promotor and to convince the companies to pursue I&FP’s strategy, as well as to come to agreement between them.

1. II) The SMEs’ point of view

The manager of Océane de Restauration does not think that the intervention of I&FP would have been different if they had taken minority stakes in the companies. Although I&FP held a majority stake, the PEF had not been intrusive. It had in no way imposed its decisions. Neither had I&FP intervened in case of conflicts, given that there had been no disagreements. If it had had to intervene, it would have made no difference whether the PEF held a majority stake or not.

In contrast, for the manager of Centrale de Restauration Martel, taking majority stakes made a great difference: it introduced the possibility of I&FP replacing the existing managers. Two points need to be considered with regard to this assertion. The first is that I&FP has only changed existing managers when they wished to retire in any case and were looking for a successor. The second point is the question of whether a link with the formation of alliances exists here. For the manager of Centrale de Restauration Martel, decisions to be taken arrived in a hierarchical way. I&FP thus had a large impact on decision-making, also in terms of the formation of synergies.

Taking account of the context, characterized here by the formation of synergies within a build-up, divergences in points of view between the two managers interviewed may be explained by the fact that the manager of Océane de Restauration belongs to the mother company in the Caterine Restauration build-up. This was the first company acquired by I&FP in the process of forming the Caterine Restauration group. This company has a particular role within the group. Many good practices have been taken from this company and redeployed in the other companies of the group. The manager also plays a particular, key role. He was suggested to lead the group, taking the role of director-general of the holding company. Although he refused this post, he nevertheless holds a special position. In particular, he supported I&FP in choosing potential targets for other acquisitions under Caterine Restauration. He was thus strongly involved in the strategic decisions of the group when it was being formed, from the outset. In contrast, the interviewed manager of Centrale de Restauration Martel succeeded the manager in place at the time of I&FP’s purchase of the company. He was subsequently appointed and had to deal with this situation without having participated in its development from the outset.

In summary, we can identify that the PEF confirms the existence of a positive correlation between taking a majority stake and the formation of alliances which are mostly intra-build-up alliances, thus intraportfolio alliances. This allows it to ensure that company managers participate in debates, in particular about implementing ideas for synergies between them. From the point of view of the companies, however, taking majority or minority stakes changes nothing for the formation of alliances. On the one hand, they will listen to the PEF’s ideas in the same way, for example about the formation of alliances. On the other hand, the companies can imagine the PEF intervening in case of conflicts, even if the situation has never actually arisen, to resolve conflicts – this, again, independently of the level of shareholding.

1. d) Anonymous PEF

The analysis was performed based on the interview transcripts. A summary is found in Table 3.16. The points of view of the PEF and the SMEs on the link between taking a majority stake and the formation of alliances are contrasted and overlap. The responses are detailed immediately below the table.

Table 3.16. The Anonymous PEF field in relation to Hypothesis 4

Stakes	PEF	Alliance 1	Alliance 2
	Anonymous PEF	YYY	WWW	XXX
General opinion	Minority stakes.	Minority stake.	WWW: minority stake with blocking minority.
Intra- and extraportfolio alliances	The fact of holding a minority or a majority of capital makes a difference in the formation of alliances. As a minority, Anonymous PEF can only suggest initiating a relationship; managing alliances remains in the hands of company managers. As a majority, they can steer the SME and thus directly manage alliances. Ideas weigh more heavily on managers’ decisions. As a minority, Anonymous PEF can intervene as an arbitrator if there are difficulties. But this situation has never arisen. As a majority, they could intervene in a more drastic way and impose rules to avoid any conflict.	Whether Anonymous PEF intervenes as a minority or a minority is unimportant for the formation of alliances. Their ideas are received and their proposals listened to in all cases. As a minority, Anonymous PEF could very well intervene in case of differences or problems to resolve them.	Anonymous PEF’s ideas are listened to independently of the percentage of capital it holds. What is important is the personal relationship and the skills of the person. It is and remains a partner independently of the percentage of capital held. Anonymous PEF might also intervene in case of difficulties as an intermediary, to grease the wheels.

1. 1) General situation
2. Anonymous PEF only holds minority stakes in the companies it supports.

1. 2) Impact of the PEF taking majority stakes on the formation of alliances
   1. I) The PEF’s point of view

According to Anonymous PEF, taking majority or minority stakes has an impact on the formation of alliances. The owner of a majority stake steers the company, which entails a much stronger responsibility for managing alliances. In taking minority stakes, one can only suggest them. The opinion and ideas of Anonymous PEF might weigh more heavily on the decisions of its portfolio companies’ managers in the case of taking a majority stake. They say: “We suggest [the alliance]. But if the managers don’t want to give it any time, we don’t have many levers of action”. Taking minority stakes, however, does not prevent Anonymous PEF from being able to intervene as an arbitrator in the case of disagreements between alliance partners. Yet such a situation has not yet arisen.

1. II) The SMEs’ point of view

In one of its portfolio companies, Anonymous PEF holds a minority stake without a controlling minority. For the company manager, the fact that Anonymous PEF holds a minority and not a majority stake in his company changes nothing at all for the formation of alliances. He listens to Anonymous PEF in the same way as if they held a majority stake. He also feels that Anonymous PEF could very well intervene in a situation of conflict of interest between partners in an alliance so as to resolve the situation, although this has not yet happened. In a situation where Anonymous PEF held a majority of shares, they could then theoretically take advantage of this and control the situation, but this is not Anonymous PEF’s approach. And, even while holding a minority stake, its impact would be positive for the alliance.

In another portfolio company, Anonymous PEF holds a minority stake of 33%, thus having a blocking minority but not controlling it. For the company manager, it is not the level of shareholding that counts for the formation of alliances but, above all, the personal relationship. Anonymous PEF is for him a partner, and thus to be listened to on principle. If their ideas are good, he follows them up. Anonymous PEF, although holding a minority stake, could still intervene as a mediator in case of conflicts between parties to an alliance, although such a situation has never occurred. In this case, what is important, again, is relations between people, and not whether Anonymous PEF holds a majority or a minority stake.

In summary, we have learned that Anonymous PEF taking minority stakes means that the PEF can only suggest ideas for alliances but cannot ensure that these are put into effect. Nevertheless, their ideas are mostly received favorably by the managers of supported companies. Anonymous PEF can also imagine itself intervening in case of conflict as an arbitrator or to grease the wheels, which is also independent of whether it takes a minority or majority stake. From the point of view of the SMEs, whether the PEF holds a minority or a majority stake changes nothing for the formation of alliances. The PEF’s ideas are listened to in the same way. Like the PEF, they can imagine the PEF intervening to resolve conflicts, which would be independent of whether it holds majority or minority stakes.

1. B) The second level of analysis

After having analyzed field by field the points of view of the PEFs, on the one hand, and the SMEs forming alliances, on the other hand, we now contrast the summary obtained for each field to the others (Table 3.17).

Table 3.17. Interfield summary in relation to Hypothesis 4

Stakes	Siparex Group	Demeter Partners	Industries et Finances Partenaires	Anonymous PEF
Supports the hypothesis	Yes	Yes	Yes	Yes
Mechanism of causality	Although participants agree with the suggested mechanism of causality, according to them, the PEF taking a minority stake has a similar impact on the formation of alliances.	Although participants agree with the suggested mechanism of causality, according to them, the PEF taking a minority stake has a similar impact on the formation of alliances.	For the PEF, taking a majority stake is necessary. From the point of view of the SMEs, one of the two managers shared the point of view of the other fields, while the other shared the point of view of I&FP.	Although participants agree with the suggested mechanism of causality, according to them, the PEF taking a minority stake has a similar impact on the formation of alliances.

All the fields of analysis support the idea that taking a majority stake has a positive impact on the formation of alliances, and thus reinforces the hypothesis. However, across the different fields, it can be seen that the effect of taking a majority stake is not greater than that of taking a minority stake for the formation of alliances. Thus, although in the case of taking a minority stake the PEFs can only suggest ideas for alliances (contrary to taking majority stakes, where they could ensure that their suggestions are enacted), the managers of companies supported by private equity agree in saying that:

• – ideas for alliances from PEFs when they hold a minority stake are listened to with the same attention as when they hold a majority stake, and that they are enacted if they are good;
• – the PEFs who take majority stakes do not in any way impose alliances, but only propose them;
• – in the end, what is most important is the personal relationship between the PEF and the managers of the companies they support.

As for intervention by PEFs in case of conflicts, everyone could imagine such an intervention, both from the point of view of the PEFs as well as the companies interviewed, independently of the level of shareholding and with the same effect.

Thus, even though the hypothesis is supported, the variable “taking majority stakes” does not seem decisive; taking a minority stake seems to have a similar effect on the roles that PEFs are supposed to play in the formation of alliances. The disciplinary role of PEFs in forming alliances is well confirmed, but not the variable which is supposed to measure that role.

The PEF’s point of view

The hypotheses from the point of view of the PEFs in light of contractual theories are divided into two sub-points. We begin with those which relate to the interests of the PEFs in the formation of alliances for their portfolio companies, and continue with the hypotheses which may relate to the relativization of the roles of the PEFs. A distinction between intra- and extraportfolio types of alliance is made, again, where necessary.

The interests of PEFs in the formation of alliances

Dealing with the question of the interest of PEFs in the formation of alliances, we put forward two hypotheses in the theoretical section. The first is as follows:

HYPOTHESIS 2.– The formation of alliances by SMEs which they support with partners who possess reputational capital constitutes, for a PEF which has weak reputational capital, a complementary or even substitute mechanism for the certification role played by well-known PEFs.

Well-known PEFs may play a role in the certification of quality of the companies which they support, which reduces the distrust of external actors. However, a PEF only plays a certification role if it possesses a certain reputational capital. This is built progressively by accumulation of experience and performance. We may thus suppose that the younger a PEF, the weaker its reputational capital, in which case, it must therefore build it up. A quicker way, if it does not possess enough of it, is to link the company it supports with established partners, for example, in the context of a strategic alliance. Thus, the formation of alliances for the SMEs it supports with partners who have some reputational capital constitutes, for a PEF with weak reputational capital, a substitute mechanism for the role of certification played by well-known PEFs.

• A) First level of analysis
  • a) Siparex Group

The points of view of the PEF and the SMEs have been analyzed on the basis of interview transcripts and grouped into Table 3.18. Their points of view in regard to the impact of weak reputation on the PEF and on the formation of alliances are contrasted therein. In general, they overlap. The responses are detailed immediately below the table.

Table 3.18. The Siparex Group field with regard to Hypothesis 2

Weak reputation	PEF	Alliance 1	Alliance 2
	Siparex Group	Sophem	Mooviin	Meseo
General opinion	Siparex has never tried to build links for its portfolio companies in any particular way with companies which are well known on the markets. Neither currently, nor when the group was still young and not yet established on the private equity market. Companies which today enjoy a certain image or reputation often did not enjoy this at the time when they were supported by Siparex.	Did not get the impression that Siparex tried to build links specifically between its portfolio companies and companies who were well known on the markets.	Did not get the impression that Siparex tried to build links specifically between its portfolio companies and companies who were well known on the markets. Rather, Mooviin was the motivating party.	Did not get the impression that Siparex tried to build links specifically between its portfolio companies and companies who were well known on the markets.

1. 1) General situation

Siparex Group is one of the main French PEFs. It has existed for 36 years. It thus already possesses strong reputational capital. It replied to the question asked by referring to the past, when Siparex Group was still young and not established on the market.

1. 2) Impact of a weak reputation of the PEF on the formation of alliances
   1. I) The PEF’s point of view

Siparex Group has never at any time tried to specifically build relationships for its portfolio companies with companies which are well known and established on the market.

The portfolio manager interviewed indicated that it was necessary to consider the following: “The Siparex group, when it began, supported companies which are now leading companies in the Lyon region. And, because it was the local economic structure with companies which were on the development path. So, today they’re European or French leaders, when at the beginning that wasn’t the case. They didn’t have the image they have today”.

1. II) The SMEs’ point of view

None of the three company managers interviewed had the impression that, at any time, Siparex Group had tried to build relationships specifically between their company and companies with good reputations on the market.

In summary, we can identify that the PEF and the mangers of its portfolio companies who were interviewed agreed in saying that Siparex Group had never tried to specifically build relationships between its portfolio companies and partners who were well known and established on the markets.

1. b) Demeter Partners

The analysis is based on interview transcripts. A summary is drawn up in Table 3.19.

There, we contrast the points of view of the PEF and the SMEs with regard to the impact of a weak reputation of the PEF and the formation of alliances. Most often, they agree.

Table 3.19. The Demeter Partners field in relation to Hypothesis 2

Weak reputation	PEF	Alliance 1		Alliance 2
	Demeter Partners	Comarth	IES	Panosol	Eurener
General opinion	Demeter does not specifically try to build relationships between its portfolio companies and well-known companies. The alliance has to make sense. What is important is not so much reputation as the personal relationship, the trust between persons.	No, Demeter has not tried, at any time, to specifically build relationships between its portfolio companies and companies who are well known on the markets.	There is no different approach.	No. The alliances suggested to us were intraportfolio alliances.

1. 1) General situation

Demeter Partners is a PEF which is long established on the French private equity market. The PEF thus replied to our question with reference to the past, when Demeter Partners was still young and not established on the market.

1. 2) Impact of a weak reputation of the PEF on the formation of alliances
   1. I) The PEF’s point of view

Demeter Partners did not specifically try to build alliance relationships between its portfolio companies and partners who were well known on the markets. The important thing was the personal relationship between the people in the companies who wished to form an alliance.

1. II) The SMEs’ point of view

None of the three company managers we spoke to had the impression that, at any given moment, Demeter Partners had tried to specifically build relationships between their company and companies who were well known on the markets.

In summary, we can identify that the PEF and the managers of its portfolio companies whom we interviewed agreed that Demeter Partners had never at any time tried to specifically build relationships between its portfolio companies and partners who were well known and established on the markets.

1. c) Industries et Finances Partenaires

The points of view of the PEF and the SMEs have been analyzed on the basis of the interview transcripts and grouped into Table 3.20. Their points of view with regard to the impact of the PEF having a weak reputation and the formation of alliances are contrasted therein. In general, they overlap. The responses are detailed immediately below the table.

Table 3.20. The Industries et Finances Partenaires field in relation to Hypothesis 2

Weak reputation	PEF	Caterine Restauration
	Industries et Finances Partenaires	Océane de Restauration	Centrale de Restauration Martel
General opinion	No, not at all. I&FP has never tried to specifically make alliances between its portfolio companies and well-known actors on the markets. In any case, I&FP’s partners already from the outset had a certain reputation and visibility in the sector. From the beginning, then, I&FP already had some kind of track record.	No, I&FP has never at any time tried to specifically build relationships between its portfolio companies and companies who were well known on the market.	No, I&FP has never at any time tried to specifically build relationships between its portfolio companies and companies who were well known on the market.

1. 1) General situation

I&FP is a PEF which is established on the private equity market. It is above all recognized as specializing in the field of build-ups. I&FP replied to the question given by referring to the past, when I&FP was still young.

1. 2) Impact of a weak reputation of the PEF on the formation of alliances
   1. I) The PEF’s point of view

I&FP has never tried to build alliance relationships between the companies it acquires with partners who are specifically well known on the markets. An element of context must be mentioned. I&FP’s reputation has never been weak, given that the founders and partners of I&FP already had a good reputation in the field, since they possessed a track record and a history in the build-up field. At its foundation, I&FP’s reputation therefore was based on that of its partners.

1. II) The SMEs’ point of view

Neither of the two company managers interviewed had the impression that at any moment, I&FP had tried to build a relationship between their company or another under the Caterine Restauration build-up specifically with companies who were well known on the markets.

In summary, we can identify that the PEF and the managers of its portfolio companies who were interviewed agreed that I&FP has never at any time tried to specifically build relationships between its portfolio companies and partners who were well known and established on the markets.

1. d) Anonymous PEF

The method of analysis is the same as for the other fields. Table 3.21 gives an overview. In general, the points of view of the participants overlap. The responses are detailed just below the table.

Table 3.21. The Anonymous PEF field in relation to Hypothesis 2

Weak reputation	PEF	Alliance 1	Alliance 2
	Anonymous PEF	YYY	WWW	XXX
General opinion	Anonymous PEF tried to build relationships between its portfolio companies and its main shareholder, but this did not succeed. After this, the portfolio companies tried to build links with major groups, but that came from them and not from Anonymous PEF. But Anonymous PEF has never tried to build relationships between its portfolio companies with well-known companies to “compensate” for its emerging reputation with respect to established PEFs, except in linking them with other PEFs for syndication.	No, Anonymous PEF has never tried, at any moment, to build relationships specifically between its portfolio companies and companies who are well known on the markets.	No, Anonymous PEF has never tried, at any moment, to build relationships specifically between its portfolio companies and companies who are well known on the markets.

1. 1) General situation

Anonymous PEF is quite a young PEF, created in 2009. It is already known in Lorraine and is beginning to make itself known throughout France. Its reputation is still under construction.

1. 2) Impact of a weak reputation of the PEF on the formation of alliances
   1. I) The PEF’s point of view

Anonymous PEF has tried to build alliance relationships between some of its portfolio companies and one of its shareholders. However, this relationship with this well-known actor was not the result of the PEF wanting to compensate for its emerging reputation with respect to well-known PEFs. The investment manager made the following comment: “No, except if that partner is a fund. As we have limited means, which may not be sufficient to support the company according to its needs, we have several times tried to reach out to other funds”.

1. II) The SMEs’ point of view

The two managers of Anonymous PEF’s portfolio companies whom we interviewed did not have the impression that Anonymous PEF had specifically tried to build relationships between their companies and partners who were well known on the markets.

In summary, we can identify that the PEF and the managers of its portfolio companies surveyed agree. They say that Anonymous PEF has not, at any time, tried to specifically put its portfolio companies into relationships with well-known and established companies on the markets to compensate for its still-emerging relationship with respect to more established and well-known companies on the French private equity market.

1. B) The second level of analysis

After having analyzed field by field the points of view of the PEFs on the one hand, and of the SMEs forming alliances on the other hand, we now contrast the summary characterizing each field to those of the others. This summary is presented in Table 3.22.

Table 3.22. Interfield summary in relation to Hypothesis 2

Weak reputation	Siparex Group	Demeter Partners	Industries et Finances Partenaires	Anonymous PEF
Supports the hypothesis	No	No	No	No
Mechanism of causality	The PEF and its portfolio companies surveyed agree that Siparex Group has never, at any time, tried to specifically build relationships between its portfolio companies with partners who are well known and established on the markets.	The PEF and its portfolio companies surveyed agree that Demeter Partners has never, at any time, tried to specifically build relationships between its portfolio companies with partners who are well known and established on the markets.	The PEF and its portfolio companies surveyed agree that I&FP has never, at any time, tried to specifically build relationships between its portfolio companies with partners who are well known and established on the markets.	The PEF and its portfolio companies surveyed agree that Anonymous PEF has never, at any time, tried to specifically build relationships between its portfolio companies with partners who are well known and established on the markets.

Over the four fields of analysis, three PEFs could be considered as already established and well known on the markets. One field of analysis includes a young PEF, still becoming established on the market and building its reputational capital. The first three PEFs thus replied from a historical point of view, when they were still young. None of the participants in the study, without exception, supported the hypothesis. These PEF have never, at any time, tried to build relationships specifically between their portfolio companies and alliance partners who are well known on the markets to make up for their own lack of reputation or visibility with regard to more established PEFs on the French private equity market.

We move on to the second hypothesis posed from the point of view of PEFs to identify their role in the formation of alliances for their portfolio companies:

HYPOTHESIS 3.– The presence of the State, of a region, or of a Competitiveness Cluster in the capital of a PEF positively impacts the formation of alliances.

In France, certain investors such as the State, a region or a Competitiveness Cluster set up vehicles of investment in the goal of promoting the development of companies in certain sectors and/or regions as well as the establishment of partnerships, alliances or cooperation between actors. This type of investor should be able to positively influence the PEF so that it becomes involved in the formation of alliances, at least in the case that this PEF takes the form of a joint-stock company (with or without the status of venture capital company), thus allowing investors the possibility of influencing the decisions taken by the PEF. So, we hypothesize that the presence of the State, a region or a Competitiveness Cluster in the capital of the PEF positively impacts the formation of alliances for the companies that it finances. We suppose that this link is strengthened when the PEF takes the legal form of a joint-stock venture capital company.

1. A) First level of analysis
   1. a) Siparex Group

The points of view of the PEF and of the SMEs have been analyzed on the basis of the interview transcripts and grouped into Table 3.23. There we contrast their opinions on the impact of investors on the PEF and on the formation of alliances. Most often, they agree. The responses are detailed just below the table.

Table 3.23. The Siparex Group field in relation to Hypothesis 3

State/region/Competitiveness Cluster	PEF	Alliance 1		Alliance 2
	Siparex Group	Sophem	Mooviin	Meseo
General opinion	The Deposits and Consignments Fund is an underwriter of Siparex. But the underwriters of the Siparex Group have no impact on the activity of forming alliances that Siparex might initiate for its portfolio companies. Siparex is an independent management company.	Up until now, the presence of such types of underwriters of Siparex or of Competitiveness Clusters has had no impact on the activity of alliance formation for Sophem.	Siparex enables Mooviin to attend regional meetings but there has been no formation of alliances due to this.	Meseo has been put into contact with regional enterprises, but the presence of certain underwriters in Siparex or of Competitiveness Clusters has played no part in this.

1. 1) General situation

Siparex Group takes the form of a VCMF with an independent management company. The underwriters of Siparex Group’s funds therefore, in principle, have no impact on the PEF’s decisions.

1. 2) Impact of the PEF’s investors on the formation of alliances
   1. I) The PEF’s point of view

CDC-Enterprises is an underwriter of Siparex Group’s funds, but has no impact on its decision-making. Siparex Group is not directly involved in Competitiveness Clusters. On the other hand, some of its portfolio companies are. According to the investment manager interviewed, this does not play a significant role in alliance formation for Siparex Group’s portfolio companies.

1. II) The SMEs’ point of view

According to the three interviewed managers of Siparex Group’s portfolio companies, the PEF does enable the building of relationships with regional businesses, yet without being limited to this role. The underwriters of Siparex Group or the involvement of Competitiveness Clusters have had no impact on the alliances they have been able to form.

In summary, we can identify that Siparex Group takes the form of a VCMF. By definition, the management society holds decision-making power independent of the wishes of underwriters. The PEF and the managers of its portfolio companies interviewed agreed that the underwriters of Siparex Group have no impact on the PEF’s formation of alliances.

1. b) Demeter Partners

Table 3.24 presents the summary of the analysis of points of view of the PEF and the SMEs concerning the impact of investors in the PEF on the formation of alliances. In general, they converge. Their opinions are explained below the table.

Table 3.24. The Demeter Partners field in relation to Hypothesis 3

State/region/Competitiveness Cluster	PEF	Alliance 1		Alliance 2
	Demeter Partners	Comarth	IES	Panosol	Eurener
General opinion	The underwriters of Demeter’s funds include in particular the Deposits and Consignments Fund, as well as the European Investment Fund for Demeter’s second fund. Demeter is also part of several Competitiveness Clusters. If that has any impact on the formation of alliances, it is an indirect one. Directly, Demeter has never noticed it. This impacts on Demeter’s presentation to potential portfolio companies. All its underwriters are passive and equal.	This does nothing in particular; no relationship has been built due to their presence or to a Competitiveness Cluster.	Involvement in a Competitiveness Cluster, thanks to Demeter, but this is not really important for the activity of alliance formation for IES.	There have been no relationships built through Demeter due to these actors.
Extraportfolio alliances	As the Competitiveness Clusters are in the field of specialization for Demeter’s investments, they allow Demeter’s portfolio companies to form relationships with other companies in the Cluster, to make themselves known. This may have an impact on the activity of alliance formation for Demeter’s portfolio companies.

1. 1) General situation
2. Demeter Partners takes the form of a VCMF with an independent management company. The underwriters of the funds thus, in principle, have no impact on the PEF’s decision-making. CDC-Enterprises and the European Investment Fund are underwriters of the PEF, which is also involved in several Competitiveness Clusters.

1. 2) Impact of the PEF’s investors on the formation of alliances
   1. I) The PEF’s point of view

CDC-Enterprises and the European Investment Fund are passive underwriters of Demeter Partners’ funds, without impact on decision-making. They thus have no influence on the activity of forming alliances. On the other hand, Demeter Partners’ participation in Competitiveness Clusters may have an impact. The clusters may give access to potential alliance partners, other members of the cluster. In this case, it involves extraportfolio alliances, thus involving an alliance partner external to Demeter Partners’ portfolio of investment.

1. II) The SMEs’ point of view

The company Panosol was not linked to either a Competitiveness Cluster nor to other companies in the region. On the other hand, the company IES was linked to a Competitiveness Cluster through Demeter Partners as well as regional companies, but that had no impact on its activity of forming alliances. The manager of the company Comarth shares these views. The underwriters of Demeter Partners or the involvement of the PEF in a Competitiveness Cluster have had no impact on the formation of alliances for Comarth.

In summary, we can identify that Demeter Partners takes the form of a VCMF. By definition, the management society holds decision-making power independent of the wishes of underwriters. The PEF and the managers of its portfolio companies interviewed agree that the underwriters of Demeter Partners have no impact on the activity of alliance formation by the PEF. Their points of view diverge slightly concerning the impact of Demeter Partners’ participation in Competitiveness Clusters. From the point of view of the PEF, this may allow access to external alliance partners for Demeter Partners’ investment portfolio. However, this has not played a role for the managers of Demeter Partners’ portfolio companies whom we interviewed.

1. c) Industries et Finances Partenaires

The points of view of the PEF and the SMEs are collected in Table 3.25. They do not overlap.

Table 3.25. The Industries et Finances Partenaires field in relation to Hypothesis 3

State/region/Competitiveness Cluster	PEF	Caterine Restauration
	Industries et Finances Partenaires	Océane de Restauration	Centrale de Restauration Martel
General opinion	The underwriters of I&FP cannot have an impact on the activities of alliance formation of I&FP’s portfolio companies. Also, there is no involvement on the part of I&FP in a Competitiveness Cluster.	There has been no involvement by the underwriters of I&FP in the formation of alliance, nor in Competitiveness Clusters.	There has been no involvement by the underwriters of I&FP in the formation of alliance, nor in Competitiveness Clusters.

1. 1) General situation

I&FP takes the form of a VCMF. Its underwriters have no impact on its decision-making. I&FP is not involved in Competitiveness Clusters. The PEF and the company managers interviewed share this point of view.

As to the influence of their legal status – joint-stock company versus mutual fund – on the independence of decision-making from their underwriters/shareholders, the associate director we interviewed at I&FP advises considering the following: “So, look at this. Our functioning at I&FP makes us really autonomous in our decision-making with regard to our investors. The mutual-fund structure was set up for this. That is, to put legal protection between us and our investors. So for us, the investment decisions are taken by I&FP. This is not always the case, though, in certain funds. In other investment funds, you have the mutual-fund structure, the management company and there’s an investment committee. And sometimes, the investment committee is representatives of the unit-holders. As for us, we really respect what the mutual-fund structure was designed for. The investment decisions are taken by us. But it does happen that other investment teams put a bias into this process, by creating an investment committee, and who are its members? Representatives of the unit-holders, who will have an influence on decisions”.

In summary, we can identify that I&FP takes the form of a VCMF and that the management company holds decision-making power independent of its underwriters. The PEF and the managers of its portfolio companies interviewed agree that the underwriters of I&FP have no impact on the activity of alliance formation by the PEF. Also, I&FP is not involved in Competitiveness Clusters.

1. d) Anonymous PEF

Again, the points of view of the participants in the study have been analyzed on the basis of interview transcripts. The conclusions are shown in Table 3.26. These points of view diverge. Participants’ responses are detailed immediately following.

Table 3.26. The Anonymous PEF field in relation to Hypothesis 3

State/region/Competitiveness Cluster	PEF	Alliance 1	Alliance 2
	Anonymous PEF	YYY	WWW	XXX
General opinion	Yes, the shareholders of Anonymous PEF (in particularly the State or the Lorraine region) may have an impact on Anonymous PEF’s activity of alliance formation.	Do not think that Anonymous PEF’s shareholders have an impact on the alliances that Anonymous PEF allows YYY to form. On the contrary, it’s one of the tasks of Anonymous PEF to link companies who can realize synergies. The Competitiveness Clusters allow YYY to access contacts for alliances, but YYY was already linked to these clusters. It’s not Anonymous PEF which allowed them to get in contact.	This does not produce alliances. On the other hand, it allows the company to be legitimized when it approaches potential partners on various levels (industrial, financial). Neither does the Competitiveness Cluster in which Anonymous PEF is involved have any influence on the alliances formed by WWW, given that its leading themes are not centered on WWW’s products or themes.

1. 1) General situation
2. Anonymous PEF takes the form of a venture-capital company, having the legal form of a joint-stock company. The shareholders may therefore influence, through the board of directors, the decisions taken by Anonymous PEF’s investment committee. Historically, Anonymous PEF was created by the Lorraine region and the Materialia Competitiveness Cluster. CDC-Enterprises and the Lorraine region are among Anonymous PEF’s shareholders. The ex-president of a Competitiveness Cluster (Materialia) is the head of Anonymous PEF’s investment committee. Also, Anonymous PEF is one of the PEFs which participate in an alliance of Competitiveness Clusters (including the Fibres and Materialia clusters) as auditors. The goal of this alliance is to award certain companies the label of “cluster innovative companies”. This is given to companies and their projects identified on the basis of their quality, and facilitates access to private equity for these companies.

1. 2) Impact of the PEF’s investors on the formation of alliances
   1. I) The PEF’s point of view

According to the investment manager for Anonymous PEF, the investors have indeed had an impact on their activity of alliance formation. He explains as follows: “… the fact that Anonymous PEF was created by the region and by Materialia in the context of revitalizing the territory, it is clear that they ask me to invest, not only to make money, but also to find and to give rise to industrial projects. So, based on that, I really have an interest in bringing people onto my territory, to try to create industrial activity; or maybe I am more sensitive to that, in fact, because of the history of the fund and its industrial positioning… Yes, absolutely. As the region has invested in Anonymous PEF, the region demands as a rule that the investments we make have economic benefits which are favorable to the Lorraine region. That means that if we invest outside the Lorraine region, we will strive to generate alliances with companies from Lorraine so as to respond to the wishes of our shareholders. So if the region did not expressly ask that of me, I would be less active in seeking out alliances”.

1. II) The SMEs’ point of view

The manager of the YYY company does not think the shareholders of Anonymous PEF have influence over the activity of alliance formation by Anonymous PEF for its portfolio companies. He thinks that this is part of Anonymous PEF’s role, to build relationships for its portfolio companies if synergies are possible. Apart from that, YYY is part of a Competitiveness Cluster, but unrelated to any intervention from Anonymous PEF. This cluster facilitates building relationships involving YYY for the formation of alliances. However, Anonymous PEF plays no role in this.

The point of view of the manager of the WWW company is similar. It is involved in different Competitiveness Clusters, in particularly Materialia, but this is independent of any involvement by Anonymous PEF. For him, the shareholders of Anonymous PEF have no particular impact on the activity of Anonymous PEF forming alliances for its portfolio companies. He thus explains: “No, that [the presence of the Lorraine Region or the State in the capital of Anonymous PEF] changes nothing. It’s a question of good sense. From the moment that it [Anonymous PEF] took a stake in us, it became our partner. It is a ‘partner’ in the good development of the company. In addition, an investment fund is interested in investing and getting a return on its investment. So, when the fund gets into a company, it is already thinking on how it will be able to exit. So, it has every interest in the company developing. And so, the more it has worked for us and for the company, the greater the chance that the company will develop, attract interest from other financial or industrial investors, or that we will be able to self-finance buying back its shares. In any case, it also minimizes its risk. So our interests are completely convergent”.

In summary, we can identify that Anonymous PEF takes the form of a venture-capital company and, thus, a joint-stock company. On the one hand, for the PEF it is clear that its shareholders – in particular the Lorraine region – have a positive impact on their activity of forming alliances. On the other hand, on the contrary, the company managers have not perceived this influence from Anonymous PEF’s shareholders. Generally, the involvement of Anonymous PEF in Competitiveness Clusters may give access to potential alliance partners, but this has not been the case for the alliances which the companies whose managers were interviewed have been able to form.

1. B) The second level of analysis

After having analyzed, field by field, the point of view of the PEF on the one hand, and that of the SMEs forming alliances on the other hand, we now contrast the summary characterizing each field to those of the others. This summary is presented in Table 3.27.

Table 3.27. Interfield summary in relation to Hypothesis 3

State/region/Competitiveness Cluster	Siparex Group	Demeter Partners	Industries et Finances Partenaires	Anonymous PEF
Supports the hypothesis	Yes	Yes	Yes	Yes
Mechanism of causality	The PEF and the managers of its portfolio companies interviewed agree that the underwriters of Siparex Group have no impact on the activity of alliance formation by the PEF.	The PEF and the managers of its portfolio companies interviewed agree that the underwriters of Demeter Partners have no impact on the activity of alliance formation by the PEF. Their points of view diverge slightly concerning the impact of Demeter Partners’ participation in Competitiveness Clusters. From the point of view of the PEF, this may allow access to alliance partners external to Demeter Partners’ investment portfolios. However, this has not played a role for the managers of Demeter Partners’ portfolio companies whom we interviewed.	The PEF and the managers of its portfolio companies interviewed agree that the underwriters of I&FP have no impact on the activity of alliance formation by the PEF. Also, I&FP is not involved in Competitiveness Clusters.	For the PEF, it is clear that its shareholders – in particular the Lorraine region – have a positive impact on its activity of forming alliances. On the other hand, the company managers have not perceived this influence from Anonymous PEF’s shareholders. Anonymous PEF’s involvement in Competitiveness Clusters may give access to potential alliance partners, but this has not been the case for the alliances which the company managers interviewed have been able to form.

Among the four fields of analysis, three of the PEFs take the form of a VCMF, consisting of a management company which, in general takes decisions independently of the underwriters. One of the PEFs takes the form of a venture-capital company, within which the shareholders can influence, through the board of directors, the investment committee’s decisions. On the one hand, we raised the hypothesis that the presence of the State or of a region in the capital of a PEF, or the involvement of the PEF in one or several Competitiveness Clusters, has a positive impact on the formation of alliances for its portfolio companies. This link has been weakly or even not at all supported. Involvement in a Competitiveness Cluster does not seem significant for the formation of alliances, although it might be a source of ideas for alliances of the extraportfolio type. The presence of the State or of a region in the capital of a PEF only had an impact for one of the four PEFs interviewed. Overall, the link is very weak. On the other hand, the mechanism of causality advanced indicates that the role of the State/a region in the capital of a PEF on the formation of alliances is reinforced when the PEF takes the form of a venture-capital joint-stock company. We therefore expected that, in the case of the PEF taking the form of a joint-stock company, the presence of certain shareholders would have a greater impact on the formation of alliances than in the case when the PEFs take the form of a venture-capital mutual fund. In fact, the case of the PEF which took the form of a venture-capital company and which had the Lorraine region in its capital, and indirectly the State via CDC-Enterprises, illustrates this mechanism of causality. In particular, the presence of the Lorraine region in its capital promotes the involvement of the PEF in the formation of alliances for its portfolio companies. This influence of the PEF’s shareholders was, however, not perceived by the managers of the supported companies, for which the involvement of the PEF in the formation of alliances for their company came from themselves. The case of the three PEFs taking the form of a venture-capital mutual fund also supports the mechanism of causality. In fact, for these PEFs, the underwriters have no impact on the decisions taken by the management company.

HYPOTHESIS 5a.– In the presence of a PEF, the number of alliances formed by a company impacts negatively on the formation of a new alliance.

According to the literature, a PEF may be reluctant for its portfolio companies to form too many alliances as, given that an alliance partner generally holds decision-making rights in the context of the alliance, this may result in a conflict of interests situation between the PEF and the alliance partner. However, according to knowledge-based theories, we might also assume the opposite effect, that is, an absence of reluctance from PEFs to the formation of a large number of alliances. This is because the more the latter form alliances, the more they gain experience, visibility and credibility, which allows them to build reputation and legitimacy on the markets. So, on the one hand, we raised the hypothesis that, in the presence of a PEF, the number of alliances formed by a company negatively impacts the formation of a new alliance. On the other hand, we formulated the alternative hypothesis that in the presence of a PEF, the number of alliances formed by a company positively impacts the formation of a new alliance.

1. A) First level of analysis
   1. a) Siparex Group

We based our analysis on the interview transcripts to summarize the points of view of the PEF and the SMEs in Table 3.28. They are in accord. The responses are detailed just below the table.

Table 3.28. The Siparex Group field in relation to Hypothesis 5

Number of previous alliances	PEF	Alliance 1	Alliance 2
	Siparex Group	Sophem	Mooviin	Meseo
General opinion	The number of alliances previously formed by a portfolio company has no impact on the formation of an additional alliance. What is important is that the alliance makes sense at the level of the strategy pursued by the company. By default, Siparex is favorable to any alliance which is in phase with the strategy pursued by one of its portfolio companies.	Assuming that the alliance brings value to Siparex, there is no reason for it to object.	Assuming that the alliance brings value to Siparex, there is no reason for it to object.	Assuming that the alliance brings value to Siparex, there is no reason for it to object.
Conflicts of interest	Theoretically, these may always occur. Siparex tries to avoid this type of situation. By default, there should never be conflicts between Siparex and an alliance partner of one of its portfolio companies.	Assuming that the alliance brings value to Siparex, there is no reason for there to be conflicts. Aside from that, it is always imaginable, but this clearly constitutes an exceptional case.	This may always happen, for example, if they take a shareholding with one of our competitors. But they avoid this, there are ethical rules.	Assuming that the alliance brings value to Siparex, there is no reason for there to be conflicts. Aside from that, it is always imaginable, but this clearly constitutes an exceptional case.
Gaining competence in the formation of alliances	This effect clearly outweighs the first.	This effect clearly outweighs the first.		This effect clearly outweighs the first.

1. 1) The PEF’s point of view

For the associate director interviewed at Siparex Group, the number of alliances previously formed by their portfolio companies has no impact on the motivation of Siparex Group to form a further alliance for the company in question. He explained as follows: “Assuming that it will serve the strategy pursued by the company and on which we are engaged with the manager, there’s no worry”. Siparex Group is favorable to all alliances assuming that they make sense at the level of the strategy pursued by the company, and that there is no situation of competition between the alliance partner and the portfolio company. In the latter case, there might be a conflict of interest between the two companies. Siparex Group may then not propose the alliance, but the situation where Siparex Group is favorable to the alliance clearly dominates.

1. 2) The SMEs’ point of view

For the manager of the Mooviin company, Siparex Group is by default favorable to any alliance which might bring added value. But the formation of alliances through Siparex Group for its companies remains very rare. The managers of the companies Meseo and Sophem share this point of view. By default, Siparex Group is favorable to any alliance as long as it makes sense at the level of the strategy pursued by the company or that it adds value. A situation where Siparex Group would be reluctant for an alliance to be formed due to conflicts of interest between the PEF and the alliance partner is always imaginable, but has not yet occurred and would constitute an exception. Siparex Group’s favorable attitude to the formation of alliances by its portfolio company largely predominates.

In summary, we can identify that the PEF as well as the company managers interviewed agree that, by default, Siparex Group is favorable to any alliance as long as it makes sense at the level of the strategy being pursued.

1. b) Demeter Partners

The points of view of the interviewees have again been grouped and contrasted in Table 3.29. Their points of view as to the impact of alliances previously formed by the portfolio companies on the motivation of the PEF to intervene in the formation of a new alliance for its companies are similar. The responses are detailed just below the table.

Table 3.29. The Demeter Partners field in relation to Hypothesis 5

Number of previous alliances	PEF	Alliance 1		Alliance 2
	Demeter Partners	Comarth	IES	Panosol	Eurener
General opinion	It is important that companies have alliances for their business.	There has never been disagreement with Demeter about the formation of alliances.	Demeter is favorable, by default, to any formation of alliances.	In the case of competing companies, Demeter will not propose alliances.
Conflicts of interest	By default, there are no conflicts of interest; Demeter is mostly favorable to the formation of alliances.	Do not see why Demeter would have a conflict of interest with an alliance partner.	Do not see why Demeter would have a conflict of interest with an alliance partner.
Gaining competence in alliance formation	Yes, the company increases its visibility and its turnover.

1. 1) The PEF’s point of view

Demeter Partners is entirely favorable to its portfolio companies forming alliances “since every time it means more projects, more business”, explains the portfolio manager. The number of alliances previously formed by its portfolio companies has no negative impact on the formation of an additional alliance for that company. In contrast, we can say that the number of alliances previously formed by its portfolio companies positively impacts the formation of a new alliance, given that it brings to the business “… visibility, reputation and turnover”, as the portfolio manager says. He adds: “We’re clearly favorable, and we’re even active for them to do it as often as possible”.

1. 2) The SMEs’ point of view

For the director of the Panosol company, there is in fact a limit on the part of Demeter Partners to the number of alliances to be formed with their company. This is because the number of possible alliances in their sector of activity is limited, all the more because Demeter Partners will not invest in companies in direct competition with Panosol. For the manager of the IES company, Demeter Partners is by default favorable to any alliance that he would envisage forming. He does not see why there would be reluctance on the part of the PEF. According to the manager of Comarth, there is no reason why the number of alliances that his company has already been able to form would have an impact on the formation of a new alliances under the auspices of Demeter Partners. There should not be any conflict of interest.

In summary, we can identify that the PEF as much as the company managers interviewed agree that Demeter Partners is favorable to any alliance as long as it makes sense at the level of the strategy being pursued, and that the knowledge-based argument prevails over the contractual argument.

1. c) Industries et Finances Partenaires

The points of view of the PEF and the SMEs regarding the impact of alliances previously formed by the portfolio companies on the inclination of the PEF to intervene in the formation of a new alliance for these companies converge. They are summarized in Table 3.30, and their responses are detailed immediately below the table.

Table 3.30. The Industries et Finances Partenaires field with regard to Hypothesis 5

Number of previous alliances	PEF	Caterine Restauration
	Industries et Finances Partenaires	Océane de Restauration	Centrale de Restauration Martel
General opinion	The number of alliances previously formed by the companies acquired by I&FP does not affect their formation of alliances.	This has no impact except that if we have already formed a large number, there may be less need to build more relationships for us to meet our needs.	By default, I&FP is favorable to any alliance as long as it makes sense and is in phase with the strategy being pursued. The number of alliances previously formed by the companies supported by I&FP does not affect the formation of a new alliance.
Conflicts of interest	No, there are no conflicts of interest.	No, there are no conflicts of interest.	No.
Gaining competences in forming alliances	Theoretically yes, but I&FP does not have many examples of alliances outside the synergies between the companies under a buildup. Theoretically this second effect outweighs the first.	Yes, maybe.

1. 1) The PEF’s point of view

For I&FP, the number of alliances previously formed by the companies they support has no impact on the motivation of I&FP to intervene in the formation of an additional alliance for these companies. This does not correspond to the logic of their thought. Taking majority stakes, I&FP cannot support a very large number of companies. If alliances are limited, it is because intraportfolio alliances are the majority in the case of I&FP and, to diversify their risks, I&FP will not invest in companies with too closely related activities.

In theory, the positive effect outweighs the negative effect. Thus, the number of alliances previously formed by the companies supported by I&FP should have a positive rather than a negative influence on the formation of a new alliance for these companies, from the fact that they gain experience. In practice, however, I&FP does not have many examples.

1. 2) The SMEs’ point of view

For the manager of Océane de Restauration, there is no reason why I&FP should be reluctant for their company to form an additional alliance even if it has already formed a certain number of them. He mostly agrees with the fact that if the number of alliances previously formed has any impact, it will be positive. However, for him, the number of previously formed alliances has no impact, neither positive nor negative, on the formation of another alliance under I&FP.

For the manager of Centrale de Restauration Martel, he does not see why I&FP would be reluctant for him to form an alliance. By default, I&FP is favorable to any alliance that his company might form, as long as it makes sense at the level of the strategy being pursued.

In conclusion, we can identify that the PEF as much as the company managers interviewed agree that, by default, there is no reason why I&FP should be opposed to the formation of an alliances for its supported companies. The knowledge-based argument prevails over the contractual argument, even if there is no practical example.

1. d) Anonymous PEF

Table 3.31 presents the points of view of the PEF and the SMEs with regard to the impact of alliances previously formed by the portfolio companies on the motivation of the PEF to intervene in the formation of a new alliance for these companies. They overlap. The responses are detailed just below the table.

Table 3.31. The Anonymous PEF field in relation to Hypothesis 5

Number of previous alliances	PEF	Alliance 1	Alliance 2
	Anonymous PEF	YYY	WWW	XXX
General opinion	The number of previous alliances has no impact on the formation of a new alliance by Anonymous PEF’s portfolio companies.	Any alliance that is well thought-out is in the interest of Anonymous PEF.	The more alliances, the better.
Conflicts of interest	No, on the contrary, Anonymous PEF is favorable to the formation of alliances. There are more opportunities than conflicts in forming alliances.	Don’t see why there would be conflicts or opposition to the formation of alliances.	No, do not agree. There are no conflicts of interest between Anonymous PEF and the alliance partners. This might possibly be different in the case of a PEF which is the subsidiary of a group or a company.
Gaining competence in forming alliances	Yes, that plays a role. The company gains experience, perhaps it opens itself to applications or markets which they had not thought of before.	Absolutely. Anonymous PEF is completely favorable to the formation of alliances.	Yes, the more alliances are formed, the more customer–supplier relations there are, the more we have a foundation, a legitimacy, the more recognition we have. And this is always true up until our business model is exhausted.

1. 1) The PEF’s point of view

There is no reluctance on the part of Anonymous PEF for its portfolio companies to form alliances, independently of the number of alliances that they have previously been able to form. Thus, Anonymous PEF’s investment manager says: “I’m favorable to it, on the contrary. I’m favorable that there should be as many alliances as possible. If there are conflicts, we will manage them. But I think there are more opportunities than conflicts in forming alliances”. He even thinks that the number of alliances previously formed may have a positive effect on the formation of a new alliance, because, as he puts it, the company: “… gains experience, it opens itself maybe to applications, to markets which they had not thought of before. Yes, that plays a role”.

1. 2) The SMEs’ point of view

For the manager of the WWW company, Anonymous PEF is totally favorable to his company forming alliances. In any case, it is not reluctant. The number of alliances previously formed by WWW has a positive impact on the formation of new alliances. He thus explains: “In any case, the more alliances we form, the more partnerships we form, the more customer–supplier relations we have, the more foundation, the more legitimacy we have, the more recognition we have. And that will always be true, whether we had a fund in our capital or not, until our business model has been exhausted, that is, until we have explored all the possible applications of our materials for the products of other companies on the market. And there is still room..”. According to WWW’s manager, this situation might be different for a purely industrial PEF: “Now, for a purely industrial fund, this might indeed be different… For example, if Anonymous PEF belonged directly to its main shareholder, that might penalize us for certain alliances if, in fact, either the alliance partner or that shareholder did not want us to form the alliance because of that fact, or, if there was a fund in our capital which directly represented a subsidiary of the shareholder, or if the shareholder rejected the alliance partner for example if they were a competitor. So, we could have the situation where “I am forbidden from working with so-and-so because so-and-so is involved”. There are people who might then find themselves in a situation of competition”. But this is not the case for Anonymous PEF.

The manager of the YYY company is of the same opinion. As long as the alliance is well thought-out, Anonymous PEF is favorable. There is no reason why they should be opposed, for example, due to conflicts of interest.

In summary, we can identify that the PEF as much as the company managers interviewed agree that there is no reason for Anonymous PEF to be opposed to the formation of an alliance for the companies it supports. The knowledge-based argument prevails over the contractual argument.

1. B) The second level of analysis

After having analyzed field by field the points of view of the PEFs on the one hand, and the SMEs forming alliances on the other hand, we now compare the summary characterizing each field to those of the others. This summary is presented in Table 3.32.

Table 3.32. Interfield summary in relation to Hypothesis 5

Number of previous alliances	Siparex Group	Demeter Partners	Industries et Finances Partenaires	Anonymous PEF
Supports the hypothesis	Knowledge based	Knowledge based	Knowledge based	Knowledge based
Mechanism of causality	The PEF as much as the company managers interviewed agree that, by default, Siparex Group is favorable to any alliance as long as it makes sense at the level of the strategy being pursued.	The PEF as much as the company managers interviewed agree that Demeter Partners is favorable to any alliance as long as it makes sense at the level of the strategy being pursued and that the knowledge-based argument prevails over the contractual argument.	The PEF as much as the company managers interviewed agree that, by default, there should be no reason why I&FP should be opposed to the formation of an alliance by the companies they support. Theoretically, the knowledge-based argument prevails over the contractual argument even if there is no practical example.	The PEF as much as the company managers interviewed agree that there is no reason why Anonymous PEF should be opposed to the formation of an alliance by the companies it supports. The knowledge-based argument prevails over the contractual argument.

All the participants in the study across the four fields of analysis agree in claiming that the number of alliances previously formed by the companies supported by private equity is not important, or matters little, to the PEF’s motivation to intervene in the formation of a new alliance by these companies. The PEFs are, by default, favorable to any alliance as long as it makes sense at the level of the strategy pursued by the supported company, and it brings value. If the number of alliances previously formed ever has an impact on the formation of a new alliance, it is then mostly positive. The interviewees recognize the relevance of the knowledge-based argument according to which the number of alliances previously formed has a positive impact on the formation of a new alliance, given that the company in question gains experience.

The relativity of the PEF’s roles

Two hypotheses have been raised which consider the fact that PEFs may not be the only mechanisms which have an impact on the formation of alliances by their portfolio companies. The first is as follows:

HYPOTHESIS 6.– The role of PEFs in strategic alliances is higher in non-publicly listed companies.

The mechanisms which may influence the discretionary space of company managers and, in light of the problematic, their decisions on forming alliances, generally arise when a company goes public and is listed on the stock exchange. A company going public and inviting other shareholders lessens the relative weight of the decisions of the PEF. We have therefore posed the hypothesis that the role of PEFs in strategic alliances is more important for non-publicly listed companies.

1. A) First level of analysis
   1. a) Siparex Group

The points of view of the PEF and of the SMEs have been analyzed on the basis of the interview transcripts and grouped into Table 3.33. Their points of view regarding the impact of the publicly listed or otherwise status of companies on the formation of alliances are contrasted there. They are in agreement. The responses are detailed just below the table.

Table 3.33. The Siparex Group field in relation to Hypothesis 6

Publicly listed status of companies	PEF	Alliance 1	Alliance 2
	Siparex Group	Sophem	Mooviin	Meseo
General opinion	Only support unlisted companies. So the fact that Siparex’s role in the formation of alliances is higher in the presence of unlisted companies naturally follows.	Sophem is unlisted and thinks that the intervention of Siparex in the formation of alliances does not differ according to whether the companies are listed or not.	Mooviin is not listed and has no alliances with listed partners. In addition, Mooviin does not think that Siparex’s intervention in the formation of alliances differs according to whether the alliances are listed or not.	Meseo is not listed and has no alliances with listed partners. In addition, Meseo does not think that Siparex’s intervention in the formation of alliances differs according to whether the alliances are listed or not.

1. 1) General situation

Siparex Group only intervenes in unlisted companies. In addition, the PEF is not in the habit of sitting on the boards (board of directors, supervisory board or other strategic boards) of its portfolio companies after exit, onto the stock exchange for example.

1. 2) Impact of listed/unlisted status of portfolio-companies on alliance formation
   1. I) The PEF’s point of view

It naturally follows that the role of Siparex Group in the formation of alliances is more intense between unlisted companies, given that the PEF only intervenes in unlisted companies.

1. II) The SMEs’ point of view

The Mooviin company is neither publicly listed, nor in any relationship with a listed company. Its manager does not think that Siparex Group’s intervention in the formation of alliances by his company would be different if his company or an alliance partner were publicly listed. The manager of the Sophem company shares this point of view. The Meseo company is not publicly listed, either. In contrast, Meseo’s manager does not know if their alliance partners are listed. However, he does not think that Siparex Group’s intervention in the formation of alliances for his company differs due to the listed/non-listed status of one of the partners in the alliances involved. For him, it is first and foremost a question of personal relations.

In summary, we can identify that for the PEF, it is clear that the role of Siparex Group in the formation of alliances is greater for non-listed companies, given that it does not intervene in listed companies. As for the company managers interviewed, they do not think that Siparex Group’s intervention would differ if one of the companies participating in alliances were listed on the stock exchange. In reality, the alliances formed under Siparex Group are between unlisted companies.

1. b) Demeter Partners

Table 3.34 presents a summary of the points of view of the PEF and the SMEs regarding the impact of the listed/unlisted status of the companies on the formation of alliances. Their points of view converge. Our summary is based on the interview transcripts. The responses are detailed just below the table.

Table 3.34. The Demeter Partners field in relation to Hypothesis 6

Publicly listed status of companies	PEF	Alliance 1		Alliance 2
	Demeter Partners	Comarth	IES	Panosol	Eurener
General opinion	Demeter also supports listed companies. Often it is Demeter who helped them go public. In fact, this does play a role for Demeter’s intervention in the formation of alliances since Demeter can intervene much less in the case where the companies are publicly listed.	Comarth is not publicly listed. Its alliance partners, sometimes. According to Comarth, Demeter’s intervention would not differ if it were publicly listed.	Neither IES nor Comarth (its alliance partner) are publicly listed and, according to IES, Demeter’s approach to forming alliances for IES would be the same if IES, its alliance partner or both were listed on the stock exchange. Demeter built a relationship between IES and BMW and directly made the contacts.	Neither Panosol nor Eurener are publicly listed. According to Panosol, Demeter’s approach would have been the same if Panosol, Eurener or both had been listed on the stock exchange.

1. 1) General situation

Demeter Partners supports some companies listed on the Alternext exchange. In general, it is Demeter Partners who allowed them to become publicly listed.

1. 2) Impact of listed/unlisted status of portfolio companies on alliance formation
   1. I) The PEF’s point of view

For Demeter Partners, its intervention differs in forming alliances depending on whether the company involved is publicly listed or not. Thus, the investment manager interviewed stated: “we can intervene much less when the company is listed than when it isn’t… This is due to the fact that, as soon as you are listed, you have rules to follow, rules which are much more stringent concerning information. You have press releases. You’re a member of the board of directors, but your relations are less close after the company is listed than before. Once the company is listed, there is obviously a barrier between the shareholders and the company at the level of transmission of information. It’s very controlled and very organized. So from then on, you have less interaction… we don’t invest in listed companies at the outset. It’s us who put them on the stock exchange. So once they’re listed, a large part of our work is done. And at the point of listing, in general, there are other shareholders who enter the company, so we generally become more of a minority in listed companies than in unlisted companies… it is mainly the handling of the relationship between shareholders and the company and the management which is different. Because you can’t commit insider trading, the same information must be distributed to all shareholders. So as the shareholder-manager relationship becomes codified, it’s less a partnership relationship, it’s less collaborative”.

1. II) The SMEs’ point of view

The company Panosol is not publicly listed. Its manager does not think that Demeter Partners’ intervention in the formation of alliances for their company would differ if either Panosol or an alliance partner were publicly listed. The Comarth company is not listed on the stock exchange either. The point of view of its manager concerning the intervention of Demeter Partners in forming alliances in the presence of publicly listed companies matches that of Panosol. He thus states: “No, there is no difference. The approach is the same”. The same applies for the IES company and the point of view of its manager: “No, no, no, there’s no difference”, he states.

In summary, we can identify that, according to the PEF, its intervention in forming alliances for publicly listed companies is different from unlisted companies. In principle, Demeter Partners intervenes mainly in unlisted companies. If some portfolio companies are listed, it’s the PEF which brought them onto the stock market. Interactions between the PEF and the company it supports strongly decrease once the company is listed on the stock exchange. The company managers interviewed, on the other hand, do not think that Demeter Partners’ intervention would differ if one of the companies in an alliance were listed on the stock exchange.

1. c) Industries et Finances Partenaires

Table 3.35 identifies, again, the points of view of the PEF and the SMEs on the basis of the interview transcripts. The opinions of the participants concerning the impact of the listed/unlisted status of companies for the formation of alliances are contrasted and overlap. The responses are detailed just below the table.

Table 3.35. The Industries et Finances Partenaires field with regard to Hypothesis 6

Publicly listed status of companies	PEF	Caterine Restauration
	Industries et Finances Partenaires	Océane de Restauration	Centrale de Restauration Martel
General opinion	There is no intervention in listed companies.	Neither Océane de Restauration nor any other companies under Caterine Restauration are listed on a market. According to Océane de Restauration’s manager, I&FP’s intervention would have been the same in the case where one or several companies under Caterine Restauration were listed on the stock exchange.	Centrale de Restauration Martel is not publicly listed. According to its manager, the fact that the companies under Caterine Restauration are unlisted and regional actually facilitates synergies, due to the fact that the companies have similar interests.

1. 1) General situation
2. I&FP only intervenes in unlisted companies.

1. 2) Impact of listed/unlisted status of portfolio companies on alliance formation
   1. I) The PEF’s point of view

1. I&FP did not comment regarding this question.
   1. II) The SMEs’ point of view

Neither Océane de Restauration nor Centrale de Restauration Martel are publicly listed. According to the manager of the first company, he does not see why the intervention of I&FP in the formation of alliances for his company would differ if Océane de Restauration or an alliance partner were listed on the stock exchange. He adds, however, that I&FP does not get involved in any case with publicly listed companies. In contrast, the manager of Centrale de Restauration Martel thinks that the fact that all the companies supported by I&FP are unlisted has an impact on the formation of alliances. The reason is that, then, the interests of the companies forming the alliance converge, which is helpful for the alliance. So, he says: “Yes, yes, I think that that played a role. The fact that the companies are unlisted regional teams helped facilitate synergies, that’s clear. Maybe also if there had been one which was publicly listed, we would have already had different interests”. Nevertheless, the manager does not think that that had the impact suggested in the theoretical section of the book, which was to say that if one of the companies had been listed on the stock market, there would have been greater openness of this company’s capital, possibly even several actors in its capital, and consequently less weight for I&FP in decision-making, in particular concerning alliance formation. On this subject, he replied: “No, no, I don’t think so”. The mechanism of causality is thus not confirmed.

In summary, we can identify that the PEF had no comment regarding this question but, in any case, it only intervenes in unlisted companies. As for the company managers interviewed, they do not think that the intervention of I&FP would differ if one of the companies participating in alliances was listed on the stock exchange. In contrast, one of the managers thinks that the fact that all the companies participating in an alliance are unlisted facilitates the alliance, given that they then share the same issues. The mechanism of causality advanced is, however, not confirmed.

1. d) Anonymous PEF

The analytical approach follows the previous cases. The points of view of the PEF and of the SMEs have been analyzed on the basis of interviews and grouped into Table 3.36. Their opinions converge concerning the impact of the listed/unlisted status of companies on the formation of alliances. The responses are detailed just below the table.

Table 3.36. The Anonymous PEF field in relation to Hypothesis 6

Listed status of companies	PEF	Alliance 1	Alliance 2
	Anonymous PEF	YYY	WWW	XXX
General opinion	One of the PEF’s portfolio companies is listed on the stock exchange. This in no way affects Anonymous PEF’s intervention in the formation of alliances for this portfolio company compared to the others.	YYY is not listed on the stock exchange. According to YYY’s manager, Anonymous PEF’s intervention in the formation of alliances would be the same if it were listed.	WWW and its partners in alliances formed under Anonymous PEF are not publicly listed. According to WWW’s manager, Anonymous PEF’s intervention would not differ if one of the companies were listed on the stock exchange.

1. 1) General situation

The large majority of companies supported by Anonymous PEF are unlisted companies. At the time of this study, one of Anonymous PEF’s portfolio companies, the company EEE, was listed on the stock exchange.

1. 2) Impact of listed/unlisted status of portfolio companies on alliance formation
   1. I) The PEF’s point of view

In relation to other companies supported by Anonymous PEF, the fact that EEE is publicly listed has no impact on the intervention of Anonymous PEF in forming alliances for its portfolio companies.

1. II) The SMEs’ point of view

Neither WWW nor YYY are publicly listed. The manager of WWW does not think that Anonymous PEF’s intervention would be different if his company or an alliance partner were listed on the stock exchange. He states as follows: “I do not think that that would change anything at the level of Anonymous PEF. I think that the intervention of the investment manager would be similar. I also think that a fund that was more aggressive in financial terms would behave differently, in fact. It would try to push up the stock price, for example… I can only speak for myself. And for me, whether we were listed or not, it would be the same”. The manager of YYY shares this point of view in the sense that, if his company were listed on the stock exchange, he would have listened to Anonymous PEF’s proposals for alliances in the same way. The intervention of Anonymous PEF in regard to forming alliances for his company would have been the same.

In summary, we can identify that Anonymous PEF mainly supports unlisted companies. One of its portfolio companies is, however, listed on the stock exchange. For the PEF, the fact that one of its portfolio companies is publicly listed changes nothing for its intervention in the formation of alliances for this company compared to others. The portfolio company managers interviewed at Anonymous PEF share this point of view.

1. B) The second level of analysis

After having analyzed field by field the points of view of the PEFs, on the one hand, and the SMEs forming alliances on the other hand, we now compare the summary of each field to those of the others. This summary is presented in Table 3.37.

Table 3.37. Interfield summary in relation to Hypothesis 6

Publicly listed status of companies	Siparex Group	Demeter Partners	Industries et Finances Partenaires	Anonymous PEF
Supports the hypothesis	Yes, but not the mechanism of causation.	Yes, but not the mechanism of causation.	Yes, but not the mechanism of causation.	Yes, but not the mechanism of causation.
Mechanism of causality	For the PEF, it is clear that the role of Siparex Group in the formation of alliances is greater in unlisted companies, given that it does not intervene in listed companies. As for the company managers interviewed, they do not think that Siparex Group’s intervention would differ if one of the companies participating in alliances were listed on the stock exchange.	According to the PEF, its intervention in forming alliances for listed companies is different than for non-listed companies. In principle, Demeter Partners intervenes mainly in unlisted companies. If some portfolio companies are listed, it is the PEF which brought them onto the stock market. Interactions between the PEF and the company it supports strongly decrease once the company is listed on the stock exchange. As for the company managers interviewed, they do not think that the intervention of Demeter Partners would differ if one of the companies participating in alliances were listed on the stock exchange.	The PEF made no comment regarding this question but, in any case, they only intervene in non-listed companies. As for the company managers interviewed, they do not think that the intervention of I&FP would differ if one of the companies participating in alliances were listed on the stock exchange.	For the PEF, the fact that one of its portfolio companies is publicly listed changes nothing for its intervention in the formation of alliances for this company compared to others. The portfolio company managers interviewed at Anonymous PEF share this point of view.

Over all the fields analyzed, it stands out that the role of PEFs in the formation of alliances is in fact greater in the presence of unlisted companies compared to a situation where it is in the presence of companies listed on the stock exchange. However, the mechanism of causality advanced did not find much support. The fact that the PEFs mainly intervene in unlisted companies in the formation of alliances is due to the nature of their activity, meaning that they mainly intervene outside the stock exchange. The mechanism of causality advanced – which is that with the opening of the companies’ capital, the role of the PEF becomes relatively weaker as the other shareholders may also influence the decision-making of the managers of the companies supported - is only confirmed by one PEF (Demeter Partners). For the other fields, the PEF’s intervention should be the same independently of whether the supported company is listed on the stock exchange or not.

The second hypothesis that we raised concerning the relativity of the PEF’s roles in the formation of alliances is as follows:

HYPOTHESIS 10.– The role of PEFs in the formation of alliances is stronger when the latter are formed between companies supported by the same PEF.

As companies financed by private equity suffer from greater informational asymmetry, potential alliance partners may display distrust as to the quality of the company and refuse to enter cooperation. PEFs may contribute to establishing trust between prospective partners by reducing informational asymmetry. In the case where all the prospective alliance partners are financed by the PEF, we may suppose that these potential partners have trust in the information transmitted by the PEF. We thus suppose that the role of PEFs in the formation of alliances is more important when the alliance is formed between companies financed by the same PEF.

1. A) First level of analysis
   1. a) Siparex Group

The points of view of the PEF and of the SMEs have been analyzed on the basis of interviews and grouped into Table 3.38. Their points of view regarding the fact that the roles of PEFs are more pronounced in the case of intraportfolio alliances are contrasted there. The responses are detailed just below the table.

Table 3.38. The Siparex Group field in relation to Hypothesis 10

Intra-/extraportfolio alliances	PEF	Alliance 1	Alliance 2
	Siparex Group	Sophem	Mooviin	Meseo
General opinion	Siparex is involved in the formation of both intra- and extraportfolio alliances. Siparex seeks partners for alliances for the purposes of the alliance, and not based on whether they are internal or external to its investment portfolios.	Through Siparex, Sophem has built more relationships within alliances with other portfolio companies of Siparex than with external companies.	Siparex’s intervention is the same whether in intra- or extraportfolio alliances. There have, however, been more intraportfolio relationships built for Mooviin.	More intraportfolio than extraportfolio relationships have been built.
Intraportfolio alliances	Siparex looks first among its portfolio companies to see if there are appropriate partners for alliances. After that, they will look outside. When Club Siparex puts companies in touch with each other, it may often be an intraportfolio relationship. They may also be extraportfolio. When it is the portfolio manager who builds the relationship, it is slightly more likely to involve extraportfolio alliances.		It is often Club Siparex which takes the initiative with companies.	These are alliances based around e-commerce.
Extraportfolio alliances	The targets come from relevant discussions with the company manager who wishes to form an alliance. They may also come from financial intermediaries who suggest cooperation to Siparex after seeing its portfolio companies online.	No relationships have yet been built with extraportfolio companies.

1. 1) General situation

Siparex Group intervenes as much in forming intraportfolio as extraportfolio alliances for its portfolio companies. The people at Siparex Group who intervene in alliance formation for their portfolio companies are, on the one hand, the investment managers and, on the other hand, people working for Club Siparex.

1. 2) The role of PEFs in the formation of intraportfolio alliances
   1. I) The PEF’s point of view

From the point of view of the investment manager interviewed at Siparex Group, the PEF intervenes as much in the formation of intraportfolio as extraportfolio alliances for its portfolio companies. It cannot be said that one form of alliance prevails over the other. In contrast, the way in which Siparex Group intervenes differs according to whether it is an intraportfolio or extraportfolio alliance. He explains as follows: “Yes, the roles are different, because it’s not addressed and not managed by the same people. On one side, there’s a kind of delegation of people who only do that [people working full-time for Club Siparex]. And, on the other hand, there are the investment managers who do this as part of their overall mission of support”. In summary, Club Siparex may be involved more often in the formation of intraportfolio alliances, while the portfolio manager is more involved in the formation of extraportfolio alliances. Thus, he says: “Two people are employed full-time in this structure [Club Siparex] which is an association and which is chaired by a company leader from one of our portfolio companies. The club has the role of building relationships between our different portfolio companies. It involves the logic of sharing information, sharing experience and business development”.

1. II) The SMEs’ point of view

Up until now, the company Mooviin has built relationships with Siparex portfolio companies for the formation of alliances. There have been no relationships built with companies external to the PEF. However, this is quite conceivable, according to the point of view of Mooviin’s manager. Thus, for him, Siparex Group does not play a more important role in the case of intraportfolio formations. He explains as follows: “It’s the same, it’s the same. After that, for us it’s the same. There are also approaches that are made outside the portfolio. If I want to meet such and such a person, I pick up my phone to meet them. Not everything goes through Siparex. Look, it’s a network like any other”. The Meseo company has also mainly built relationships with other portfolio companies. Unlike the previous company manager, it is clear for Meseo’s manager that the intervention of Siparex Group is greater in the case of intraportfolio alliances. Thus, asked whether Siparex Group plays a greater role in the formation of intraportfolio alliances for his company, he states: “Yes, of course, clearly. These are alliances based around the e-commerce sector, of a customer–supplier type”.

The Sophem company has also been brought into alliances of the intraportfolio type. Its manager thinks, in fact, that the role of Siparex Group is more important for the formation of intraportfolio than extraportfolio alliances, given that the group has more facilities involved in the relationship.

In summary, we can identify that, from the point of view of Siparex Group, the PEF enables its portfolio companies to build relationships within alliances both of the intraportfolio type and the extraportfolio type. We cannot conclude that one type of alliance prevails over the other. In contrast, the PEF’s intervention differs between the two cases, as the alliances are not managed by the same people. Intraportfolio alliances most often come under the work of Club Siparex, while extraportfolio alliances may also be part of the work of the portfolio managers supporting the company forming the alliance. Siparex Group’s portfolio companies who took part in the study had been brought into intraportfolio alliances. For the three managers, this did not at all exclude the PEF from intervening in the formation of extraportfolio alliances. In contrast, two of them thought that the role of Siparex Group was greater in the case of intraportfolio alliances, in particular, because it is easier for Siparex Group to build relationships in that case.

1. b) Demeter Partners

Table 3.39 groups and summarizes the points of view of the PEF and the SMEs regarding whether the PEF’s roles are more important in the case of forming intraportfolio alliances. They more or less overlap. The analysis is supported by the interview transcripts. The responses are detailed just below the table.

Table 3.39. The Demeter Partners field in relation to Hypothesis 10

Intra-/extraportfolio alliances	PEF	Alliance 1		Alliance 2
	Demeter Partners	Comarth	IES	Panosol	Eurener
General opinion	Two-thirds of ideas for alliances for their portfolio companies come from Demeter. There are more extraportfolio than intraportfolio alliances.	Demeter’s role in the formation of alliances is no greater in the case of intraportfolio than extraportfolio alliances. It may be a little different.	Demeter’s intervention is the same and just as important whether it involves the formation of intraportfolio or extraportfolio alliances.	For Panosol, Demeter has played a larger role in the formation of intraportfolio than extraportfolio alliances.
Intraportfolio alliances	Cross-border alliances are generally intraportfolio alliances.
Extraportfoli o alliances	Alliances within France are mainly extraportfolio alliances.

1. 1) General situation

Demeter Partners is involved as much in the formation of intraportfolio as extraportfolio alliances. The most common case, however, is the formation of extraportfolio alliances.

1. 2) The role of PEFs in the formation of intraportfolio alliances
   1. I) The PEF’s point of view

According to the portfolio manager interviewed at Demeter Partners, the two types of alliances, intra- and extraportfolio, go hand in hand with two other types of alliance. Thus, cross-border alliances are mostly of the intraportfolio type. As for alliances between French companies, they are most often extraportfolio alliances, given that there are more possibilities for building relations outside the portfolio than within it. Thus, the portfolio manager interviewed stated: “Within the same country it is easier to build relationships between companies within our investment portfolios with companies which we don’t finance”. This is moreover the most common case. He adds: “Yes, because, of course, we know more companies outside of the portfolio”. In contrast, in the case of cross-border alliances, he states: “When it’s cross-border, it’s quite difficult to do that with companies which are outside the investment portfolios”.

The idea of partners for alliances generally comes, at least in more or less two-thirds of cases, from Demeter Partners. This is the case whether it involves intraportfolio or extraportfolio alliances. Demeter Partners often knows the external companies because, as the portfolio manager interviewed said: “We know them because, as we’re sectoral, we know most of the companies in the sector. And whether we’ve been in contact with them because at one time they wanted to engage in an operation, etc., with us and in the end we didn’t do it, or it’s them who in the end didn’t want to go through with, so they didn’t want to open up their capital and it’s us who wanted to enter into them. And we stayed in contact and we know that they have a good product and they could work with one of our companies”.

1. II) The SMEs’ point of view

The company Comarth has been linked with a company from Demeter Partners’ investment portfolio but not with external companies. According to Comarth’s manager, the idea often comes from the PEF in the case of intraportfolio alliances. In contrast, when it involves building an alliance relationship between his company and a company outside Demeter Partners’ portfolio, the idea is frequently his. Nevertheless, Demeter Partners plays an important role in it. First, indirectly, because the PEF supports and helps its portfolio company simply because it is part of the company’s capital. Second, it plays a direct role because it is sometimes present at the first negotiations and the first exchanges, given that its portfolio company brings the potential alliance partners together in the Demeter Partners premises for these initial discussions so as to gain their support.

The company IES has formed one intraportfolio and one extraportfolio alliance under Demeter Partners’ auspices. For its manager, Demeter Partners’ role in the formation of alliances for his company takes the same importance in the case of intraportfolio and extraportfolio alliances. The intervention only differs on one point: in the case of an intraportfolio alliance, where Demeter Partners is part of the capital of all the companies forming the alliance, the PEF intervenes in the case of an imbalance in the alliance relationship for one of the companies. In contrast, in the case of an extraportfolio alliance, Demeter Partners only intervenes in the case where there is an imbalance within the alliance relationship to the disadvantage of its portfolio company.

The company Panosol has built a relationship with another company within Demeter Partners’ portfolio. It has not formed extraportfolio alliances, though there may have been discussions about this between Panosol’s manager and the investment manager. For its manager, the role of the PEF was greater in the formation of intraportfolio alliances, but that does not exclude it from intervening in the formation of extraportfolio alliances.

In summary, we can identify that, from the point of view of Demeter Partners, the PEF intervenes as much in the formation of alliances of the intraportfolio type as extraportfolio. The most common case, however, is the formation of extraportfolio alliances. According to the PEF, the intervention is not different whether it involves intraportfolio or extraportfolio alliances. In contrast, intraportfolio alliances are often cross-border alliances, while extraportfolio alliances are mostly alliances between French companies. As the PEF is sectoral, it knows the companies in the French sector in which it invests, and there are more opportunities for building relationships outside of rather than inside its investment portfolios. From the point of view of the SMEs, they agree that Demeter Partners intervenes as much in the formation of intraportfolio as extraportfolio alliances. In contrast, according to the company managers interviewed, the PEF’s roles differ between the two cases. The idea for the alliance most often comes from the PEF in the case of intraportfolio alliances.

1. c) Industries et Finances Partenaires

The analysis of the points of view of the PEF and the SMEs with regard to the hypothesis that the PEF’s roles would be greater in the case of formation of intraportfolio alliances are grouped into Table 3.40. They have been analyzed on the basis of the interviews. In general, they overlap. The responses are detailed just below the table.

Table 3.40. The Industries et Finances Partenaires field with regard to Hypothesis 10

Intra-/extraportfolio alliances	PEF	Caterine Restauration
	Industries et Finances Partenaires	Océane de Restauration	Centrale de Restauration Martel
General opinion	The role of I&FP is greater in the case of the formation of intraportfolio alliances.	The role of I&FP is especially high in the realization of synergies between companies within a build-up.	The role of I&FP is especially high in the realization of synergies between companies within a build-up.
Intraportfolio alliances	I&FP is especially involved in the detection and implementation of synergies between companies in a buildup. There may also be relationships built on the margin between companies from different build-ups.	Océane de Restauration was put in contact with a company from another build-up to form an alliance.	There have been no relationships built with companies outside of Caterine Restauration.
Extra-portfolio alliances	That may happen, but it is very rare.	There have been no relationships built with external companies.	There have been no relationships built with external companies.

1. 1) General situation

Alliances are formed mainly between companies within build-ups. Alliances are also possible between companies from different build-ups, although they are much less frequent. In both cases, this involves intraportfolio alliances. Formation of extraportfolio alliances cannot be excluded but are even more extremely rare than alliances between build-ups.

1. 2) The role of PEFs in the formation of intraportfolio alliances
   1. I) The PEF’s point of view

The role of I&FP is greater in the formation of intraportfolio than extraportfolio alliances. The case of formation of an extraportfolio alliance is very rare. I&FP’s core business consists of creating build-ups. Thus, the alliances formed most often are those with the aim of realizing long-term synergies between companies in the same build-up. The associate director interviewed stated as follows: “If we take one of the funds that we’ve managed, fund number one for example where Caterine Restauration was, the alliances or fields of alliances that we have been able to put in place between the various firms, in creating wealth for all the funds, overall the resale price of our portfolio companies is zero. Zero, zero, if you have to measure it. What the presence of I&FP brings to companies is its know-how in the build-up period. And that’s what companies seek because for everything else, they work it out for themselves very, very well”. In the case of forming extraportfolio alliances, he states: “that happens. For example, I’m involved with a factory which makes melted cheese for cooked dishes. In the context of my research into buying other companies in this field, in the field of intermediary food products, well, I came across companies who really had know-how but whose bosses did not want to sell. In that case, it’s a great pity to say: “Good, okay, we can’t do anything on the capital level”, but it’s also a great pity to say: “This company which could possibly allow a commercial relationship to develop, etc.”. Good, so let’s still talk to the managers and let’s ensure that these managers talk to each other. If I as a shareholder see an interest in these companies after all working a bit together, I give them a hint… On the other hand, I’m not really going to spend a lot of time on it”.

1. II) The SMEs’ point of view

The company Centrale de Restauration Martel has not been linked with other companies in alliances through I&FP outside of the Caterine Restauration build-up. According to its manager, the role of I&FP is much greater in the formation of intraportfolio and even intra-build-up alliances than extraportfolio alliances. It is the same for the other companies included in the Caterine Restauration build-up, except for Océane de Restauration.

The company Océane de Restauration has been linked under I&FP not only with other companies involved in Caterine Restauration, but also with a company from another build-up, which is also part of I&FP’s portfolio. However, according to the point of view of the manager interviewed, I&FP building relations between companies outside of the build-up to which they belong is very rare. Thus, he says: “But building relationships in general, outside the build-up, remains very marginal”. We can say that the role of I&FP is more important in the detection of synergies between companies within a build-up. This is their core business. In contrast, outside of build-ups, whether intra- or extraportfolio alliances, it remains very rare. The manager of Océane de Restauration says: “Yes, that’s right. They could do this more. There is another company in their portfolio, a company specializing in the automatic distribution called European Food. We might be able to learn by collaborating with them. It’s true that they don’t present this kind of idea to collaborate with other companies in the fund outside of build-ups, that’s less preferred”. He adds: “… they don’t really promote it. But they’re not against it”.

In summary we can identify that, in the case of I&FP, its role is clearly greater in the formation of intraportfolio than extraportfolio alliances. This point of view is shared by all the people interviewed, from the PEF as well as the companies.

1. d) Anonymous PEF

The points of view of the PEF and the SMEs have been analyzed on the basis of the interview transcripts and grouped into Table 3.41. Their points of view as to whether the PEF’s roles are heightened in the case of the formation of intraportfolio alliances are contrasted there. They more or less converge. The responses are detailed just below the table.

Table 3.41. The Anonymous PEF field in relation to Hypothesis 10

Intra-/extraportfolio alliances	PEF	Alliance 1	Alliance 2
	Anonymous PEF	YYY	WWW	XXX
General opinion	Mainly, alliances are formed within Anonymous PEF’s investment portfolio. But this does not exclude Anonymous PEF intervening in the formation of extraportfolio alliances for its portfolio companies. According to Anonymous PEF, its intervention does not differ between the cases of intra- and extraportfolio alliances.	For YYY, the role of Anonymous PEF is much greater in the case of forming intraportfolio alliances. There has been no formation of extraportfolio alliances for their company.	Anonymous PEF builds relationships for WWW as much with other portfolio companies as with external companies. Anonymous PEF’s roles in building these relationships are important in both cases, but more so in the formation of intraportfolio alliances.
Intraportfolio alliances	The idea mainly comes from Anonymous PEF.		The idea comes as a rule from Anonymous PEF. It is at the initiative of the alliance. It proposes them even before taking the company into its portfolio. After that, it acts as an intermediary.
Extraportfolio alliances	The idea comes either from the portfolio company, from Anonymous PEF or from a combination of the two. Anonymous PEF mainly plays a role in giving credibility to its portfolio companies when the idea of the alliance or of the alliance partner does not come from them. The presence of Anonymous PEF reassures potential partners.	There has been no relationship built between YYY and companies outside Anonymous PEF.	The idea may come equally from WWW or from Anonymous PEF. They discuss it and, afterwards, it is WWW who continues the initiative to meet the potential alliance partner.

1. 1) General situation

Anonymous PEF is mainly involved in the formation of intraportfolio alliances. Extraportfolio alliances are, however, not excluded. There may be alliances between Anonymous PEF’s portfolio companies and one of its shareholders. Anonymous PEF may also be involved in building relationships between one of its portfolio companies and a company which is completely external to the PEF.

1. 2) The role of PEFs in the formation of intraportfolio alliances
   1. I) The PEF’s point of view

The investment manager of Anonymous PEF interviewed does not think that his intervention is different in the case of intraportfolio or extraportfolio alliances. “I don’t see why,” he says.

1. II) The SMEs’ point of view

For the manager of YYY, the role of Anonymous PEF in the formation of alliances for his company is clearly greater in the case of intraportfolio rather than extraportfolio alliances. For the YYY company, there have been no relationships built for forming alliances with companies outside of Anonymous PEF’s investment portfolio. In contrast, it has formed an intraportfolio alliance with the WWW company. The WWW company, in contrast, has formed an intraportfolio alliance under Anonymous PEF and has discussed extraportfolio alliances. In regard to discussions of extraportfolio alliances, the manager of WWW says: “Outside, yes. We reflected on it together, with which companies I might form alliances. We reflected together on the company’s strategy and, in particular, on how to position our targets. As I told you, it’s a partner [Anonymous PEF]”. However, up until now, no concrete contact for forming an extraportfolio alliance has been suggested by Anonymous PEF, which does not mean they did not discuss it. Thus, WWW’s manager adds: “Yes. I’m part of a participatory company management. I’m not all-knowing, so I like to have the opinions of my partners. And I insist, we really are “partners”. So we consult, we exchange”. For the WWW company, the roles played by the PEF slightly differ depending on whether intra- or extraportfolio alliances are being formed. Thus, in the former case, the idea often comes from the PEF. The company manager thus states: “Yes, in fact, they are the ones who know the companies in their portfolio. Often as well, they suggest relationships with other companies with us even before they’ve taken the decision to invest in them. So at that time, we don’t know of each other’s existence. When they make their statement, they’re already thinking of companies they have in their portfolio and wonder whether it might be interesting to introduce so-and-so to so-and-so. So, once they’ve signed and Anonymous PEF invests in the company, they’ll then suggest that the companies who might form an alliance should meet each other. It’s always them who make the first move. It’s them who have the initiative”. In contrast, when it comes to discussing extraportfolio alliances, the idea may come from the PEF or it may come from the company manager, and the PEF will not generally initiate the first meeting, in contrast to the situation with forming intraportfolio alliances. The manager of WWW states as follows: “Yes, when that’s the case, generally, we discuss it and afterwards it’s me who makes the approaches to set up a meeting… they [Anonymous PEF] give their advice, we discuss it and they tell me if they find it interesting or not. They may also tell me that they’ve seen an article in the paper saying that Renault or Peugeot want to integrate eco-materials and composites into their cars, and will then ask me if we’ve thought about it, if there aren’t things we could do together with those companies”.

To the direct question whether the role of the PEF is then more important in the case of forming intraportfolio rather than extraportfolio alliances, the manager interviewed concludes: “Yes, of course. So in that situation, they act as the intermediary”.

In summary, we can identify that, for the PEF, its investment manager does not see why he would intervene differently for the formation of intraportfolio rather than extraportfolio alliances. For the managers of supported SMEs, the PEF intervenes both in the formation of intraportfolio and extraportfolio alliances, but its intervention is greater and more facilitated for intraportfolio alliances. In particular, ideas for alliances most often come from the PEF when intraportfolio alliances are involved.

1. B) The second level of analysis

After having analyzed, field by field, the points of view of the PEF, on the one hand, and the SMEs forming alliances, on the other hand, we now compare the summary obtained for each field to those of the others. This summary is presented in Table 3.42.

The hypothesis according to which the role of PEFs is more important in the formation of intraportfolio alliances is partly validated. Over all the fields analyzed, we see that for some PEFs, it is easier to form alliances between their portfolio companies (for example Anonymous PEF or I&FP). For others, in contrast, it is just as simple outside their investment portfolios as inside them, even easier outside. This is due to the fact that the number of potential alliance partners is often limited internally. We cannot therefore conclude that relationships are built more or less frequently within investment portfolios or outside them. In contrast, where the hypothesis is validated is that the roles played in the formation of intraportfolio alliances differ from those played in the case of forming extraportfolio alliances.

The biggest difference is that, in the case of forming intraportfolio alliances, the idea almost routinely comes from the PEF. In the case of forming an extraportfolio alliance, the idea may come equally from the PEF as from the manager of the supported company. Nor does the PEF always initiate the first contact between prospective alliance partners, contrary to the situation in forming intraportfolio alliances.

Table 3.42. Interfield summary in relation to Hypothesis 10

Intra-/extraportfolio alliances	Siparex Group	Demeter Partners	Industries et Finances Partenaires	Anonymous PEF
Supports the hypothesis	No (especially not the mechanism of causality)	Yes and no	Yes	Yes
Mechanism of causality	The PEF’s intervention differs in the two cases because the alliances are not managed by the same people. Intraportfolio alliances most often come under the work of Club Siparex, while extraportfolio alliances may also be part of the work of the portfolio managers supporting the company forming the alliance. Managers of supported SMEs think that intervention in the case of intraportfolio alliances is nevertheless facilitated for the PEF.	Intervention is different in the two cases. For intraportfolio alliances, the idea comes mostly from the PEF; for extraportfolio alliances, from the PEF and/or from the SME managers.	Alliances are mostly intraportfolio and even intra-build-up.	For the PEF, its investment manager does not see why he would intervene differently for the formation of intraportfolio rather than extraportfolio alliances. For the managers of the supported SMEs, the PEF intervenes as much in the formation of intraportfolio as extraportfolio alliances, but its intervention is stronger and facilitated for intraportfolio alliances. In summary, for intraportfolio alliances, there is more of a chance that the idea of the alliance comes from the PEF.

3.2.2.5.2. Knowledge-based hypotheses

We begin the analysis of the knowledge-based hypotheses by explaining the point of view of the SMEs forming the alliance. We continue with those of the PEF. The approach of the analysis follows the two-level plan presented in Section 3.2.2.5.

The SMEs’ point of view

Adopting a knowledge-based perspective to answer our research question, we have proposed two main paths of intervention for PEFs from the point of view of SMEs. First, there is a possible role for PEFs in detecting opportunities for growth for their portfolio companies and helping the latter to implement them, through collectively learning via the formation of alliances. Apart from that, we considered the role of PEFs in the formation of strategic alliances designed to create new knowledge. Second, once the idea of an alliance has been raised, the next step is approaching the managers of companies who are candidates for forming an alliance. A second hypothesis is also linked to the role of PEFs in the facilitation of the first exchanges.

Detecting opportunities

Concerning the role of PEFs in forming opportunities for growth for their portfolio companies through the formation of alliances, we argued that it is dependent on two factors. The first factor is the expertise of the PEF and the second, its access to information of a strategic nature.

The PEF’s expertise

Regarding the first factor, we suggested the two following subhypotheses:

HYPOTHESIS 7a.– We expect increased formation of intrasectoral alliances in the presence of PEFs which are regional or specialized in financing a specific sector of activity.

HYPOTHESIS 7b.– We expect increased formation of alliances allowing development at an international level of the participating alliances, when the PEF holds investment funds in different countries.

The PEFs detect and allow their portfolio companies to seize opportunities for growth by promoting the growth of strategic alliances. These opportunities and organizational practices depend, in particular, on the competences and expertise of the PEF. We thus posed the hypothesis of a link between the expertise of the PEF and the type of allowances it allows to form (H7). Concretely, we expect increased formation of intraportfolio and intersectoral alliances in the presence of a PEF which focuses their investments in certain sectors (H7a), and increased formation of alliances allowing international development for participating companies when the PEF holds investment funds in different countries (H7b). Let us now check the plausibility of this mechanism of causality across our different cases.

1. A) First level of analysis
   1. a) Siparex Group

The points of view of the PEF and of the SMEs have been analyzed on the basis of the interview transcripts and grouped in Table 3.43. We identify and contrast the points of view of the PEF and the SMEs on the link between the expertise of Siparex Group and the formation of alliances for its portfolio companies. Most often, they coincide. The responses are given in detail after the table.

Table 3.43. The Siparex Group field with regard to Hypotheses 7a and 7b

Expertise	PEF	Alliance 1		Alliance 2
	Siparex Group	Sophem	Mooviin	Meseo
General opinion	Is a facilitator for the formation of alliances of whatever nature. There is no link with Siparex’s expertise. Ideas for alliances result from discussions between Siparex and the managers of its portfolio companies.	Sophem most often comes into contact with companies in the same sector of activity (e-commerce).	There is no link between Siparex’s expertise and the type of alliances that Siparex allows Mooviin to form. What is important is that the alliance makes sense.	There is a link mostly in that Siparex has a regional organization. The alliances formed are regional. Ideas for alliances come as much from Meseo as from Siparex.
Intrasectoral alliances		Formation essentially of intrasectoral alliances which are facilitated due to the fact that Siparex has several holdings in Sophem’s sector of activity.

1. 1) General situation

Siparex Group is a generalist PEF, intervening in different types of sectors. It has a national, even international foothold, with a regional organization.

1. 2) Impact of the PEF’s expertise on the formation of alliances
   1. I) The PEF’s point of view

For the PEF, there is a link between its expertise and the formation of alliances in the sense that Siparex Group is a generalist PEF and it intervenes in the formation of all types of alliances. Intraportfolio alliances form throughout all of the portfolio companies from every sector, and not mostly within a single sector of activity. The associate director interviewed says: “… it’s across all the portfolio companies. It’s the advantage of having a central structure with the two people who work on this full time, and who have an overview of all our portfolio companies”.

Due to its several offices situated outside of France, Siparex Group also intervenes in the formation of alliances at the international level. In this type of alliance, however, the desire or idea for the alliance generally comes from the manager of the supported company. Siparex Group therefore puts forth every effort to aid the manager in their project. The interviewee states: “Yes, we would put them in contact with a specialist, an international expert, with a lawyer specializing in contracts with parties in several countries, etc. If, for example, if there is a company which wants to find a Moroccan supplier in a very specific sector, we will ask our counterparts in Morocco if they have any names or contacts, and if so, we’ll put the companies into contact”.

1. II) The SMEs’ point of view

For the manager of the company Mooviin, Siparex intervenes to build relationships with the company for all types of alliances, as long as they make sense. There is no specific intervention by Siparex Group for a particular type. For the manager of the company Sophem, Siparex Group helps their business in particular to form intrasectoral alliances. This is because this corresponds to its needs, and because Siparex Group has portfolio companies in Sophem’s sector of activity. However, the PEF also invests in other areas and could build relationships between Sophem and companies belonging to other sectors. Currently, this does not correspond to the needs of the company. For the manager of the company Meseo, there is a link between the expertise of Siparex Group and the type of alliances that it helps form. According to him, this expertise is regional.

In summary, we can identify that all of the interviewees concluded that there is a link between the expertise of Siparex Group and the type of alliances that the PEF helps to form. Apart from the manager of the company Meseo, the interviewees agree that the expertise of Siparex Group, which is a generalist PEF, enables the formation of all types of alliances equally, across all sectors.

1. b) Demeter Partners

Table 3.44 presents the points of view of the PEF and the SMEs on the link between Demeter Partners’ expertise and the formation of alliances for its portfolio companies. In general, they are similar. The analysis is based on interviews. The responses are given in detail after the table.

Table 3.44. The Demeter Partners field with regard to Hypotheses 7a and 7b

Expertise	PEF	Alliance 1		Alliance 2
Demeter Partners	Comarth	IES	Panosol	Eurener
General opinion	There is more formation of extraportfolio than intraportfolio alliances because Demeter, as a sectoral organization, is familiar with more companies in the sector than it has in its portfolio. But there is actually a link between Demeter’s expertise and the type of alliances formed because it involves either national or international sectoral alliances. This is linked to the fact that Demeter is a sectoral fund.	There is a link between Demeter’s expertise and the type of alliance formed when the idea for the alliance comes from Demeter.	There is a link between Demeter’s expertise and the type of alliances it enables for IES because Demeter is a sectoral fund.	There is a link between Demeter’s expertise and the type of alliances it enables for Panosol. For a relationship to be built quickly, Demeter must have the company in its portfolio. However, this does not exclude building relationships with external companies.
Intrasectoral alliances	In the case of international alliances, these are generally intraportfolio alliances. Within a single country, it is easier to build relations between portfolio companies and companies which are not supported by Demeter.			This is especially the case for intrasectoral alliances. More recently, a relationship has been built within an intersectoral alliance.
International alliances

1. 1) General situation

Demeter Partners is a sectoral PEF, of national scope or even international in the countries where the PEF is present: in Belgium, Spain, Germany, the United Kingdom and the United States.

1. 2) Impact of the PEF’s expertise on the formation of alliances
   1. I) The PEF’s point of view

According to the portfolio manager interviewed at Demeter Partners, there is indeed a link between the PEF’s expertise and the type of alliances it enables to form. In particular, there is a link with the fact that Demeter Partners is a sectoral PEF; it thus enables frequent formation of sectoral alliances, either national or cross-border, in the countries where it is based. The interviewee explains: “Yes, entirely, yes. And again, we’re a bit particular, because we’re a sectoral fund. So, we can more easily do this kind of thing than a generalist fund. We have companies who are in the same field of business. Except maybe one company we have in our portfolio which makes compatibility software. Because everyone needs compatibility software… But okay, it’s easier when you’re a sectoral fund to do this kind of thing [form alliances for portfolio companies] than when you’re a generalist fund”.

1. II) The SMEs’ point of view

For the manager of the company Comarth, there is a link between Demeter Partners’ expertise and the formation of alliances for his company when the idea for the alliance comes from the PEF. In the opposite case, the PEF may also be involved in alliances which are not directly linked to its expertise. According to the manager of the company IES, an alliance partner of Comarth, there is a link between Demeter Partners’ expertise and the type of alliances it enables to form. So, in response to the question of whether there is a link, he says: “Yes, because Demeter is sectoral. So, it has quite a rapid understanding of the business of its portfolio companies”.

The manager of the company Panosol shares the point of view of a link between the expertise of Demeter Partners and the type of alliances it enables to form. He states as follows: “In an indirect way, the link is in the fact that it’s in their portfolio. We’ve already spoken about other companies. It’s true that to build relations quickly with the president, they must be a shareholder of the company”. In their case, both intrasectoral and intersectoral alliances are formed: “Yes, absolutely, intrasectoral. We’ve recently formed another relationship which was intersectoral”. Concerning the intersectoral alliance, this was formed with a company in a sector of activity in which Panosol was not yet present but wanted to enter. Thus, Panosol’s manager explains: “It was heat pumps, a sector we wanted to enter, concerning which we wanted a relationship. This was done immediately, and there we go. Contacts were made little by little, quietly, and even today when Demeter is no longer a shareholder, we continue the relationship, making exchanges, because there is a trust which has been built”.

In summary, we can identify that all of the interviewees attest to a positive correlation between Demeter Partners’ expertise and the type of alliances that the PEF enables to be formed. It is a sectoral PEF of national or international scope which mainly enables the formation of intrasectoral or intersectoral alliances. It also enables the formation of both national and cross-border alliances, mainly between companies in countries where it is present.

1. c) Industries et Finances Partenaires

As for the previous fields, the points of view of the PEF and the SMEs have been analyzed on the basis of the interview transcripts and grouped in Table 3.45. Generally, their points of view on the link between I&FP’s expertise and the formation of alliances for its portfolio companies converge. The responses are given in detail just below the table.

Table 3.45. The I&FP field with regard to Hypotheses 7a and 7b

Expertise	PEF	Caterine Restauration
	Industries et Finances Partenaires	Océane de Restauration	Centrale de Restauration Martel
General opinion	Yes, there is a link between I&FP’s expertise and the types of alliances and build-ups, because I&FP’s areas of competence are also linked to their experience and understanding of certain sectors/markets. Ideas also come more easily in these sectors.	This mainly involves relationships built within a build-up, thus relationships built within the same sector, “intrasectoral”. As to relationships built outside of a build-up, I&FP is not opposed but does not specifically promote them. This remains very rare.	I&FP’s expertise is that it has been able to detect synergies which can be realized by the companies under the Caterine Restauration build-up and thus for CRM.
Intrasectoral alliances	This is based on synergies between companies within a build-up and therefore in the same sector. There may be synergies between companies from different build-ups and thus intersectoral synergies, but this is very rare.
International alliances

1. 1) General situation

I&FP specializes in the formation of build-ups, so sectoral consolidation at the national level. The PEF is particularly competent in certain sectors such as distribution, agro-food or construction and public works.

1. 2) Impact of the PEF’s expertise on the formation of alliances
   1. I) The PEF’s point of view

According to the associate director interviewed at I&FP, there is clearly a link between I&FP’s expertise and the formation of alliances for its portfolio companies. Thus, he says: “Well yes, in the sense that we carry out quite a few operations in these sectors, for example distribution… Well, that will lead to us being a bit better in alliances, in build-ups in those areas. Because we understand how that works. Same thing in agro-food, same thing in construction and public works. People like Cerea will answer you, for them it’s obvious because they do nothing but agro-food. Yes, very clearly, that plays an enormous role. But for us, in fact…”. Responding to whether ideas for alliances come to him more easily in some sectors, he replies: “More easily, yes. Certainly”. His response to whether we should see, consequently, more alliances in and within certain types of sectors, is also affirmative.

1. II) The SMEs’ point of view

For the manager of the company Océane de Restauration, there is clearly a link between I&FP’s expertise and the type of alliances formed. It essentially involves intrasectoral alliances, given that the PEF specializes in sectoral consolidation. He thus explains: “Yes, these are consolidations of companies belonging to the same sector. So, yes, the fund mainly intervenes to build relationships between companies belonging to the same sector”. The manager of the company Centrale de Restauration Martel does not see this in the same way. He thus explains: “No, not necessarily. No, because I&FP did not know the business. The synergies happened more in relation to immediate needs and not in regard to a specific competence. It’s not really a build-up that they mastered at the level of a core business. So that happened more on the level of actions for synergies, of reducing costs than by applying competences”. In contrast, he does admit to having been introduced to companies in the same sector of activity, and ideas for synergies between companies also came from the PEF. Their implementation, however, resulted from collegial discussions between the PEF and the members of the build-up to which the company belongs.

In conclusion, we can identify that there is a link between I&FP’s expertise and the type of alliances that the PEF enables to be formed. Its field of expertise is sectoral consolidation at the national level. Alliances are thus made mainly between companies participating in sectoral consolidation. They are thus for the most part intrasectoral alliances on a national scale.

1. d) Anonymous PEF

Based on the interview transcripts, we analyze the points of view of the PEF and the SMEs on the link between the expertise of Anonymous PEF and the formation of alliances for its portfolio companies. They are grouped in Table 3.46. In general, they more or less overlap. The responses are given in detail after the table.

Table 3.46. The Anonymous PEF field with regard to Hypotheses 7a and 7b

Expertise	PEF	Alliance 1	Alliance 2
	Anonymous PEF	YYY	WWW	XXX
General opinion	Has many ideas for alliances because Anonymous PEF invests in closely related fields of activity and develops over a restricted territory Based on this, the alliances which are formed are mainly intrasectoral (understood to be in the field of materials) or regional. There is therefore indeed a link between Anonymous PEF’s expertise and the type of alliances it enables its portfolio companies to form.	Yes, there is a link between Anonymous PEF’s expertise and the type of alliances it has enabled YYY to form. That depends on the companies in the portfolio, which depend on the investment criteria of Anonymous PEF. The idea then comes from Anonymous PEF who initiates the alliance.	Difficult to say, not convinced of a link between Anonymous PEF and the type of alliance it enables to be formed for its portfolio companies including WWW. It is not important for Anonymous PEF to respond to all kinds of requests from its portfolio companies.

1. 1) General situation

Anonymous PEF is a regional and sectoral PEF. It intervenes in Lorraine and in the area of materials.

1. 2) Impact of the PEF’s expertise on the formation of alliances
   1. I) The PEF’s point of view

According to the investment manager interviewed at Anonymous PEF, there is a link between the PEF’s expertise and the type of alliances it enables its portfolio companies to form. “Yes, indeed”, he says. It mainly involves regional and intrasectoral alliances. More generally, he says: “I think we are more likely to generate alliances because we are developing over a smaller territory, in related industrial areas. So, we’ll have more ideas, more opportunities and probability in forming alliances than funds which are very generalist and international, for example”.

1. II) The SMEs’ point of view

The manager of the company WWW is not convinced of a link between the expertise of Anonymous PEF and the type of alliances that the PEF enables to be formed. He agrees that Anonymous PEF has a certain amount of skill in the formation of several types of alliance (intersectoral, for example). In contrast, Anonymous PEF would make every effort to support and help its portfolio company in the formation of any other type of alliance, if it were desired and if it made sense at the level of the strategy being pursued. According to the manager of the company YYY, there is a link between Anonymous PEF’s expertise and the type of alliances it enables to be formed for its portfolio companies. He justifies it like this: “Yes. Yes, because Anonymous PEF has diverse portfolio companies and, for now, there are only two which could form a relationship with us”.

In summary, we can identify that there is a link between Anonymous PEF’s expertise and the type of alliances that the PEF enables to be formed. The PEF is sectoral and regional and intervenes mainly and with greater ease in forming intrasectoral or intersectoral and regional alliances. This however does not prevent the PEF from also supporting its portfolio companies in forming other types of alliances.

1. B) The second level of analysis

After having analyzed, field by field, the points of view of the PEFs, on the one hand, and the SMEs forming alliances, on the other hand, we now contrast the summary obtained for each field with the others (Table 3.47).

Table 3.47. Interfield summary with regard to Hypotheses 7a and 7b

Expertise	Siparex Group	Demeter Partners	Industries et Finances Partenaires	Anonymous PEF
Supports the hypothesis	Yes	Yes	Yes	Yes
Mechanism of causality	Generalist PEF of national and international scope. It enables the formation of all kinds of alliances, across all sectors, even cross-border.	Sectoral PEF of national, even international scope. It mainly enables the formation of intrasectoral, but also intersectoral alliances. It also enables the formation of both national and cross-border alliances, mainly between companies in countries where it is present.	The PEF’s field of expertise is sectoral consolidation at the national level. Alliances are thus made mainly between companies participating in sectoral consolidation. They are therefore mainly intrasectoral alliances, located at the national level.	Sectoral and regional PEF. It intervenes mainly and with ease in the formation of intra- or intersectoral and regional alliances.

The general hypothesis of a link between the PEF’s expertise and the type of alliances formed finds support across the four fields of analysis. The sectoral and/or regional PEFs intervene mainly in the formation of intra- or intersectoral alliances and/or those of regional scope. As for the generalist type of PEF, it intervenes in the formation of alliances across all types of sectors. PEFs of national scope have more of a tendency to enable the formation of alliances at national level, while PEFs with offices overseas more frequently enable the formation of cross-border alliances between their portfolio companies. These facts do not prevent PEFs from also intervening in the formation of alliances beyond their field of expertise. Hypotheses 7a and 7b, assuming greater formation of intrasectoral or intersectoral alliances in the case of a sectoral PEF and greater formation of cross-border alliances in the presence of a PEF of international scope, are thus confirmed.

The PEF’s access to information of a strategic nature

The second factor which determines the PEF’s detection of growth opportunities for its portfolio companies is its access to information of a strategic nature. On this topic, we raised the following two subhypotheses:

HYPOTHESIS 11a.– The number of companies in which the PEF holds one (or several) seat(s) on the board of directors or the supervisory board has a positive effect on the formation of alliances between the companies it supports.

HYPOTHESIS 11b.– Participation by the PEF in associations (AFIC/France Invest, EVCA/Invest Europe, Unicer, etc.) has a positive impact on the formation of alliances between the companies it supports.

We assume a positive impact of the PEF having access to information of a strategic nature on the formation of alliances (H11). PEFs usually sit on strategic boards, on the board of directors or the supervisory board of companies they support. These boards represent places for discussion and thus allow access to information of a strategic nature, which may be a source of ideas. We thus expect that the number of companies in which the PEF holds one (or several) seat(s) on the board of directors or the supervisory board will have a positive effect on the formation of alliances between the companies it supports (H11a). Similarly, the PEF participating in associations which discuss specific themes linked with the sectors of activity of their supported companies, or associations such as AFIC/Invest France, EVCA/Invest Europe or Unicer, may also be sources of information and ideas. We therefore expect that the participation of the PEF in this kind of event would have a positive effect on the formation of alliances for the companies they support (H11b). We now test the hypotheses and the suggested mechanism of causality.

1. A) First level of analysis
   1. a) Siparex Group

Table 3.48 presents a summary of the points of view of the PEF and the SMEs on the link between Siparex Group’s access to information of a strategic nature and alliance formation for its portfolio companies. Generally, they overlap closely. The responses are given in detail just below the table.

Table 3.48. The Siparex Group field with regard to Hypotheses 11a and 11b

Seat(s) on strategic boards/associations	PEF	Alliance 1	Alliance 2
	Siparex Group	Sophem	Mooviin	Meseo
General opinion			For Mooviin, the formation of alliances is almost secondary.
Strategic boards	Yes, seats on strategic boards have a positive influence on forming alliances. These boards are where strategic decisions are made; having a seat there therefore allows access to information which may be a source of ideas for alliances.	Yes, the seats held by Siparex on strategic boards may be a source of ideas for alliances.	The seats held by Siparex on strategic boards may be a small benefit to the formation of alliances for Mooviin. It’s a source of information, a sign of a guarantee of quality.	Yes, the seats held by Siparex on strategic boards definitely have an impact on the formation of alliances.
Associations	Siparex is present in associations but that has no impact on the formation of alliances. It is more about working groups on the issues for capital investors, not involving their portfolio companies.	Do not know if Siparex is part of associations. Internally, it formed Club Siparex with a working group on e-commerce. In any case, its investment manager is very well informed on this subject, and this clearly has a positive impact for Sophem with regard to the formation of alliances.	Yes, people from Siparex are involved in associations, forums, etc., and that may have a favorable effect on the formation of alliances, or at least, it has no negative effect.	Suppose that Siparex is part of associations, but don’t know.

1. 1) General situation

Siparex Group holds seats on strategic boards (board of directors, supervisory board, etc.) in the companies it supports. The PEF is also involved in associations such as AFIC/Invest France, Unicer, AFG (French Management Association).

1. 2) Impact of specific sources of information on the formation of alliances
   1. I) The PEF’s point of view

For the portfolio manager interviewed at Siparex Group, the number of seats held on strategic boards, of whatever kind, is clearly a source of ideas and beneficial for the formation of alliances for its portfolio companies. He explains: “Having a seat on a company’s governing board, of whatever sort, so board of directors, supervisory board, strategic board, as the name suggests these are governing boards, these are places where decisions are made, these are places for discussion and work whether about strategic alliances or otherwise. So, yes, the fact of having a seat or an administrative post increases the chances of forming alliances”.

The involvement of Siparex Group in associations like those cited above has no impact on the formation of alliances for its portfolio companies, states the interviewee. Responding to whether it may be a source of ideas for alliances, he responds: “No. It’s more working groups on issues for capital investors, rather than issues for the companies in which we invest”.

1. II) The SMEs’ point of view

For the manager of the company Sophem, the fact that Siparex Group sits on strategic boards clearly gives it access to information which may be a source of ideas for alliances. In contrast, she cannot say if Siparex Group is involved in associations which debate the key themes linked to her sector of activity. She imagines so. The investment manager in charge of supporting her company has various kinds of information which he must obtain from one source or another. This information is beneficial to Sophem for forming alliances. The manager of Meseo shares these opinions. For the manager of the company Mooviin, it is clear that holding seats on boards gives access to information and may have an impact on his company forming alliances. The same goes for the PEF’s participation in associations which debate key themes related to its portfolio companies’ sector of activity. For the manager of Mooviin, it is clear that Siparex Group attends events that play a role - at least, not a negative one - in the formation of alliances.

In summary, we can identify that the PEF holding seats on strategic boards may be a source of ideas and favorable to the formation of alliances for the companies they support. As for the involvement of the PEF in associations, the perceived link is weaker or even absent. The PEF says that it does not participate in associations which give it access to information which may be favorable to the formation of alliances. The managers of the SMEs - except for one - did not have an opinion. The one who did have an opinion perceived a link which was at least not negative.

1. b) Demeter Partners

The analytical approach is the same as for the previous field . The points of view of the PEF and the SMEs on the link between Demeter Partners’ access to information of a strategic nature and the formation of alliances for its portfolio companies have been grouped in Table 3.49. In general, they overlap. The responses are given in detail just below the table.

Table 3.49. The Demeter Partners field with regard to Hypotheses 11a and 11b

Seat(s) on strategic boards/associations	PEF	Alliance 1		Alliance 2
	Demeter Partners	Comarth	IES	Panosol	Eurener
General opinion	Demeter Partners holds seats as a rule on the strategic boards of the companies it supports. It is difficult for them, therefore, to compare if this facilitates access to certain information in relation to a situation where it did not sit on these boards. Demeter supposes that this facilitates access to some information, and that may be favorable to ideas for alliances.
Strategic boards		The fact that Demeter sits on the board of directors has a positive impact on the formation of alliances for Comarth, because absolutely everything is discussed on this board.	Seats held on strategic boards will give Demeter ideas for forming alliances.	Yes, the seats held by Demeter on strategic boards may have a favorable impact on the formation of alliances for their portfolio companies. This also gives credibility in the sense that for Demeter to propose a company as an alliance partner to another company is a sign of quality if it has been chosen among several candidates.
Associations	Demeter Partners participates in several associations and Competitiveness Clusters. This may have an impact on the formation of alliances and ideas for alliances, especially extraportfolio ones, and thus with alliance partners which are external to Demeter’s portfolios of investment.	Demeter as well as Comarth attend events, forums etc., and this may be a source of ideas for alliances which Demeter and Comarth may exchange.	Demeter attends events and forums and suggests to IES and its other portfolio companies that they attend such events. It is a source of ideas and contacts which may lead to alliances.	Relatively little presence of Demeter in associations or forums which relate to the business of Panosol because it involves B2B. This fact is important to Panosol’s investors, but not so much for its business and the alliances that Panosol might form.

1. 1) General situation

Demeter Partners holds seats as a rule on the strategic boards of the companies it supports. In addition, Demeter Partners takes part in associations such as AFIC/Invest France, EVCA/Invest Europe, Unicer or Cleantech Group. Demeter Partners is also a member of the Renewable Energy Professional Organization and is involved in several Competitiveness Clusters.

1. 2) Impact of the PEF’s specific sources of information on the formation of alliances
   1. I) The PEF’s point of view

Demeter Partners takes seats on strategic boards as a rule. It is thus difficult for it to imagine a situation where it did not do so. But the interviewee imagines that this facilitates access to information. He thus explains: “As there is no company for which we are not on the board, I would not really be able to tell you what the difference is. But I imagine that if we weren’t on the board, in certain companies it would be less easy to have ideas for building relationships”. With regard to the involvement of Demeter Partners in associations, the PEF is indeed involved, in particular, in those cited above. They are indeed a source of information which might give ideas for forming alliances for the companies in which Demeter Partners has a shareholding. The investment manager interviewed states as follows in response to whether the fact that Demeter Partners belongs to associations has an impact on the formation of alliances: “Yes. Especially with companies which are outside our portfolios. Finally, basically, for us, these are sources of “deal flow”, of prospects. And therefore, this allows us to get to know the companies in the sector better, and possibly build relationships between our portfolio companies and external companies”. Their response to whether it is a source of ideas and contacts for alliances is affirmative.

1. II) The SMEs’ point of view

For the manager of the Comarth company, it is certain that the fact that Demeter Partners holds seats on the board of directors or supervisory board of the companies it supports gives access to information of a strategic nature that may be a source of ideas for alliances. It also happens that the portfolio manager shares ideas which may help the formation of alliances after attending conferences, associations, forums or other events. The manager of the IES company shares this point of view. Concerning the importance of attending events and forums, he states: “We’re really in a market where we have to follow the market and be present. Demeter, as a sectoral fund, is well known and they invite us to events. And it’s really fundamental for finding clients to attend these events… It’s fundamental to find contacts and clients. It’s at cocktail parties, salons etc., where you meet people”. The manager of the company Panosol shares the point of view of the other managers concerning the impact of the PEF holding seats on strategic boards on the formation of alliances. In contrast, as to whether there is interest in the PEF’s investment manager attending events or forums which discuss certain topics for the activity of forming alliances for his company, he states: “For the business we’re involved in, very little, because it still involves B2B business. Even if it might be interesting for people, in practice it’s rare. Also, there’s interest, it’s enjoyable, it builds reputation for managers. It’s in our interest if we ever needed Demeter to exit and another fund to enter. It holds more interest for our investor and not much for business benefits”.

In summary, we can identify that all of the people interviewed testified to a positive link, both for holding seats on strategic boards and for the PEF being part of associations, going to forums, salons etc., on the formation of alliances for its portfolio companies. Both give access to information which may be the source of ideas for alliances. Only one of the company managers testified to a weak link to the formation of alliances for the PEF going to events. This, however, seems linked to his sector of activity, as he states.

1. c) Industries et Finances Partenaires

Again, the points of view of the PEF and the SMEs have been analyzed on the basis of the interview transcripts and grouped into Table 3.50. We identify and contrast the points of view of the PEF and the SMEs on the link between I&FP’s access to information of a strategic nature and the formation of alliances for its portfolio companies. In general, they overlap. The responses are given in detail after the table.

Table 3.50. The Industries et Finances Partenaires field with regard to Hypotheses 11a and 11b

Seat(s) on strategic boards/associations	PEF	Caterine Restauration
	Industries et Finances Partenaires	Océane de Restauration	Centrale de Restauration Martel
General opinion	The number of seats held by I&FP on boards has no impact on the formation of alliances, given that I&FP in any case holds the majority of shares for the companies it supports. Most of the value added by I&FP does not happen on boards, but in day-to-day relations with the managers of the companies it supports.
Strategic boards	What is important is the almost daily dialog with company managers, so the relationship between I&FP and the managers.	The number of seats held on strategic boards by I&FP is important in the sense that the more different companies they hold seats in, the more issues they see, which may be a source of ideas but not necessarily for the formation of alliances.	The number of seats held on strategic boards by I&FP has no impact on the formation of alliances.
Associations	I&FP is part of associations such as AFIC/Invest France or EVCA/Invest Europe and ensures that the companies it supports participate in associations or events which concern them. It is important as a source of information, as well as for ideas for alliances.	If I&FP participates in associations, this may constitute a source of ideas but the activity of forming alliances is really marginal for I&FP. What might be more important is the providers of business with whom I&FP engages, and the company managers who approach them to sell their companies.	I&FP has not participated in associations.

1. 1) General situation

I&FP takes majority stakes in the companies it supports. This means that the PEF also holds a majority on its boards. I&FP is part of associations such as AFIC/Invest France or EVCA/Invest Europe.

1. 2) Impact of the PEF’s specific sources of information on the formation of alliances
   1. I) The PEF’s point of view

According to the PEF’s point of view, given that it takes majority stakes, the fact of holding seats on strategic boards does not enable important access to information that can be the source of ideas for alliances. Thus, the associate director interviewed says: “No, not at all. No, we’re in a majority everywhere. Actually yes, there is one. The corollary of being the majority shareholder is that we’re a majority on the boards as well. But still, the essential part of the value added by I&FP does not take place on the boards, but rather in the day-to-day relationship we have with the managers and the dialog we have on a daily basis. You don’t make decisions on the boards, in fact”. He continues: “it’s rather linked to the fact that we’re very close to our companies and their managers which means we have a dialog, not daily, but quite close to it”.

Concerning participation in forums and events, it’s not I&FP who participates directly. The PEF ensures that the companies it supports attend. Thus, the interviewee says: “We’re a member of AFIC/Invest France and EVCA/Invest Europe but we have no activity, we don’t actually participate. Still, while it’s clear that concerning our portfolio companies and others, I&FP is not directly a member of any sort of professional group or association, in contrast we make sure that the companies in which we invest do so, when they exist in their profession”. Responding to the question of whether attending such events has an impact on the formation of alliances for their supported companies, the associate director states: “Of course. For example, in mass catering, Océane de Restauration was a member of a professional group, SNERS – well, in fact, being there allowed me to spread the word about I&FP as a shareholder in Océane Restauration and to clearly send a message to the members of SNERS saying: ‘If you’re interested in joining the adventure of I&FP and Caterine Restauration, let’s talk’. And often, the build-up brought to Océane de Restauration something that SNERS had never managed. Namely, discussions about how to improve production efficiency, etc. And it’s really what I was saying to you. I think that there’s not a lot of efficiency. Associations are fine, but if you don’t have at any given moment a central point where people’s interests are the same, you won’t get there. And one central point is to be a shareholder”.

1. II) The SMEs’ point of view

According to Océane de Restauration’s manager, the more seats the PEF has on strategic boards, the more it can identify issues. This may be a source of ideas, but not for the formation of alliances. He does not know if I&FP takes part in associations. When asked if the PEF’s participation in associations may be a source of ideas for alliances, he replies: “Maybe. But as I said, this relationship-building is already very marginal. I think that what is a source of ideas is rather the providers of business whom they engage and the offers of sale from managers who want to sell their company”.

The manager of Centrale de Restauration Martel entirely agrees with the above point of view. He thinks that neither the seats held on strategic boards by I&FP nor participation in specific events or forums plays a role in the formation of alliances for the companies which I&FP supports.

In summary, we can identify that in the case of I&FP, holding seats on strategic boards of the companies it supports is not a primary source of information of a strategic nature which may give ideas for the formation of alliances. The same goes for the participation of the PEF in forums or events. Moreover, the PEF does not participate much in such things. Rather, it advises its supported companies to attend them. This is explained by the context. I&FP specializes in sectoral consolidation. To do this, it takes majority stakes in the companies it buys through a holding company created for this purpose. Holding majority stakes, access to information is not provided through seats on boards.

1. d) Anonymous PEF

On the basis of the interview transcripts, we have analyzed the points of view of the PEF and the SMEs on the link between Anonymous PEF’s access to information of a strategic nature and the formation of alliances for its portfolio companies. Most often, they only partly converge. A summary can be seen in Table 3.51. The responses are given in detail after the table.

Table 3.51. The Anonymous PEF field with regard to Hypotheses 11a and 11b

Seat(s) on strategic boards/associations	PEF	Alliance 1	Alliance 2
	Anonymous PEF	YYY	WWW	XXX
General opinion		Ideas for alliances may especially come from the files which Anonymous PEF analyzes, even if they choose not to support that company.
Strategic boards	The number of seats held on strategic boards has an impact on the formation of alliances; the more Anonymous PEF is present on company boards, the more ideas for alliances it may have.	The number of seats held by Anonymous PEF on strategic boards has a positive impact on the formation of alliances. The larger the number of companies, the larger the number of possible synergies in the portfolio.	The number of seats held by Anonymous PEF on strategic boards has no direct impact on the formation of alliances. Ideas for alliances come more often to Anonymous PEF because of its economic intelligence in the broad sense (watching the markets, etc.), from going to events, forums, and conferences, and from the files that it receives.
Associations	Anonymous PEF participates in associations, events or forums and that constitutes a source of ideas for alliances. This mainly allows it to guide strategic discussions that can lead to discussions of alliances.	Do not know if Anonymous PEF is part of associations or if that may have an impact.	Yes, Anonymous PEF participates in events, forums and conferences and this has an impact on ideas for forming alliances. If Anonymous PEF is part of associations like AFIC/France Invest or EVCA/Invest Europe, WWW does not know.

1. 1) General situation

Anonymous PEF generally holds seats on the strategic boards of the companies it supports. The PEF is also a member of Unicer, AFIC/France Invest and Green Universe. It is involved in Competitiveness Clusters.

1. 2) Impact of the PEF’s specific sources of information on the formation of alliances
   1. I) The PEF’s point of view

According to Anonymous PEF’s investment manager, sitting on strategic boards has an impact on the formation of alliances for its portfolio companies. “The more we follow companies, the more we can get ideas and examples of useful alliances”, he states. He adds that this has certainly played a role in the alliances that Anonymous PEF has already been able to form: “Yes, that has played a role. Not being on the board means having no information on the company, or in any case very vague information. So, it would be difficult to suggest and guide alliances”.

For Anonymous PEF’s investment manager, participation in associations is also a source of ideas. He says: “Ideas, yes. It’s mainly, for example, we are members of an association called Green Universe. I don’t know if we’re members; we attend all their events. These are people who give their opinion on what is happening in the environment and we follow their conferences, because that gives us, in fact, information on market trends which we, in turn, pass on to our portfolio companies”. In reality, this has not yet had a direct impact on an alliance. In contrast, as the interviewee states: “Not in reality on an alliance, but it enables us to orient strategic discussions with the companies”.

1. II) The SMEs’ point of view

For the manager of the company WWW, the fact that Anonymous PEF holds seats on strategic boards has no significant impact on the formation of alliances for the companies it supports. In contrast, Anonymous PEF may go to conferences which may be sources of ideas of alliances. Thus, in answer to whether the PEF holding seats on boards may have an impact on the formation of alliances, he explains: “No. I think the inspiration that it gets for building relationships comes more from its technical intelligence and its economic intelligence in the broad sense (watching the markets, etc.), from its overall intelligence. And when I say intelligence, it means that it’s them who will make the effort to go to conferences, go to events, etc. Then, another source, in my sense, there are people who go to ask them to intervene in their company. So, they will inspect the files, study them, etc. And in that case, this may also be a source of ideas… I know that they go to congresses on biomaterials, on the more technical points”.

For the manager of the YYY company, the fact that Anonymous PEF holds seats on strategic boards is a source of ideas and has a positive impact on the formation of alliances. He says: “It’s positive. It’s certain that if it finds synergies in one way or another, it has every interest in building relationships between the companies in which it has a foothold. So, the larger the number of companies, the larger for the number of relationships in the portfolio”. Concerning the impact of Anonymous PEF going to events or forums on the formation of alliances, he cannot comment.

In summary, we can identify that the number of seats held on strategic boards by the PEF, like its participation in forums, events and associations, may be sources of information and give ideas for alliances. Opinions differ, however, concerning the relative importance of these sources of information. According to one of the managers interviewed, another source of information may be more important than the two mentioned: the files that pass over the PEF’s desk, which it analyzes before making a decision to invest.

1. B) The second level of analysis

After having analyzed, field by field, the points of view of the PEFs on the one hand, and the SMEs forming alliances on the other hand, we now contrast the summary obtained for each field with the others (Table 3.52).

Table 3.52. Interfield summary with regard to Hypotheses 11a and 11b

Seat(s) on strategic boards/associations	Siparex Group	Demeter Partners	Industries et Finances Partenaires	Anonymous PEF
	Yes	Yes	No	Yes
Mechanism of causality	Holding seats on strategic boards on the part of the PEF may be a source of ideas and favorable to the formation of alliances for the companies it supports. Concerning the PEF’s involvement in associations, the PEF is not a member of any associations which might have a favorable impact on the formation of alliances.	Holding seats on strategic boards on the part of the PEF gives access to information and is a source of ideas for the formation of alliances for the companies it supports. The same goes for the PEF’s involvement in associations.	Neither seats held on the strategic boards of the companies it supports, nor participation in forums or associations, constitute important sources of access to information favorable to the formation of alliances. This is due to the context, the fact that I&FP organizes build-ups and takes majority stakes.	The number of seats held on strategic boards by the PEF, as well as its participation in forums, events and associations, may be a source of information and give ideas for alliances. Opinions differ, however, as to the relative importance of these sources of information. Another source may be more important: the files on SMEs that are candidates for financing which are analyzed by the PEF.

The hypothesis that the PEF having access to information of a strategic nature has a positive impact on the formation of alliances for its portfolio companies is supported, in all fields except for I&FP. The number of seats held on strategic boards (board of directors or supervisory board) seems to have a greater positive impact on the formation of alliances than involvement of the PEFs in associations. This latter source of ideas only has an impact if the association or event deals with themes relating to the sectors of activity of the supported companies.

Other sources, not mentioned, may also have an impact on the formation of alliances, and possibly to a greater extent. These may include, for example, files on firms which are candidates for financing which are analyzed by the PEF’s investment committees.

The I&FP field does not confirm the hypothesis. This is due to the context. Although the number of seats held on strategic boards and attending certain events may be sources of information and may give ideas for alliances, I&FP obtains such information and ideas in a more direct manner due to its taking majority stakes.

Facilitation of exchanges

Once potential partners for an alliance are found, the next step is to approach the managers of the businesses taking part. We have argued that the presence of a PEF allows the facilitation of exchanges between potential alliance partners who are supported by this PEF and, thus, has a positive impact on the formation of the alliance (Hypothesis 8). However, this intervention as a facilitator of exchanges assumes that the PEF has the time to devote to such a support service which does not constitute its primary task. Consequently, we suppose a negative link between the number of portfolio companies to be managed by each investment manager (because of lack of time) and the formation of alliances. Similarly, we suppose a negative effect associated with the geographical distance which separates the portfolio companies from the investment manager in charge of their file (H8a).

HYPOTHESIS 8a.– (1) A high number of portfolio companies to be managed by an investment manager (lack of time) and (2) a large geographical distance separating the portfolio companies from the investment manager in charge of their file, have a negative impact on the formation of alliances.

A purely knowledge-based argument, however, allows us also to consider a positive impact associated with the number of portfolio companies to be managed by each investment manager and the formation of alliances. The more the latter examines the files, the more they may be a source of ideas. We will therefore also test the following impact:

HYPOTHESIS 8a (knowledge-based argument).– (1) A high number of participations to be managed by an investment manager (source of ideas) has a positive impact on the formation of alliances.

To test if, in contrast, the fact of having relatively few portfolio companies to manage and a small geographical distance has the inverse effect and is favorable to the PEF becoming more involved in the formation of alliances for its portfolio companies, we also pose the following hypothesis:

HYPOTHESIS 8b.– (1) A small number of portfolio companies to be managed by an investment manager and (2) a small geographical distance separating the portfolio companies from the investment manager in charge of their file, have a positive impact on the formation of alliances.

Let us now put these hypotheses and their mechanism of causality to the test.

1. A) First level of analysis
   1. a) Siparex Group

The points of view of the PEF and the SMEs have been analyzed on the basis of the interview transcripts and grouped into Table 3.53. We identify and contrast the points of view of the PEF and of the SMEs on the link between the number of portfolio companies to be managed by each investment manager at Siparex Group in the formation of alliances for its portfolio companies, and then on the link between the geographical distance that separates the PEF from the SMEs and the formation of alliances. They more or less overlap. The responses are given in detail just below the table.

Table 3.53. The Siparex Group field with regard to Hypotheses 8a and 8b

Facilitation of exchanges	PEF	Alliance 1	Alliance 2
	Siparex Group	Sophem	Mooviin	Meseo
General opinion	Siparex Group does intervene to facilitate exchanges between alliance partners.	Siparex Group does intervene to facilitate exchanges between alliance partners.		Siparex Group does intervene to facilitate exchanges between alliance partners.
Number of portfolio companies per investment manager	The investment managers in charge of the SMEs manage approximately a dozen portfolio companies. Those in charge of mid-sized companies manage two or three of them. The number of portfolio companies to be managed has an impact on the formation of alliances. If there are too many to manage, the person has no time to correctly support the company. However, this does not affect the activity of forming alliances, because he will then do overtime. In addition, the number of companies to be managed may also be a source of ideas for alliances. This latter effect outweighs the first.	The more files to be managed by an investment manager, the less time they will have to devote to each portfolio company, and thus also to forming alliances.	The number of portfolio companies has a positive impact on the formation of alliances. The more there are, the more they see issues and strategies, which provides ideas and experiences which may be useful for other portfolio companies.	The more portfolio companies, the more chances to form alliances.
Geographic distance	Yes, geographical distance is important. Proximity is important. If the PEF is far away from its portfolio company, it is difficult for it to understand the local context of the company and its potential partners. The idea of knowing the portfolio company’s network of local actors is thus no longer present.	Sophem is half an hour away from Siparex by car. There is no geographical distance that could impact on the formation of alliances and, in any case, a lot is done by telephone and that works very well. In contrast, the geographical distance between Sophem and Mooviin (their alliance partner) is important. It is important that the two companies are geographically close to each other.	Proximity is a plus, but it is not fundamental.	Geographical distance has no impact. Meseo and Siparex call each other and communicate by mail regularly.

1. 1) General situation

Siparex Group intervenes in facilitating the first exchanges between partners or prospective partners for an alliance. The number of portfolio companies to be managed by each investment manager at Siparex Group varies depending on the type of company being supported. In general, investment managers in charge of SMEs manage a dozen portfolio companies, while the investment managers in charge of mid-sized companies manage more or less three. As for the geographical distance separating the portfolio companies for the PEF, Siparex Group has an organization with branches at regional level, allowing the PEF to ensure a certain amount of geographical proximity to its portfolio companies.

1. 2) Impact of variables relating to the concept of time on the formation of alliances
   1. I) The PEF’s point of view

According to the associate director interviewed at Siparex Group, the number of portfolio companies to be managed by each investment manager does have an impact on the formation of alliances for its portfolio companies. He says: “If there are too many companies to manage, you don’t have enough time to support the company as you would like, in particular on the strategic level which may involve the formation of alliances”. Asked whether there may also be a positive effect on the formation of alliances, he states as follows: “Yes, you can submit proposals”. For him, one effect prevails over the other. This is the second one. He explains: “The second one prevails. The first effect, the lack of time, does not impact the formation of alliances in a negative way. If the investment manager lacks time, he will work a little more”.

Concerning the geographical distance separating the PEF from its portfolio companies, it does play a role for the formation of alliances, according to the interviewee. He explains: “If you’re more than two hours away from a portfolio company, that means that you won’t go to see them often. That means that if you’re based at Lyon and you’re supporting a company which is in Nantes, you’re not going to know the local context around the company very well, and potential partners or others. There’s this notion of knowing the network of local actors, whether these are industrial actors, economic actors or others. You also have to integrate the need to claim a geographical proximity with your portfolio companies”. Concerning Siparex Group, the PEF is generally always close enough to its portfolio companies to be able to go to them more or less quickly.

1. II) The SMEs’ point of view

For the manager of the company Mooviin, Siparex Group clearly plays a facilitating role in the first exchanges between alliance partners. He considers that the number of portfolio companies monitored by each portfolio manager does play a role in the formation of alliances. According to him, the link is positive. The more there are to monitor, the more it is a source of ideas. He says: “Yes. And then, they see different strategies, they see issues for one sector that they can duplicate in another sector. They have an advantage, it’s that they collect, they are faced with a number of situations which are slightly different. So, the experiences of one side can be of benefit to other companies”. The manager of the company Sophem shares the idea that Siparex Group plays the role of facilitator of exchanges. In contrast, for her, one effect prevails. This is that the number of portfolio companies overseen by an investment manager may represent an overload and thus a lack of time for the formation of alliances. She states: “Well, I think that, in fact, the more he has the more he’ll be overwhelmed. In general, yes that’s right, I think that he’ll then be a little bit less available for us”. The manager of the company Meseo, in contrast, states that there is no impact on the level of time: “No, that doesn’t require a lot of time”. If it has an impact, it is a favorable one for him. It is a source of ideas.

As for the geographical distance separating the PEF from its portfolio companies and the impact on the formation of alliances, the manager of the company Mooviin states: “Proximity is a plus. As for me, I love everything that is nearby. When we choose between two providers, it may play a role, yes. But it’s not fundamental”. For the manager of the company Sophem, geographic distance does not have much impact. A lot can be done by telephone, she says. Now, in her case, the portfolio manager in charge of monitoring her company is only an hour and a half away by car from her business. The manager of the company Meseo shares this point of view. For him, geographical distance does not exist. A lot can be done by telephone and by email.

To conclude, we can identify that all of the interviewees agree that the PEF plays a role in facilitating the first exchanges between partners or potential partners for alliances. Except for one person, their points of view converge on the fact that building relationships between companies is not a very time-consuming task and that, consequently, even if a high number of portfolio companies to monitor for each investment manager may result in a lack of time, this does not negatively affect their involvement in the formation of alliances. On the contrary, the number of portfolio companies monitored by an investment manager may be sources of ideas, and thus have a favorable impact on the formation of alliances. The positive, knowledge-based effect prevails over the negative effect. As for geographical distance, it has no negative effect on the formation of alliances. In contrast, proximity may be a plus.

1. b) Demeter Partners

Table 3.54 gives an overview of the points of view of the PEF on the links between: (1) the number of portfolio companies to be managed by each investment manager at Demeter Partners and the formation of alliances for its portfolio companies; (2) the geographic distance separating the PEF from the SMEs and the formation of alliances. In general, they overlap. The responses are given in detail just below the table.

Table 3.54. The Demeter Partners field with regard to Hypotheses 8a and 8b

Facilitation of exchanges	PEF	Alliance 1		Alliance 2
	Demeter Partners	Comarth	IES	Panosol	Eurener
General opinion	Demeter does intervene to facilitate exchanges between alliance partners.	Demeter does intervene to facilitate exchanges between alliance partners.	Demeter does intervene to facilitate exchanges between alliance partners.	Demeter does intervene to facilitate exchanges between alliance partners.
Number of portfolio companies per investment manager	The more portfolio companies each investment manager has to manage, the less time they have to devote to forming alliances. Within Demeter, the managers look after more or less eight portfolio companies; it varies. This does not impact the formation of alliances.	The number of portfolio companies to be managed by each investment manager does not affect Demeter’s involvement in the formation of alliances.	An investment manager should not have too many portfolio companies to manage.	The more portfolio companies there are to manage, the more beneficial this may be for the formation of alliances for the portfolio companies. This does not imply a lack of time, because the relationships are built very quickly.
Geographic distance	Geographic distance does not affect the formation of alliances because Demeter does a lot of work by telephone and by email.	Geographic distance does not affect Demeter’s activity of forming alliances.	Geographic distance does not affect the formation of alliances much. There is always a way to communicate. By phone or by email especially.	Geographic distance has no impact because there are ways to communicate by telephone.

1. 1) General situation

At Demeter Partners, various people manage different numbers of portfolio companies. We were given the example of partners, who manage approximately eight portfolio companies. The PEF comprises four partners for about 30 portfolio companies.

1. 2) Impact of variables relating to the concept of time on the formation of alliances
   1. I) The PEF’s point of view

Demeter Partners does play a role in facilitating the first exchanges between partners or prospective partners for alliances. In the case of the person interviewed at Demeter Partners, the number of portfolio companies managed has not negatively affected their involvement in alliance formation. Both effects, positive (source of ideas) and negative (lack of time), are present. But, the higher the number of portfolio companies there are to manage, the greater the negative effect, lack of time, prevails over the knowledge-based effect. A balance must therefore be found. Thus, he says: “… I think some of us are at the limit. So, I’d say six (6) is good, ten (10) is too much. The balance is quite precarious. Because if you go through files one after the other and you haven’t sold any of them in the meantime… But that affects everything, not just building relationships. It affects your reactivity, etc.”

Concerning geographical distance, it does not affect the formation of alliances. A lot is done by email and by phone.

1. II) The SMEs’ point of view

The manager of the company Comarth agrees that Demeter Partners facilitates the first exchanges between alliance partners. According to him, the number of portfolio companies supported by an investment manager or a partner in the PEF does not affect their involvement in the formation of alliances. The same goes for geographical distance, which according to him has no impact on the formation of alliances.

The manager of the company IES shares the point of view that Demeter Partners intervenes as a facilitator of exchanges in the formation of alliances. With regard to the number of portfolio companies to be managed by each investment manager and the impact on the formation of alliances, he says: “But, there’s just a balance to be found. It should not be a bad thing to encounter different people, different issues, to obtain different information. I think that there’s a good balance to find, that they should not be overwhelmed with portfolio companies to manage, but that there should be a reasonable number to manage, that there’s a certain efficiency”. Concerning the impact of geographical distance, he says: “Yes, actually, no, not a lot. Not a lot, really. We also have international relationships and there’s always a way to meet each other or to communicate by phone or email”.

The manager of the company Panosol shares the two preceding points of view on the facilitation of exchanges between alliance partners. The number to be managed by each investment manager has, according to him, a positive impact on the formation of alliances and does not constitute a loss of time. He explains: “No, because building relationships is quite fast anyway. He has to see us, so that it takes no extra time to discuss a relationship… No, on the contrary, because it’s very quick to bring companies to a first meeting. So, the more portfolio companies, the better that may be”. Concerning the impact of geographical distance separating the PEF from its supported companies on the formation of alliances, the manager of the company Panosol shares the preceding points of view. He says: “No, I don’t think so either, because we’re brought to meet each other or to communicate by telephone, so there’s no worries”.

In summary, we can identify that all of the interviewees agree that the PEF plays a role as facilitator of the first exchanges between partners or prospective partners in an alliance. Concerning the impact of the number of portfolio companies managed by each investment manager or partner at Demeter Partners, a good balance must be found so that the number of portfolio companies managed is high enough that the knowledge-based effect, the source of ideas, prevails over the negative impact of a lack of time. This is generally the case, but some people are at their limit. In summary, the effect should be mainly positive or neutral on the formation of alliances. As for geographical distance, it does not play a significant role in the formation of alliances.

1. c) Industries et Finances Partenaires

Based on the interviews we have constructed Table 3.55. It contrasts the points of view of the PEF and of the SMEs on the link between the number of portfolio companies to be managed by each investment manager at I&FP and the formation of alliances for its portfolio companies, then on the link between the geographical distance separating the PEF from the SMEs and the formation of alliances. In general, they are similar. The responses are given in detail just below the table.

Table 3.55. The I&FP field with regard to Hypotheses 8a and 8b

Facilitation of exchanges	PEF	Caterine Restauration
	Industries et Finances Partenaires	Océane de Restauration	Centrale de Restauration Martel
General opinion	I&FP does intervene to facilitate exchanges between alliance partners.	I&FP does intervene to facilitate exchanges between alliance partners.	I&FP does intervene to facilitate exchanges between alliance partners.
Number of portfolio companies per investment manager	Given that a build-up involves several acquisitions of companies, one person manages approximately two build-ups. You could say that the fewer companies one has to manage, the more time one has to spend on them.	If an investment manager has too many portfolio companies to manage, they will lack time to worry about other things. At the same time, there are then more possible exchanges between the companies. This latter effect should outweigh the first. Exchanges between the companies and the PEF are very important.
Geographic distance	Geographic distance has no impact on the formation of alliances.	Geographic distance has no impact on the formation of alliances.	Geographic distance does not matter.

1. 1) General situation

Since I&FP takes majority stakes, the number of companies they support is limited. Geographically, the PEF invests at the national level.

1. 2) Impact of variables relating to the concept of time on the formation of alliances
   1. I) The PEF’s point of view

I&FP does intervene as a facilitator of exchanges between companies forming an alliance. For the PEF, the formation of alliances is in no way linked to the number of portfolio companies to be managed. Geographic distance has no impact on the formation of alliances either.

1. II) The SMEs’ point of view

For the manager of the company Centrale de Restauration Martel, I&FP has indeed facilitated exchanges between the companies participating in the build-up. This person did not comment on the impact of the number of portfolio companies to be managed by each investment manager on the formation of alliances. Concerning geographical distance, this has played no role according to the company manager.

According to the manager of the company Océane de Restauration, I&FP does play a role in facilitating exchanges between companies in an alliance. The number of portfolio companies managed by an investment manager has, on the one hand, a negative effect on the formation of alliances, as this implies a lack of time. Thus, he says: “It is obvious that if an investment manager has too many portfolio companies to manage, he will be preoccupied by other things”. However, it should be noted that the context of the Caterine Restauration build-up means that the investment manager carries out the purchase of the companies which constitute the build-up, and that they engage a director of the holding company in charge of coordinating the activities of the companies under Caterine Restauration. Thus, even if the portfolio manager lacks time, that should have no direct impact on the formation of alliances, as the latter are put in place and managed by the director of the holding company. When asked if one might then say that lack of time does not have much impact on exchanges within the build-up, but it does for exchanges between companies in a build-up with other portfolio companies outside of it, the manager interviewed responds: “Yes, indeed”. At the same time, he agrees that the number of companies to be managed by each investment manager may also have a positive impact. “Yes, there are then more exchanges possible”, he states. In summary, the positive effect prevails. He concludes: “The positive point, that of exchanges, should prevail if communication between the portfolio companies is well structured by the investment fund. The culture of the fund is very important. In the Compass group which bought Caterine Restauration from I&FP, the culture is such that there are no exchanges between the group’s portfolio companies. We hardly know them. This is very different from the culture of the I&FP fund that we used to know. It’s easy to see. Even between the companies of Caterine Restauration, it’s no longer the same”.

Concerning the geographical distance separating the PEF from its supported companies, this has played no role. He says: “No. In any case, the companies under Caterine Restauration do come from all over France and the fund is in Paris. Geographical distance has not posed a problem”.

In summary, we can identify that the PEF does play a role in the facilitation of exchanges between partners or prospective partners for an alliance. In contrast, the number of portfolio companies to be managed by each investment manager, as with geographical distance, does not affect the formation of alliances. This is linked to the context. I&FP specializes in sectoral consolidation and takes majority stakes. Due to this, the number of companies acquired by I&FP is naturally limited. In addition, build-ups are formed in such a way that the director of the holding company created at the outset is in charge of implementing, with the company managers, the strategy to be pursued and detecting and implementing possible synergies between the companies. Because of this separation of tasks, a possible lack of time on the part of the investment managers at I&FP does not directly affect the formation of alliances for companies participating in a build-up.

1. d) Anonymous PEF

The points of view of the PEF and the SMEs have been analyzed on the basis of the interview transcripts and grouped in Table 3.56. We identify and contrast the points of view of the PEF and the SMEs on the link between the number of portfolio companies to be managed by each investment manager at Anonymous PEF on the one hand, and the geographical distance separating the PEF from the SMEs on the other hand, and the formation of alliances for its portfolio companies. In general, they agree. The responses are given in detail after the table.

Table 3.56. The Anonymous PEF field with regard to Hypotheses 8a and 8b

Facilitation of exchanges	PEF	Alliance 1	Alliance 2
	Anonymous PEF	YYY	WWW	XXX
General opinion	Anonymous PEF does intervene to facilitate exchanges between alliance partners.	Anonymous PEF does intervene to facilitate exchanges between alliance partners.	Anonymous PEF does intervene to facilitate exchanges between alliance partners.
Number of portfolio companies per investment manager	The number of portfolio companies to be managed by each investment manager does have an impact on the formation of alliances. The more there are to manage, the less time they have to spend on each of them and on the formation of alliances. At the same time, the more portfolio companies there are, the more that may be a source of ideas for alliances. The second effect outweighs the first.	Many portfolio companies to manage leads to a lack of time. Now, that should not impact the formation of alliances too much, because building relationships is not a very time-consuming activity. There is no real link.	The number of portfolio companies to be managed by each investment manager does not impact the activity of forming alliances. Now, the more portfolio companies, the more internal possibilities of realizing synergies, but that is not very important either.
Geographic distance	Geographic distance between Anonymous PEF and its portfolio companies plays a role, but mainly the geographic distance separating the alliance partners. It is more difficult to be in an alliance with a distant partner; it is more difficult to support companies that are far away from the PEF.	Proximity is clearly favorable to exchanges. The closer you are, the more that promotes exchanges. Nothing replaces physical contact.	Geographic distance has no impact either. These days there are various substitutes: Skype, telephone, email, etc. Physical presence is not mandatory.

1. 1) General situation

Within Anonymous PEF, an investment manager looks after eight to ten portfolio companies. Geographically, the PEF mainly intervenes in Lorraine and neighboring countries. It may also intervene at the national level.

1. 2) Impact of variables relating to the concept of time on the formation of alliances
   1. I) The PEF’s point of view

Anonymous PEF plays a role in facilitating exchanges between alliance partners. The number of portfolio companies to be managed by each investment director has an impact on its involvement in the formation of alliances for its supported companies. The investment manager we interviewed says: “Yes. The more portfolio companies you manage, the less time you have to devote to each one. But, at the same time, the more ideas you may have for collaborations between your portfolio companies. I think that the more portfolio companies you manage, the more probability there is of creating alliances… the source of ideas prevails. Yes, because generating the idea for an alliance does not require much time for the fund, it’s just an idea, a suggestion and a relationship to be formed. And if it’s relevant, it will follow its own path; the two managers will pursue the idea. But you must have the idea and you must find the right partner”. Concerning the geographical distance separating the PEF from the companies it supports, he says: “Yes. We don’t invest too far from our bases because we know it’s difficult to steer companies when you’re far away”. He also says: “Yes, that plays a role. Especially the geographical distance separating the two companies forming the alliance. It’s difficult to enter into an alliance with companies which are very far away”.

1. II) The SMEs’ point of view

For the manager of the company WWW, Anonymous PEF plays a role in facilitating exchanges in the formation of alliances. Concerning the impact of the number of portfolio companies to be managed by each investment manager on the formation of alliances, he says: “I think that, in fact, if in their portfolio, they end up managing a thousand portfolio companies on their own, the poor person will end up being a bit overwhelmed. Apart from monitoring their portfolio companies, they will have to always look at new requests, new portfolio companies. So, that will have an impact. Aside from that, putting them into relationships with each other, that happens in quite a natural way because they’re in contact with the companies they follow and, in those companies, when they have one on the line, they will tell them that they thought that such and such a company that makes this thing might interest them”. However, he does not think that this is a source of ideas either. He explains: “No, I think that the idea is a prerequisite. They have the idea in their head and they will submit it to the two parties independently, and if both reply that they’re interested, they will put them into contact and tell them to call back, and there it is. So, this doesn’t involve time-consuming work either”. He thus does not really confirm either effect. Concerning geographical distance, he indicates that proximity is favorable to the formation of alliances: “It’s certain that proximity is favorable to exchanges. The closer you are, the more that promotes exchanges… Nothing replaces physical contact”.

For the manager of the company YYY, Anonymous PEF plays a role as a facilitator of exchanges in the formation of alliances. According to him, the number of portfolio companies to be managed by each investment manager has no impact on this role of facilitator of exchanges. This in no way constitutes a lack of time. On the other hand, it does not constitute a source of ideas either. He says: “That is, the more there are, the more possibilities of synergies. But it doesn’t harm or improve anything”. Geographic distance has no impact on the formation of alliances either. He explains: “Skype works everywhere. You can discuss quite easily. You can exchange over your laptop and Skype and over Internet and mail, there are all the means of internal communication. Regardless of where it is. Distance is no longer a factor. That is, contrary to what a lot of French people think, the Earth is a very small planet… Physical presence is not necessary. There is no more distance. You can discuss with strangers no matter when, day or night”.

In summary, we can identify that all the interviewees agree on the fact that Anonymous PEF intervenes in the facilitation of exchanges between partners or prospective partners for an alliance. In contrast, points of view diverge as to the effects of the number of portfolio companies to be managed by each investment manager and of geographical distance. As a generalization, the number of portfolio companies to be managed by each investment manager has an impact that is either mildly positive or neutral on the formation of alliances. The number of portfolio companies may be a source of ideas. However, other sources of ideas may be more important, like the number of files analyzed by the investment committee before deciding whether to invest in a prospect. The number may also have the impact of a lack of time, but this should not be very important given that intervention in the formation of alliances is not a time-consuming act. As for geographical distance, it has no negative impact. In contrast, geographical proximity may have a mildly positive effect on the formation of alliances.

1. b) The second level of analysis

After having analyzed field by field the points of view of the PEFs on the one hand, and the SMEs forming alliances on the other hand, we now contrast the summary obtained for each field with the others (Table 3.57).

Table 3.57. Interfield summary with regard to Hypotheses 8a and 8b

Facilitation of exchanges	Siparex Group	Demeter Partners	Industries et Finances Partenaires	Anonymous PEF
	Yes	Yes	No	Yes
Mechanism of causality	The PEF plays a positive role in facilitating exchanges between alliance partners. For the number of portfolio companies to be managed by each investment manager, the knowledge-based, positive effect of a source of ideas prevails over the contractual, negative effect of lack of time. Geographical distance does not appear to have a significant effect on the formation of alliances. The effect is not negative, but rather neutral or even mildly positive.	The PEF plays a positive role in facilitating exchanges between alliance partners. For the number of portfolio companies to be managed by each investment manager, there is a balance to be found. Both effects (knowledge-based and contractual) play a role. The positive, knowledge-based effect of a source of ideas prevails over the contractual effect, if the number of portfolio companies to be managed is not too high, which is generally the case but may also be at its limit. In summary, the effect is either positive or neutral. Geographic distance has no impact on the formation of alliances.	The PEF plays a positive role in facilitating exchanges between alliance partners. Neither the number of portfolio companies to be managed by each investment manager nor geographical distance plays a role in the formation of alliances.	The PEF plays a positive role in facilitating exchanges between alliance partners. For the number of portfolio companies to be managed by each investment manager, the effect is neutral, even slightly positive. Both effects (knowledge-based and contractual) play a role. The positive, knowledge-based effect of a source of ideas prevails slightly over the contractual effect, but other sources of ideas may be more important. Either geographical distance has no impact on the formation of alliances, or geographical proximity has a slightly favorable link for the formation of alliances.

All of the participants in the study and for all four fields agree that the PEFs are involved as facilitators of exchanges – in general, the first – between partners or prospective partners in an alliance. Three of the four fields of analysis affirm that the number of portfolio companies to be managed by each investment manager may be a source of ideas for alliances. At the same time, too high a number of portfolio companies to manage may lead to a lack of time. The knowledge-based and contractual mechanisms of causality thus seem plausible. Generally, the purely knowledge-based effect prevails. The intervention of a PEF in the formation of alliances does not constitute a time-consuming task. The I&FP field is the only field to not agree with these points of view. This is, again, due to the context. Since the PEF takes majority stakes, the number of companies it buys via holding companies created for this purpose is limited, as well as the number of build-ups (which constitute I&FP’s portfolio companies) managed by each investment manager. In addition, investment managers generally hire directors for the holding companies, who are in charge of detecting and implementing alliances. The question of the number of portfolio companies managed therefore does not make much sense for the formation of alliances in the case of this PEF.

Geographical distance does not receive so much support across the four fields of analysis. It has no impact on the formation of alliances. A lot can be done by telephone and by e-mail, which means distance no longer plays a role. On the other hand, geographical proximity may have a mildly positive impact on the formation of alliances.

The PEF’s point of view

In the light of knowledge-based theories and from the point of view of the PEFs, we argued that the formation of alliances constitutes a means of differentiation on the private equity market, which gives the PEF a competitive advantage (Hypothesis 9). It is time for this hypothesis to be tested empirically.

HYPOTHESIS 9.– Providing contacts allowing the formation of alliances constitutes a means of differentiation on the private equity markets for PEFs.

1. A) First level of analysis
   1. a) Siparex Group

A summary of the points of view of the PEF and the SMEs on the link between the PEF’s activity of forming alliances and its activity of differentiation on the private equity market is presented in Table 3.58. The analysis is performed on the basis of the interview transcripts. Generally, the opinions of the different participants do not overlap. The responses are given in detail just below the table.

Table 3.58. The Siparex Group field in relation to Hypothesis 9

Differentiation	PEF	Alliance 1	Alliance 2
	Siparex Group	Sophem	Mooviin	Meseo
General opinion	The formation of alliances for its portfolio companies does constitute a service for Siparex, which allows it to differentiate itself on the private equity market.	Siparex promoted itself by providing contacts for alliances.	Providing contacts and the formation of alliances are not at all put forward by Siparex when it approaches companies.	Providing contacts for the formation of alliances was not a sales pitch from Siparex. It was offered a little bit, as a complementary service, but did not have much importance.

1. 1) The PEF’s point of view

The formation of alliances for its portfolio companies does constitute for Siparex a means of differentiation on the private equity market. Club Siparex was created in particular for this goal, more than 36 years ago. The interviewee states: “It’s always been a desire of Siparex to cultivate this dimension of building relationships, which for us is a differentiating factor with respect to other actors in private equity”.

1. 2) The SMEs’ point of view

The manager of the company Sophem stated: “In their presentation, they promoted it, it was a plus because, well, we’d also contacted other funds, and they promoted themselves with respect to that, yes”. For the manager of the company Meseo, building relationships within alliances was not perceived by him as a means of differentiation on the part of Siparex Group. In any case, for him, it is not a service of great importance. It’s a complementary service. The manager of the company Mooviin shares the above point of view. He says: “No. No, absolutely not. When Siparex comes, that’s not what they will promote”.

In summary, these testimonies indicate a willingness from Siparex Group to differentiate itself on the private equity market, among other ways by providing contacts for the formation of alliances. This fact is, however, not always perceived by the managers of the companies it supports.

1. b) Demeter Partners

Based on the interview transcripts, we have analyzed and summarized the points of view of the PEF and the SMEs on the activity of formation of alliances by the PEF as a means of differentiation on the private equity market. An overview is given in Table 3.59. Overall, the points of view of the PEF and the SMEs do not coincide. The responses are given in detail just below the table.

Table 3.59. The Demeter Partners field in relation to Hypothesis 9

Differentiation	PEF	Alliance 1		Alliance 2
	Demeter Partners	Comarth	IES	Panosol	Eurener
General opinion	The formation of alliances for its portfolio companies does constitute a service for Demeter that allows it to differentiate itself on the private equity market. This service is promoted as a rule and is also found on their two-sided brochures.	Difficult to respond because no experience in comparison to other PEFs. In any case, the service of forming alliances did not weigh at all on the decision of Comarth’s manager to work with Demeter.	It appealed to IES that Demeter could provide contacts, but they did not differentiate themselves in relation to providing contacts or forming alliances, but because they are a sectoral PEF, not solely financial but also possessing an engineering and business culture.	It is more the sectoral side which was promoted, and not providing contacts for the formation of alliances.

1. 1) The PEF’s point of view

For Demeter Partners, the service of forming alliances is indeed a way of differentiating themselves on the private equity market. The portfolio manager we talked to said: “Yes. Within Demeter, we also have what we call Club Demeter, the club for entrepreneurs. That’s what we also promote as a rule. It’s even in our two-sided brochure… Compared with a generalist fund, we can tell them that they have a place to meet up; we have a forum on the site which is reserved for Club members. And then, every three months, they meet, they do business together; we gather them around a theme linked to their sector in general. So, yes, there are things which are hard for a generalist fund to offer”.

1. 2) The SMEs’ point of view

For the manager of the company Comarth, it is difficult for him to say if Demeter Partners promoted its service of forming alliances when they approached his company, given that he was not present then. But he does not think so, and this fact did not weigh on his decision to work with the PEF. The manager of the IES company shares this point of view. He says: “No. Although, yes, it interested us that they brought contacts, but they differentiated themselves as being sectoral and, mainly, not being just financiers but having a whole engineering and business culture”.

In summary, the testimonies indicate that for Demeter Partners, providing contacts for the formation of alliances is one of the elements in their package of differentiation. The element of alliance formation which is part of this is not, however, vital or perceived as decisive by the company managers.

1. c) Industries et Finances Partenaires

The points of view of the PEF and the SMEs have been analyzed on the basis of the interview transcripts and grouped in Table 3.60. We identify and contrast the points of view of the PEF and the SMEs on the activity of formation of alliances by the PEF as a means of differentiation on the private equity market. On the whole, the points of view of the PEF and the SMEs do not agree. The responses are given in detail after the table.

Table 3.60. The Industries et Finances Partenaires field in relation to Hypothesis 9

Differentiation	PEF	Caterine Restauration
	Industries et Finances Partenaires	Océane de Restauration	Centrale de Restauration Martel
General opinion	Everyone claims to have a network. I&FP says that it exists, but it is not necessary used to form alliances. However, in a commercial approach, we are able to offer contacts to introduce companies to others outside of the build-up.	I&FP did not differentiate itself as a PEF by providing contacts for the formation of alliances, but by the industrial project they proposed to Océane de Restauration, which was to become part of a build-up.	Providing contacts for the formation of alliances is a plus which was subsequently provided but was not a sales pitch or a means for I&FP to differentiate itself on the private equity market.

1. 1) The PEF’s point of view

I&FP does use its network of contacts to differentiate itself on the private equity market. So, according to the associate director: “Everyone claims to have a network. Except that we said that it exists”. I&FP’s main distinction for differentiating itself on the private equity market, however, remains specialization in sectoral consolidation.

1. 2) The SMEs’ point of view

For the manager of the company Centrale de Restauration Martel, providing contacts for forming alliances was not specifically promoted by I&FP to differentiate itself on the private equity market. It was rather a bonus which was subsequently provided. The manager of the company Océane de Restauration shares this point of view. He says: “No. In any case, I did not perceive it as such. It differentiated itself through the industrial project that it had for Océane de Restauration”.

In summary, the testimonies indicate that for I&FP, providing contacts for the formation of alliances is one of the elements in its package of differentiation. The element of alliance formation which is part of this is not, however, vital or perceived as decisive by the company managers. What is decisive is I&FP’s core business, specialization in sectoral consolidation and, thus, the projects proposed to managers who want to sell their company.

1. d) Anonymous PEF

As for the previous fields, the points of view of the PEF and the SMEs have been analyzed on the basis of the interview transcripts and grouped in Table 3.61. We contrast the points of view of the PEF and the SMEs on the activity of formation of alliances by the PEF as a means of differentiation on the private equity market. Generally, the points of view of the PEF and the SMEs are distinct. The responses are given in detail just below the table.

Table 3.61. The Anonymous PEF field in relation to Hypothesis 9

Differentiation	PEF	Alliance 1	Alliance 2
	Anonymous PEF	YYY	WWW	XXX
General opinion	To differentiate itself from other PEFs, Anonymous PEF especially promotes its ability to build relationships between its portfolio companies in alliances.	For YYY, providing contacts for the formation of alliances was not promoted by Anonymous PEF at the time of their approach. A possibility of building relationships was mentioned, but in no way influenced the manager of YYY’s decision to offer their company for support by Anonymous PEF. This was rather provided subsequently.	For WWW, providing contacts for the formation of alliances was not promoted by Anonymous PEF at the time of their approach. It was a bonus provided subsequently.

1. 1) The PEF’s point of view

According to the investment manager interviewed at Anonymous PEF, the formation of alliances for its portfolio companies constitutes a means of differentiation on the private equity market.

1. 2) The SMEs’ point of view

According to the manager of the company WWW, Anonymous PEF did not promote the service of forming alliances to differentiate themselves on the private equity market. It was rather a service which was provided to him later. Thus, he says: “Yes. I do not think that that’s the way they differentiate themselves from others. Although, I don’t know, I don’t have that experience of the others either. Where there is a great difference with respect to the others, it’s that it’s territorialized, and that it’s a fund specializing in materials. It’s a fund based on materials and not on agro-food which may have portfolio companies in agro-food, in energy, in aeronautics, in service, in information technology or anything else, but always linked to materials. So, by necessity, the companies in the fund, in their portfolio, all have something in common: materials. Either they are users, or producers, or transformers or constructors of materials. So, by necessity, there’s naturally a non-zero potential that companies in this fund might have common interests and be able to work together”.

The manager of the company YYY shares this point of view. He says: “No, no, that wasn’t promoted. That was a bonus which came subsequently… This possibility [building relationships] had been mentioned. But that in no way influenced the decision to build a relationship between Anonymous PEF and YYY. It wasn’t a determining factor. For them, perhaps it was a little. That is, in the choice of companies, there had to be anyway a kind of synergy or certain contacts like, being on the board of directors of the Fibres cluster. Now, it’s certain that in their choice of portfolio, there had to be anyway a kind of synergy, a kind of complementarity or a certain unity in choosing their portfolio companies. That’s for sure. But, for me, it was not determining at all; for them, more so”.

In summary, these testimonies indicate that for Anonymous PEF, providing contacts for the formation of alliances is one of the elements in its package of differentiation. The element of alliance formation which is part of this is not, however, vital or perceived as decisive by the company managers. What is decisive is that Anonymous PEF is a regional PEF, specializing in the field of materials.

1. B) The second level of analysis

After having analyzed, field by field, the points of view of the PEFs on the one hand, and the SMEs forming alliances on the other hand, we now contrast the summary obtained for each field with the others (Table 3.62).

Table 3.62. Interfield summary in relation to Hypothesis 9

Differentiation	Siparex Group	Demeter Partners	Industries et Finances Partenaires	Anonymous PEF
Supports the hypothesis	Yes/no	Yes/no	Yes/no	Yes/no
Mechanism of causality	Providing contacts for the formation of alliances is part of the package of differentiation, according to the PEF. It is not perceived as such by the company managers interviewed.	Providing contacts for the formation of alliances is part of the package of differentiation, according to the PEF. It is not perceived as such by the company managers interviewed.	Providing contacts for the formation of alliances is part of the package of differentiation, according to the PEF. It is not perceived as such by the company managers interviewed.	Providing contacts for the formation of alliances is part of the package of differentiation, according to the PEF. It is not perceived as such by the company managers interviewed.

Across the four fields of analysis, a clear disparity is visible between the points of view of the PEFs on the one hand, and those of the SMEs on the other hand, as regards Hypothesis 9. For all the PEFs who were objects of analysis, the provision of contacts for the formation of alliances is part of their package of differentiation on the private equity market. In contrast, this does not seem to be particularly perceived by the SME managers interviewed. For the latter, if this fact was mentioned during the PEF’s approach, it was then not decisive in their decision to work with the PEF. Often, also, forming alliances thanks to the PEF appeared more as a bonus which was brought to them subsequently. For the company managers interviewed, PEFs are distinguished either by the project which they propose to implement (this is the case for I&FP), or again by the fact that they are sectoral (this is the case for Demeter Partners and Anonymous PEF), or by their regional, national or international foothold.

3.2.2.6. Conclusion of the multiple case study

Table 3.63 presents, for each hypothesis, the conclusion of the case study as to its validity and the plausibility of the mechanism of causality advanced.

Table 3.63. Summary of results for the multiple case study by hypothesis

Hypothesis	Independent variable	Support or rejection of the hypothesis and of the mechanism of causality
Contractual argument
H1	Reputation of the PEF	The hypothesis and the mechanism of causality are confirmed.
H2	Weak reputation of the PEF	The hypothesis and the mechanism of causality are unanimously rejected.
H3	State/region/Competitiveness Cluster	The impact of the State/region seems weak, or even neutral, as does the impact of the Clusters. In contrast, the mechanism of causality indicating that the impact is stronger when the PEF takes the form of a venture-capital joint-stock company is confirmed.
H4	Taking majority stakes	The hypothesis is confirmed, but not the mechanism of causality. Taking a majority stake does not necessarily have more impact on the formation of alliances than taking a minority stake.
H6	Unlisted status of portfolio companies	The hypothesis is confirmed, but not so much the mechanism of causality.
H10	Intraportfolio alliances	The hypothesis is confirmed in part. The roles of PEFs in the formation of intraportfolio alliances are much more intensive than in the case of the formation of extraportfolio alliances. In contrast, it does not follow that there is more frequent formation of intraportfolio than extraportfolio alliances.
Contractual and knowledge-based arguments
H5	Previous alliances	The hypothesis is confirmed in its knowledge-based dimension. The impact of previous alliances is slightly positive or neutral. The contractual dimension does not seem to have an impact, even if the mechanism of causality is plausible.
H8	Facilitation of exchanges	The hypothesis and the mechanism of causality are confirmed.
H8a	Number of portfolio companies	The hypothesis and the contractual and knowledge-based mechanisms of causality are confirmed. The knowledge-based effect prevails over the contractual effect.
H8b	Geographic distance	The impact of geographic distance on the formation of alliances is weak or even neutral. In contrast, geographic proximity has a slightly favorable impact.
Knowledge-based arguments
H7	Expertise of PEF	The hypothesis and the mechanism of causality are confirmed.
H7a	Sectoral PEF	The hypothesis and the mechanism of causality are confirmed.
H7b	Cross-border PEF	The hypothesis and the mechanism of causality are confirmed.
H9	PEF diversification	The hypothesis and the mechanism of causality are confirmed by all the PEFs participating in the study. In contrast, none of the SMEs participating in the study could confirm it.
H11	Access to information	The hypothesis and the mechanism of causality are confirmed, except for one field due to context.
H11a	Seats on board of directors	The hypothesis and the mechanism of causality are confirmed.
H11b	Associations	The hypothesis and the mechanism of causality are confirmed, but the impact on the formation of alliances is weak compared to H11a.

The multiple case study provides support for some hypotheses and their mechanism of causality. At the level of the contractual hypotheses, this is Hypothesis 1, of a link between the PEF’s reputation and the formation of alliances for its portfolio companies, and of the mechanism of causality advanced.

At the level of the knowledge-based hypotheses, the multiple case study supports the hypotheses and the plausibility of the mechanism of causality advanced in the following cases:

• – Hypothesis 7, of a link between the PEF’s expertise and the type of alliances it enables to be formed, as well as Hypotheses 7a and 7b;
• – Hypothesis 8 of the role of the PEF as a facilitator of exchanges between (prospective) partners to an alliance and Hypotheses 8a and 8b. For Hypothesis 8a, of the impact of the number of portfolio companies to be managed by each investment manager, the knowledge-based argument (the source of ideas) outweighs the argument of a lack of time, although both impacts are plausible;
• – Hypothesis 11 of the PEF’s access to certain sources of information and the formation of alliances, and Hypotheses 11a and 11b. The source of information “holding seats on strategic boards” (11a) is more important than the source “participation in associations” (11b).

Some hypotheses and their mechanisms of causality are confirmed in part. At the level of the contractual hypothesis, these are:

• – Hypothesis 3, of a link between the presence of the State or of a region in the capital of the PEF and also the involvement of a Competitiveness Cluster and the formation of alliances for its portfolio companies. Overall, the presence of the State, of a region or of a Competitiveness Cluster has quite a weak or even a non-existent impact on the formation of alliances. In contrast, the mechanism of causality advanced whereby the impact of the State, of a region or of a Competitiveness Cluster on the formation of alliances is strengthened when the PEF takes the form of a venture-capital joint-stock company is confirmed;
• – Hypothesis 4, of a link between the PEF taking a majority stake and the formation of alliances for its portfolio companies. The case study does confirm a positive link between the two variables. In contrast, the mechanism of causality is not at all confirmed. The variable “majority stake” does not appear to be a decisive indicator. It does not have much more impact on the formation of alliances than taking a minority stake;
• – Hypothesis 6, of a link between the unlisted status of portfolio companies and the formation of alliances. The hypothesis predicting greater roles for the PEF in the formation of alliances in the presence of unlisted companies is confirmed. In contrast, the mechanism of causality advanced did not find much support.
• – Hypothesis 10, of a greater role for PEFs in the case of the formation of intraportfolio alliances. The roles of PEFs in the formation of alliances are indeed more important in the case of forming intraportfolio alliances. In contrast, we cannot confirm a greater formation of intraportfolio than extraportfolio alliances.

At the level of knowledge-based hypotheses, this involves:

• – Hypothesis 9, of a link between the PEF’s strategy of diversification on the private equity market and the formation of alliances. The hypothesis and the mechanism of causality are unanimously confirmed by the PEFs, but have not been perceived as such by any of the company managers participating in the study;
• – Hypothesis 5, of a link between the number of alliances previously formed by a portfolio company on the inclination of the PEF to form an additional alliance for this portfolio company. The hypothesis is supported in its knowledge-based dimension. The contractual impact, although plausible, is not confirmed by the facts. The knowledge-based effect is plausible and finds confirmation. Nevertheless, its impact on the formation of alliances is weak, or even neutral.

The multiple case study also allows us to unanimously reject one of the hypotheses as well as the mechanism of causality advanced. This is Hypothesis 2, of the positive impact of a weak reputation of the PEF on the formation of alliances for its portfolio companies.

3.2.3. Reconciliation of the econometric study and the multiple case study and discussion of the results

3.2.3.1. Reconciling the two empirical studies

Having presented the econometric study and the multiple case study, we now confront and compare their results in order to reach a conclusion at the level of the multimethod study as a whole. Before proceeding with this comparison, let us recall its main objectives.

The multimethod study, comprising an econometric study and a multiple case study, was chosen to answer three questions:

1. – why does the phenomenon (formation of alliances in the presence of a PEF) take place?
2. – how do the PEFs intervene in the formation of alliances for the companies they support?
3. – are the phenomenon and the explanations advanced generalizable to all French PEFs?

The multiple case study was intended to provide answers to the first two questions, and the econometric study, to the third.

We proceed to reconcile the results of the two studies in the following way: first, we will combine the results table for the multiple case study that we have just presented (Table 3.63) with that arising from the econometric study (Table 3.3). The table determines the rejection or acceptance of the hypotheses across the multimethod study as a whole. We then discuss the summary of the results obtained by the two methods. We will then take an overview of the results obtained, attempting to reply to the three main questions that were raised.

The integration of the results obtained from the econometric study (Table 3.3) into the table summarizing the results of the multiple case studies (Table 3.63) allows us to present Table 3.64.

Table 3.64. Reconciliation of the results of the econometric study and the multiple case study

Hypothesis	Independent variable	Expected link to the formation of alliances	Support/rejection of the hypothesis by the econometric study	Support/rejection of the hypothesis by the multiple case study	Support/rejection of the mechanism of causality advanced by the multiple case study	Conclusion of the multimethod study
Contractual argument
H1	Reputation of the PEF	+, Especially for the formation of extraportfolio alliances	Support for allex, not significant for allin but expected sign	Support	Confirmed (and supported by the overall results of the econometric study)	Support
H2	Weak reputation of the PEF	+ On building relations with well-known alliance partners	Not testable	Unanimous rejection	Unanimous rejection	Rejection
H3	State/region/Competitiveness Cluster	+ And more important role when the PEF takes the form of a venture-capital joint-stock company	Rejection for the State, not significant for Competitiveness Cluster	The impact of the State/region seems weak, even neutral, as for the impact of Clusters	The mechanism of causality indicating that the impact is stronger when the PEF takes the form of a venture-capital joint-stock company is confirmed	Ambiguous, even neutral results leading us to conclude a rejection. The mechanism of causality is confirmed
H4	Taking majority stakes	+	Rejection for allin, not significant for allex	Support	Rejection of mechanism of causality. Taking a majority stake does not necessarily have more impact on the formation of alliances than taking a minority stake	We thus conclude rejection of the variable chosen to test the hypothesis, and rejection of the mechanism of causality. In contrast, the case study confirms the “disciplinary” or advisory role in the weaker version played by the PEFs
H6	Unlisted status of portfolio companies	+	Not significant	Support	Mostly rejection of mechanism of causality	Tendency to reject the hypothesis which is “only” supported by the fact that by nature, the PEFs essentially only support unlisted SMEs
H10	Intraportfolio alliances	+	Tendency to not support	Support in part	The roles of PEFs in the formation of intraportfolio alliances are much more intensive than in the case of extraportfolio alliance formation. In contrast, it does not follow that there is more frequent formation of intraportfolio than extraportfolio alliances	Support in part. The intensity of the roles played is stronger in intraportfolio alliances, but the PEFs play an important role in intraportfolio as much as in extraportfolio alliances
Contractual and knowledge-based arguments
H5	Previous alliances	-/+ (contractual/knowledge-based argument)	Not testable	The hypothesis is confirmed in its knowledge-based dimension	The impact of previous alliances is slightly positive or neutral. The contractual dimension does not seem to have an impact, even if the mechanism of causality is plausible	Support for the knowledge-based dimension
H8	Facilitation of exchanges	+		Support	Support	Support
H8a	Number of portfolio companies	-/+ (contractual/knowledge-based argument)	Not significant	Support	Support. The knowledge-based effect prevails over the contractual effect	Tendency to support the knowledge-based dimension
H8b	Geographic distance	-	Not testable	Mostly rejection	The impact of geographic distance on the formation of alliances is weak or even neutral. In contrast, geographic proximity has a slightly favorable impact	Rejection
Knowledge-based argument
H7	Expertise of PEF	+		Support	Support	Support
H7a	Sectoral PEF	+ On the formation of sectoral/regional alliances	Tendency to support with no conclusion possible for allex	Support	Support	Support
H7b	Cross-border PEF	+ On the formation of alliances of an international character	Tendency to support with no conclusion possible for allin	Support	Support	Support
H9	PEF diversification	+	Support	Support by the PEFs, no support by the SMEs	Support by all the PEFs participating in the study. In contrast, none of the SMEs participating in the study could confirm it	Support from the PEFs’ point of view
H11	Access to information	+		Support	Support	Support
H11a	Seats on board of directors	+	Not significant	Support	Support	Tendency to support
H11b	Associations	+	Not testable	Support	Support, but the impact on the formation of alliances is weak compared to H11a	Support but weak

Let us discuss the results of Table 3.64. To do this, it is necessary to recall the goal of comparing the results arising from the two studies, as well as certain limitations encountered in their implementation, as they have an impact on the achievement of the goals pursued by the overall study. This mainly concerns the limits related to our econometric study. Aside from the main objective of the multimethod study and its components, which is to respond to the general questions that we have just restated, comparing the results of the two studies should allow the triangulation of results. In concrete terms, this means seeing if the results of the two analyzes overlap or not in their trends. If this is the case, the econometric study will then allow a statistical generalization of the results, while the multiple case study will enable the refinement of the analysis by testing the results in specific contexts and, to go beyond the correlations detected by the econometric analysis by evaluating the plausibility of the mechanisms of causality advanced, in order to result in a generalization of the analytical type.

The limits of our econometric study, mentioned in the relevant section, do not however make it entirely possible to triangulate the results obtained with those from the multiple case study. Besides the difficulties encountered in the study due mainly to our small sample size, the main limit relates to the fact that the questionnaire intended for SMEs supported by private equity did not succeed. Consequently, because of its general design, the econometric study boils down to an analysis of the correlations between, on the one hand, the perceptions that French PEFs who are members of AFIC/Invest France have of the impact of certain variables on the formation of alliances for the companies they support and, on the other hand, estimation of the actual rate of alliances that their portfolio companies have been able to form. In light of the goal of triangulation of the results via the two methods (econometric and case study), we cannot claim anything but a partial triangulation of the results, for the PEFs’ points of view only.

This limitation impairs the realization of the goal of statistical generalization of the obtained results. These can only be established, on the one hand, from the point of view of the PEFs and, on the other hand, must be interpreted with caution due to the weak robustness of our results. However, the descriptive data collected testify to a generalized importance of the formation of alliances for companies supported by private equity. This may be a first indicator of a tendency to the generality of the results. A second econometric investigation could therefore be interesting in order to further explore not only the phenomenon of the formation of alliances, but also the roles that PEFs seem to play in it.

Because of these limitations, the results of the multimethod study essentially rely on those of the multiple case study. The econometric study’s results are mainly useful to confirm those of the case study. While being aware of this point, we draw up a summary and recapitulation of the results of both methods in interpreting Table 3.64. Generally, we notice that all the knowledge-based variables have been confirmed across the multimethod study and, in particular, across the multiple case study. When results were obtained from the econometric study, they support (or tend to support) those of the case study. As for the contractual variables, they are not all confirmed across the study. Some are rejected (Hypothesis 2), other variables are only confirmed in part. Over all the contractual variables, only the variable Rp (reputation of the PEF) finds its assumed positive impact on the formation of alliances confirmed across both studies (econometric and case study) and enables confirmation of the mechanism of causality advanced. Finally, for the variables of mixed nature (contractual and knowledge-based), the knowledge-based variant prevails over the contractual variant. Let us now discuss these different results.

3.2.3.2. Discussion of results

The discussion of results includes three parts. First, we discuss the results hypothesis by hypothesis. Then, we undertake a reflection on their presentation. Finally, we try to take an overview of the results obtained in response to the questions that we posed at the beginning of the study.

Discussion of the results hypothesis by hypothesis

We begin with the results of the hypotheses arising from the contractual approach. We continue with those of the hypotheses resting on argumentation of mixed (contractual and knowledge-based) character, and end with the results of the empirical testing of the assumptions arising from the knowledge-based approach.

Results for the contractual approach hypotheses

The contractual approach led us to propose seven hypotheses, including one of mixed character (contractual and knowledge-based: Hypotheses 5a and 5b), presented in a separate section. The six others are Hypotheses 1–4, 6 and 10, whose results appear at the beginning of Table 3.64. The contractual approach mainly proposes two mechanisms of intervention: reputation (Hypotheses 1 and 2) and discipline (Hypothesis 4). This approach then allows us to take account of the interests of certain investors in the formation of alliances (Hypothesis 3). As for Hypotheses 6 and 10, they allow us to relativize the role of the PEFs, either because they only constitute one mechanism among others which may have an impact on the formation of alliances (Hypothesis 6), or because their roles may be more or less intense depending on their position within the structure of the alliance formed (intraportfolio vs. extraportfolio – Hypothesis 10).

Hypothesis 1 predicts a positive impact for the reputation of the PEF on the establishment of trust between prospective alliance partners, enabling the reduction of transaction and agency costs. It thus has a positive impact on the formation of alliances. Our argument stated that the PEFs’ reputation is more important in the establishment of a situation of trust between the actors in the case of the formation of extraportfolio alliances. For intraportfolio alliances, trust can be established more easily without necessarily requiring the reinforcement of the mechanism of reputation. The hypothesis and the advanced arguments were confirmed throughout the totality of the multimethod study. The multiple case study confirmed the hypothesis and emphasized that the reputation of PEFs includes two components: the reputation of PEFs in general, and the specific reputation of the PEF compared to its peers. The component which is always involved is the first. The second is, in general, only perceived by the PEFs themselves. Company managers only perceive it if they have already been in contact with the PEF on a previous occasion. This may be the case if they have, for example, at one time or another, been in the position of a prospect, or if the PEF possesses a strong regional or sectoral presence. The result of the econometric study also corroborates these predictions. The allin and allex regressions both presented a positive sign, but the coefficient only appeared significant for the allex regression.

With regard to the literature, our result is consistent with that of Hsu [HSU 06], who found a positive link between the reputation of venture capital firms and the formation of alliances. However, our study provides complementary information. Apart from the confirmation of the plausibility of the mechanism of causality advanced by the case study, our study also enables us to state that the mechanism of reputation plays a more pronounced role in the establishment of trust between actors – on the basis of the minimization of transaction and agency costs – within extraportfolio alliances. In the context of forming an intraportfolio alliance, the mere presence of the PEF is enough to establish a link of trust without necessitating the intervention of the mechanism of reputation. In other words, even a PEF of weak reputational capital may have a positive impact on the establishment of trust between actors in the context of forming intraportfolio alliances.

Hypothesis 2, according to which building relationships within alliances for the SMEs that they support with well-known partners on the market constitutes for a PEF with a weak reputation a complementary or even substitute mechanism for the role of certification played by well-known PEFs, was not testable through our econometric study. On the other hand, it could be tested through the multiple case study. The hypothesis was unanimously rejected by all the participants (both the SMEs’ and the PEFs’ points of view) and across all eight cases. The literature discusses the role of certification of the quality of start-ups by well-known partners on the markets (including alliance partners), well-known PEFs [MEG 91] and the interaction between the two [OZM 13]. However, in our study, the PEFs do not seem to choose potential alliance partners with reputations on the markets as a mechanism enabling better valuation of their portfolio companies on the markets in view of a more rapid exit, or a better valuation. Neither do the managers of the SMEs supported by private equity adopt such a strategy when they choose alliance partners.

Hypothesis 3 predicted a positive impact on the formation of alliances of certain investors in the PEF’s capital, such as the State or a region, or again the involvement of the PEF in a Competitiveness Cluster. Our argument stated that the impact of investors is more important in the case where the PEF takes the legal form of a venture-capital joint-stock company. These actors effectively promote networking between actors, which may suggest a positive effect for their presence on the formation of alliances. Our study shows, however, that these actors do not seem to have a decisive impact on the formation of alliances, in light of the results of the multiple case study. The mechanism of causality according to which the impact should be more important in the case where the PEF takes the form of a venture-capital joint-stock company is, however, confirmed. The results of the econometric study allow us to reach a conclusion only as to the presence of the State. However, they indicate a sign contrary to that expected for this link: the PEFs who indicated a positive link between the presence of the State and the formation of alliances tend to form fewer alliances than the PEFs who indicated a neutral or negative impact. This result seems curious and must be relativized. First, the economic study does now allow us, in view of our data, to distinguish PEFs taking the form of a venture-capital joint-stock company from those taking the form of a VCMF. Now, the impact is assumed to be greater in the first case. Thus, if the PEFs who replied to the questionnaire mainly take the legal form of a mutual fund, it is understandable that the expected impact would not be decisive. Only, in this case, we would thus expect a rather neutral impact. However, the result indicates a negative link. This result may then be linked to the limits already mentioned of our econometric study. Another possible explanation is that the PEFs who replied in favor of the existence of a positive effect of the presence of the State on the formation of alliances are specific kinds of PEF, public ones for example, who by nature would invest in companies less prone to the formation of alliances, or, again, that they are recently created PEFs who at the time form fewer alliances than some of their peers. It might be interesting, consequently, to study this point more deeply. Our data does not allow us to test these hypotheses at present.

Hypothesis 4 predicts that PEFs taking majority stakes in the companies they support has a positive impact on the formation of alliances. The argument advanced, of a contractual nature, is that the PEFs can resolve situations of conflicts of interest between alliance partners and intervene in the presence of uncooperative behaviors. They thus play a disciplinary role (or advisory role, in the weaker version). This intervention is supposed to be facilitated when the PEF holds a majority stake in the companies it supports which form the alliance. The multiple case study enables the corroboration of the hypothesis of the potential disciplinary role played by the PEFs. The study also reveals that the term “disciplinary” is too strong to describe the role of the PEFs, who play more of an advisory role (or weakened disciplinary role). The PEFs will not punish behavior, nor intervene in the details of management, but may intervene to grease the wheels. The study allows us to conclude, at the same time, on the rejection of the mechanism of causality advanced, linked to taking majority stakes. Thus, this variable does not seem to be an adequate proxy to identify this function, as PEFs play this role independently of the majority or minority character of their stake.

Our study thus enables us to confirm the argument advanced in the literature which, however, has not been explicitly tested. Thus, Lindsey [LIN 08] in particular advances this explanation to justify her result establishing an increased probability of formation of alliances in the presence of a PEF if they are of the intraportfolio type. However, she does not explicitly test a contractual variable to explain this result. The use of the case study, in our work, does however allow us to verify the plausibility of the mechanism of causality advanced.

Hypotheses 6 and 10 enable us to relativize the roles played by PEFs in the formation of alliances. The first arises from the argument according to which PEFs may only be one mechanism of governance among others having an impact on the formation of alliances. The second, based on the mobilization of the concept of social capital in a contractual perspective, enables investigation of the intensity of the roles played by PEFs according to the intraportfolio and extraportfolio structures of alliances, and the respective position occupied by the PEFs within these structures.

Hypothesis 6 thus rests on the argument that the role of PEFs tends to reduce with the opening up of the capital of the companies they support, due to the arrival of new investors. The hypothesis predicts a greater role for PEFs in the formation of alliances between unlisted companies. In accordance with Lindsey’s [LIN 08] study, alliances seem to be formed more frequently in the presence of unlisted SMEs. In contrast, our case study, which allows us to check the plausibility of the mechanism of causality advanced, shows that the latter is not supported. The listed/unlisted status of the companies also appears as not significant in the econometric study. The explanation of this result therefore seems to be simply linked to the fact that, by nature, PEFs intervene mainly in unlisted companies.

According to Hypothesis 10, the role of PEFs in the formation of alliances is greater in the case of the formation of intraportfolio alliances. This hypothesis finds confirmation in the sense that the roles they play are more intensive in the context of intraportfolio alliances. Contrary to Lindsey’s [LIN 08] result, we cannot however conclude on more frequent formation of intraportfolio rather than extraportfolio alliances. Two points should, however, be considered. The first is that Lindsey shows that intraportfolio alliances are more common than alliances where the SMEs forming the alliance are supported by different PEFs. She thus implicitly contrasts the case of the formation of intraportfolio alliances with that of the formation of inter-portfolio alliances involving different PEFs. Our case of extraportfolio alliances may include this situation, but also includes all the situations where one company is supported by private equity and the other not. The case of extraportfolio alliances thus includes more possibilities for forming alliances than the case of inter-portfolio alliances in Lindsey’s study, which may be an explanation. Second, Lindsey’s results are based on a calculation of the probability of alliance formation including data stemming from a “theoretical” calculation, only partly based on verifiable empirical data. We might then also raise the question of whether her fictional sample of opportunities for alliances reflects reality. Finally, note that our study is the only one in the literature on the subject to consider non-formalized alliances, often of the intraportfolio type. The studies relate, in fact, either to the American market or to the Italian market, which offer a field of analysis allowing the researcher to access data published in databases. The latter do not necessarily register all alliances formed, but often selectively those for which written records exist (whether contracts or press articles, alliances containing at least one listed company or whose exit was made onto the stock exchange, etc.). Divergence in results may also arise from the fact that the analyzed alliances are not all of the same nature according to the mode of access to data.

Results for the mixed character (contractual and knowledge-based) hypotheses

Two of our hypotheses are of mixed character, thus contrasting a contractual argument to a knowledge-based argument. These are Hypotheses 5a/5b and 8a/8b. In both cases, the results of our study are in favor of the knowledge-based argument.

Hypotheses 5a and 5b state that in the presence of a PEF, the number of alliances formed by a company affects the formation of alliances in a negative (contractual argument) or positive manner, respectively. We were only able to test the hypothesis through the multiple case study. The negative effect is the result of potential conflicts between the PEF and the alliance partner. Although this negative effect is plausible in itself, the multiple case study shows that the presumed positive effect clearly prevails. Rather, previous alliances have a positive effect on the formation of a new alliance by the SME supported by private equity, because it gains in experience, in visibility and in awareness. This result differs from those present in the literature, on at least two levels. First, we find no confirmation of an actual negative effect, although the effect seems plausible. Ozmel et al. [OZM 13] among others, however, analyze this effect in the specific area of biotechnology SMEs, of venture capital and of alliances between start-ups and pharmaceutical companies. Because of the context of their study, it may be that this result pertains to the alliance partners of the company supported by venture capital which, like the venture-capital company, brings a contribution of capital to the start-up while taking a shareholding in the start-ups in the context of the alliance. Decision-making rights may then be shared in a different way compared to the case of inter-company alliances, which are the object of our study, and where there is in principle no financing of the start-up by their alliance partner. Second, we present an alternative outcome, following a knowledge-based argument. We show, in fact, that the number of alliances previously formed by a start-up may have a positive effect on the formation of new alliances.

Hypothesis 8 concerns the role of PEFs in the facilitation of exchanges between potential alliance partners. The mobilization of the concept of social capital enables us to argue that PEFs can only ensure this role if they have the time and attention necessary to offer this service, which does not represent their main activity. Thus, through Hypotheses 8a and 8b we test the negative (positive) impact of a large (small) number of portfolio companies to be managed (lack of time/non-lack of time) and of a large geographical distance separating the PEF from their supported SMEs on the formation of alliances. In addition, we empirically test the purely knowledge-based argument, linked to a high number of portfolio companies to be managed by each director of investment, predicting a positive role in the formation of alliances. The argument is that the more the portfolio manager monitors SMEs, the more this may be a source of ideas for establishing alliances. This positive effect contrasts with the negative effect assumed to be associated with a lack of time.

All these hypotheses are new with respect to the existing literature which directly relates to our research question. The econometric study reveals a non-significant impact for the number of portfolio companies to be managed by each investment manager on the formation of alliances. As for geographical distance, this could not be tested through the allin and allex regressions. In contrast, the multiple case study allowed us to empirically study the different hypotheses. As for geographical distance, we found that it impacts little on the formation of alliances. New technologies these days allow the facilitation of exchanges between actors, which means that physical presence is less important and geographical distance does not seem to have a negative effect. In contrast, once knowledge-based dimensions are considered, geographic proximity seems to have a positive effect on the formation of alliances. For the number of portfolio companies to be managed by each investment manager, the positive, knowledge-based effect – the greater the number, the more potential new ideas – clearly prevails over the effect of lack of time. In a more general sense, apart from the effect of these variables (geographical distance and number of portfolio companies), the role of PEFs as facilitators of the first exchanges between prospective alliance partners is confirmed by the multiple case study.

Results for the knowledge-based hypotheses

Knowledge-based arguments based on the strict mobilization of knowledge-based theories (excluding social capital) have allowed us to form two main hypotheses. They propose that the alliances that PEFs enable to be formed are dependent, in part, on their own expertise and competences (Hypothesis 7). From the point of view of the PEF, the formation of alliances may be pursued in the context of a strategy of differentiation on the private equity market (Hypothesis 9). Hypothesis 11 enables us, by resorting to the mobilization of the concept of social capital in a knowledge-based perspective, to consider that PEFs may have privileged access to the information and competences held by other actors which may be sources of ideas for alliances.

The results linked to Hypotheses 7a and 7b propose that the type of alliance that PEFs enable their portfolio companies to form depend in part on their own competences. The multiple case study corroborates the hypotheses according to which sectoral PEFs tend to intervene more frequently in the formation of intra or inter-sectoral alliances, while cross-border PEFs tend to intervene more frequently in the formation of alliances of an international character (7b).

Hypothesis 9, of a link between the diversification strategy adopted by the PEF and the formation of alliances, is confirmed across the whole of the multimethod study. The statistical study reveals a significant positive impact on the formation of alliances, both intra- and extraportfolio. The multiple case study enables the confirmation of these results from the point of view of the PEFs. In contrast, the managers of SMEs do not perceive the relational service provided by the PEF as a way for them to diversify themselves on the private equity market. As mentioned previously, we think that this point of view shared by the SME managers results from the fact that the PEFs do not communicate enough about their strategy for the SME managers to be aware of it. This result appears to be new with respect to the literature linked to our research question. Previous studies have mainly analyzed the impact of alliances on the speed and the method of exit of SMEs supported by private equity. Wang et al. [WAN 12] argue that PEFs form alliances in an international manner for the companies they support so as to alleviate some of the environmental risks which their portfolio companies face. According to us, PEFs may provide this relational service of alliance formation in order to stand out on the private equity market. In a wider sense, this result is part of the literature related to diversification linked to the theory of resources. Founded on the analysis of Penrose [PEN 59], this literature provides the explanation that the orientation of the diversification strategy pursued by a company depends, in particular, on its resources and competences, and the opportunities that its environment offers [MAH 92, pp. 366–367]. The PEF’s resources and competences may thus constitute a driving force for the diversification strategy being pursued. In this sense, the service of alliance formation arises from the PEF’s competences as active investors and from their resources in terms of the contacts which they may provide to the companies they support.

Hypothesis 11, of a positive link between the PEF’s access to information of a strategic nature and the formation of alliances, is confirmed across the study. The impact of the two sources of information specified in Hypotheses 11a and 11b, nevertheless, seems to be weak. Thus, the number of seats held on strategic boards by the PEFs (Hypothesis 11a) is only corroborated as a trend in the multiple case study, and this variable appears non-significant in the econometric study. The participation of PEFs in associations (Hypothesis 11b) and its positive impact on the formation of alliances only seems weakly confirmed by the multiple case study. The hypothesis is not testable through the econometric study. The impact of these two channels of access to information appears weak, in particular for participation in associations compared to seats held on boards. We may thus be led to wonder about the existence of other privileged means of access to information or ideas for alliances. From our discussions in the context of the multiple case study, it appears that the number of files that an investment manager analyzes in the process of choosing companies to finance and support seems to be a more important source of ideas for alliances.

Does the role of PEFs in the formation of alliances vary according to the types of alliances?

In light of our research question, in particular linked to the initial statistical part of our study (presented in the introductory section of our work), we might expect an attempt at restatement of the results according to a typology of alliances. Works in the strategic literature on modes of coordination within alliances often result in such a presentation of the results [JOL 01, JAO 06]. In fact, we do have elements of categorization in the results of the statistical study. Why, then, not mobilize the categories of alliances proposed by our survey (based on the objective of the alliance or the type of formalization) to restate the results in this section?

The reason we have not proceeded in this way is strictly linked to the approach of our study, which is by nature hypothetico-deductive. The empirical component of our work had the goal of testing the hypotheses arising from the theoretical analysis. Now, these hypotheses did not predict different results depending on the types of alliances (based on the goal of the alliances, for example, or on their mode of formalization). Only two of our hypotheses proposed a link between the role of the PEF and a certain type of alliances: these are Hypothesis 10 (greater role of PEFs in the context of forming intraportfolio alliances) and Hypotheses 7a and 7b (link between the PEF’s expertise and the type of alliances formed [sectoral and regional/international]). Apart from these two hypotheses, the roles of PEFs proposed in the formation of alliances among the other hypotheses are independent of the type of alliance formed. Attempting to restate these results according to the types of alliances proposed in the descriptive section, in the introduction of the work, would thus not offer different information depending on the types of alliance and, as indicated, would not be coherent with our hypothetico-deductive approach.

The concept of our multimethod study, which is justified in the light of the goals of our study, also does not enable us to restate the results based on a typology other than that which contrasts intra- and extraportfolio alliances, which, again, is logical given that our theoretical section is based on this distinction. A restatement of the results based on a typology of alliances, such as that contained in the statistical study at the beginning of the work, would only be possible with rigor for the econometric study. On the contrast, it would not be so for the case study. The different cases were chosen in a theoretical and intentional matter in line with the theoretical section, in such a way as to ensure that our research variables were represented in at least two cases, as well as the opposite cases, to guarantee at best a literal replication. However, our theoretical section did not predict results based on types of alliances (except the intra-/extraportfolio alliance distinction). It is therefore to be expected that our case does not allow us to restate the results according, for example, to the types of alliances mentioned in the initial descriptive section. To do this, we would need at least one case per type of alliance, but they were not chosen with this in mind. As to the econometric study, even if this possibility exists in principle, this study was part of the multimethod study. To be able to restate the results of this multimethod study in a synthetic way, there needs to be equivalence, on the level of design, between its two components (the econometric study and the multiple case study), which again justifies the absence of the presentation of our results according to a certain typology of alliances in line with the descriptive section. We must also add that this would have led to adding an additional variable to the regression, which would only have increased the econometric problems we encountered.

We must also underline that our work differs from those cited above, which resulted in a typology of alliances, mainly on two points: these works do not adopt a hypothetico-deductive approach, but often an exploratory approach; and some of them were interested in modes of coordination during the life of the alliance. However, we focused our attention on the role of PEFs in the formation of alliances. Certainly, we think that the question of the roles of PEFs according to a typology of alliances (other than the distinction between intra-/extraportfolio alliances that we adopted in this work) is in principle much more relevant for studying actual interventions by PEFs during the lifetime of alliances, rather than during their formation. However, in the precise context of our research, where we are interested in the role of one particular actor - the PEF - in the formation of alliances, it is clear from many discussions which took place in the context of the multiple case study that PEFs play much more of a role in building the relationship between companies at the time of the formation of the alliance, than during the actual life of the alliance. In general, it seems that once the relationship has been initiated, the PEFs only intervene at the request of the managers of the SMEs in the alliance, in the goal of smoothing things over in case of slight misunderstandings. This point moreover confirms the interest of having centered our study on the analysis of PEFs in the formation phase of alliances.

Responses to questions raised

After having discussed the results one by one, we wish to reply more directly to the different questions that we have raised. We begin with the questions relating to the “why” and the “how” of the intervention of French PEFs in the formation of alliances for their portfolio companies, then we explain their impact on the associated creation of value, referring to the hypotheses and the mechanisms of causality that have been confirmed. We continue by responding to the two secondary questions raised:

• – Are the main arguments advanced in the literature valid in the French context?
• – What is the respective importance of the contractual and knowledge-based arguments?

Finally, we discuss the external validity of our results.

Most of all, the study enables us to conclude that French PEFs do play a role in the formation of strategic alliances for the companies they support. Now, we reply to the various questions raised.

Why do PEFs intervene in the formation of strategic alliances?

This question refers to the interest that PEFs have in intervening in the formation of alliances for their portfolio companies. The question has been analyzed according to the points of view of the PEFs and SMEs. For the PEFs, the argument seems mainly of a knowledge-based nature: the formation of alliances for their portfolio companies seems in particular to arise from a strategy of differentiation for the PEFs on the private equity market (Hypothesis 9). This hypothesis is confirmed from the point of view of the PEFs, both through the econometric study and through the multiple case study. In contrast, the case study indicates that the managers of the portfolio companies forming alliances do not perceive this fact. Although the hypothesis is verified by triangulation of methods, it is not verified by triangulation of points of view. However, it seems to us that in this case, the interpretation of the results does not seem to bring into question the acceptance of the hypothesis. The results obtained seem rather to indicate a lack of communication from the PEFs with respect to their portfolio companies concerning their differentiation strategies. According to the contractual argument, the presence of certain investors in the capital of the PEFs, such as the State or a region, seems to be able to encourage the activity of alliance formation by the PEF, only in the case where the PEF takes the legal form of a venture-capital joint-stock company. However, the impact is nevertheless low or even neutral (Hypothesis 3).

From the point of view of the SMEs, on the one hand, the PEFs help them in their development by detecting and allowing them to seize opportunities for growth through building relationships within alliances. On the other hand, in comparison with a situation of alliance formation without the presence of a PEF, this kind of company enables them to overcome difficulties in building relationships which they might encounter, mainly due to a lack of visibility or past performance (track record). They therefore enable them to reduce inefficiencies linked to the transaction.

How do PEFs intervene in the formation of strategic alliances?

Our modeling has allowed us to analyze two main channels of intervention for PEFs in the formation of alliances: a “passive” channel of intervention and an “active”, intentional channel. Here, we present the main results of this, considering the variables and mechanisms of causality that have been confirmed by the study.

The first channel of intervention, through contractual variables, indicates that PEFs are able to reduce the inefficiencies that may typically be encountered by companies supported by private equity in the formation of alliances. In light of the different hypotheses raised and the variables that have been confirmed, PEFs seem above all to play the role of a mechanism enabling the establishment of a situation of trust between prospective alliance partners. Before the alliance is formed, this situation mainly enables the reduction of potential distrust by prospective alliance partners toward the company supported by private equity that, by its nature, is less visible on the markets. In the context of extraportfolio alliances, the reputation of the PEF enables facilitation of the establishment of such a situation of trust (Hypothesis 1 which was confirmed). After the alliance is formed, the PEF may intervene as advisors (a “weakened” disciplinary mechanism) to grease the wheels in case of difficulties between alliance partners (Hypothesis 4, whose mechanism was confirmed but not the variable that was supposed to measure this effect. The variable “taking majority stakes” does not seem decisive for PEFs to play this role, as was indicated to us by all the interviewees participating in the multiple case study.)

Through the second lever of analysis, the hypotheses arising from the application of knowledge-based theories to our problem indicate the channels of intervention for PEFs in alliance formations: the PEF may in particular be the source of the idea of the alliance. This idea may arise from privileged access to information, particularly of the strategic type. The seats that they hold on strategic boards (board of directors, supervisory board) constitute such a source (Hypothesis 11), as do the number of portfolio companies being supported (Hypothesis 8). The latter entails that the PEF will have seen a certain number of files and issues that may serve as the basis of ideas for alliances. Whether the actual idea for the alliance comes from the PEF or from the SMEs, the PEF may help its portfolio company to realize the project. It may thus contribute to detecting opportunities for forming alliances. The type of opportunities that it enables to be detected seems to be, at least in part, dependent on its own skills and experiences (Hypothesis 7). Thus, a PEF specializing in investments at regional/national or sectoral level is thus more likely to build relationships for its portfolio companies within alliances of regional/national scope, or with partners coming from the sector of specialization in question. PEFs of international scope may facilitate access to international alliance partners. This fact, however, does not prevent these PEFs from building relationships for their supported companies within alliances with partners outside of their areas of specialization. Finally, PEFs may facilitate the first exchanges between prospective alliance partners, while assuring cognitive coordination and thus reducing the risks of mutual misunderstandings (Hypothesis 8).

Altogether, PEFs intervene both in the formation of intraportfolio and extraportfolio types of alliances. We cannot conclude a greater formation of one of these types of alliance in the presence of a PEF. In contrast, although PEFs play the roles mentioned in both cases, they seem more intensive in the formation of intraportfolio alliances (Hypothesis 10).

What is the impact of PEFs on the creation of value resulting from the transaction?

The analysis was carried out in the framework of the efficiency paradigm, in the light of shareholder value creation through the mobilization of both contractual and knowledge-based theories. Thus, our analysis also enables us to reach a conclusion as to the impact of PEFs on the creation of value for the companies they support through alliance formation. We have already partially answered this question through the previous two. PEFs intervene in the creation of value that results from alliance formation thanks to two main levers: the creation of value itself and the reduction of losses of value. On the one hand, they directly allow the creation of value by detecting opportunities for growth or possibilities for generation of new knowledge through the creation of alliances. On the other hand, they reduce the inefficiencies, and thus losses of value, which might prevent such a transaction by reducing the costs linked to it.

Secondary goals of the empirical study

The empirical study allows us to provide answers to two other questions that we raised. They are the following questions:

• – Do the main arguments of the literature hold in the French context?
• – What is the respective importance of the contractual and knowledge-based arguments?

Concerning the literature (question 1), we find that five hypothesis which are usually presented are not confirmed, or are only partly confirmed, in the French context. First, our results indicate that previous alliances have no negative impact on the formation of an additional alliance for a start-up. On the contrary, they tend to have a positive impact (Hypothesis 5b). The assertion that PEFs with weak reputational capital would put their portfolio companies more into alliances with well-known actors on the markets (Hypothesis 2) is rejected. The fact that the PEFs’ roles are more pronounced in the formation of alliances between unlisted companies is only verified in that, by nature, PEFs mainly support unlisted companies (Hypothesis 6). Finally, the hypothesis of greater roles for PEFs in the formation of alliances when it comes to intraportfolio alliances (Hypothesis 10) is only partly confirmed. French PEFs are no more involved in the formation of intraportfolio than extraportfolio alliances. In contrast, certain roles may be more intense in the case of forming intraportfolio alliances. Geographical distance does not appear to have a decisive effect on the formation of alliances (Hypothesis 8b). Furthermore, although we did not explicitly test this hypothesis, responses to the question of whether the formation of alliances constitutes a substitute for the capital brought by PEFs all rejected this hypothesis.

Concerning the relative importance of contractual and knowledge-based arguments, we notice that both theoretical frameworks have their place in explaining the phenomenon. However, we also notice that across all the hypotheses raised, all the knowledge-based hypotheses are validated, while for the contractual hypotheses the validation was only partial. Some hypotheses were even rejected. For the hypotheses including a contractual and a knowledge-based variable at the same time, the knowledge-based variant prevails. Knowledge-based argumentation therefore prevails over contractual argumentation, even if the latter remains relevant. In summary, these results support the complementary use of the two frameworks in the explanation of the phenomenon. In our case, contractual roles appear to be at the service of knowledge-based roles. In other words, the roles promoted by the contractual argument enable the establishment of the necessary conditions for the proper conduct of the transaction from a knowledge-based point of view, which confirms to the way in which we have conjointly used the two theoretical frameworks. This result supports the point of view of Cohendet and Llenera [COH 05, p. 181]. The authors argue that, unlike companies operating in environments of certainty, those acting in uncertain environments are led to develop capabilities to innovate. The knowledge-based dimension is therefore more valuable than the contractual dimension.

Are these results generalizable?

We have already discussed the limits of the econometric study and, consequently, the statistical generalization of the results. As our main results arise from the case study, the question thus arises of its external validity. In other terms, can we conclude an analytical generalization of our results?

For the hypotheses which were confirmed, the theoretical arguments underpinning the mechanisms of causality that arose from the application of contractual and knowledge-based theories seem to be confirmed in the context of our study. Through the combination of various cases and the econometric study, our conclusions are analytically generalizable across all alliances formed in analogous situations and contexts (in the presence of SMEs supported by at least one PEF in the French context). More generally, our results add to the catalog of results available concerning the roles of PEFs in the formation of alliances and confirm that PEFs play a role. The combination of our results with those of the literature allows us to conclude on a generalization of the roles of PEFs in the formation of alliances beyond the specific context of the French private equity sector. For the hypotheses common to other studies, which were confirmed through our own, we may conclude on an analytical generalization of the phenomenon, beyond the strict context of our study, the French private equity market. However, our study also shows that certain roles proposed by the literature do not appear in the specific context of the French private equity market.

On a more general level, our study contributes to the studies advocating a complementary utilization of contractual and knowledge-based theories. It is thus in combination with other studies aiming at a complementary application of these theoretical frameworks that a process of analytical generalization of the complementary application of theories is progressively carried out [DAV 05]. Our study is one element in this process.

1 Modality 3 occupies the fourth position in the schema because modality 0 = not relevant. The positions of the modalities are as follows: 0 = not relevant (position 1); 1 = totally negative modality (position 2); 2 = rather negative modality (position 3); 3 = neutral modality (position 4); 4 = rather positive modality (position 5); 5 = totally positive modality (position 6).

2 www.ats.ucla.edu/stat/stata/dae/intreg.htm.

3 In Stata, this can be done as follows (example given for allin): “replace allin = 3 if allin == 4”.

4 The regroupings were again obtained in the following way in Stata: taking the variable Rp, simply indicate replace Rp = 2 if Rp == 1 for the negative modalities and replace Rp = 4 if Rp == 5 for the positive modalities.

5 Ordered logit and probit treat our dependent variable as an ordinal with several modalities, while intreg allows us to treat the variable as basically quantitative (a percentage varying naturally from 0% to 100% in our case), but whose observations can only be provided by intervals.

6 For each interval, we thus define a “lower limit” y (y_low) and an “upper limit” y (y_up). For example, for allin, this gives (before having decided to consolidate intervals 3 and 4):

• gen y_low = 0
• replace y_low = 1 if allin == 1
• replace y_low = 11 if allin == 2
• replace y_low = 51 if allin == 3
• gen y_up = 0
• replace y_up = 10 if allin == 1
• replace y_up = 50 if allin == 2
• replace y_up = 100 if allin == 3

We proceed in the same way to define the lower and upper limits of the dependent variable (allex) for “extra” alliances.

7 In Stata, this is done by means of the command estat vif.

8 In Stata, the test is carried out with the order sktest. It is necessary to generate the residuals variable before being able to carry out the test. This is done with the order predict residus, stdp.

9 Order hettest in Stata.

10 In Stata this gives: intreg y_low y_up ib3.Rp ib2.Ple ib2.Etat ib2.maj ib3.Cot ib0.Spsec ib3.rgInt i.Diff ib3.Nbrepart ib3.Ca ib3.Spfin ib2.indus ib2.finan ib3.Exp ib3.Nbresyn.

11 The reference base (here “ib followed by a number”) is the modality that the software uses as a reference to calculate the coefficients of the other modalities.

12 To recall, the meaning of the categories is as follows: 0 = not relevant; 1 = totally negative; 2 = rather negative; 3 = neutral; 4 = rather positive; 5 = totally positive. After our first consolidations (of the negative modalities into a single category, and of the positive modalities also into a single category), there are then only four categories: 0 = not relevant; 2 = negative; 3 = neutral; 4 = positive.

13 The table indicates for each modality, taken separately, if it is significant or not. The table does not indicate whether the variable, through all of the modalities which comprise it, is significant as a whole.

14 In Stata, this is done using the commands testparm or contrast.

15 In Stata this gives: intreg y_low y_up ib3.Rp ib2.Ple ib2.Etat ib2.maj ib3.Cot ib0.Spsec ib3.rgInt i.Diff ib3.Nbrepart ib3.Ca ib3.Spfin ib2.indus ib2.finan ib3.Exp ib2.Nbresyn.

16 For a discussion of the use of case studies with an explanatory design, see, in particular, Koenig [KOE 09a, KOE 09b] and Scapens [SCA 90, p. 265, pp. 270–272], as well as more generally Yin [YIN 09] and Miles and Huberman [MIL 03, p. 84].

17 www.siparex.com.

18 www.ifpart.com.

19 www.caterine.fr/.