2
The Role of Private Equity Firms in Alliance Formation from the Perspective of Value Creation

In this chapter, we refer to three theoretical frameworks in order to analyze our problem and then we apply the theory to the French market in Chapter 3. We begin our analysis with the use of contractual theories (section 2.1), and then discuss knowledge-based theories (section 2.2). The final section will turn to sociological network theories (section 2.3) with the sole aim of completing the contractual and knowledge-based argumentation.

The contractual theories we allude to include the transaction cost theory (TCT) and the positive agency theory (PAT). The so-called “knowledge-based” theories are based on strategic or heterodox economic streams. They include the resource-based view (RBV), competence-based view (CBV), evolutionary economic theory and behavioral theory of the firm. They are sometimes grouped under the term knowledge-based view (KBV). These two theoretical frameworks are part of the efficiency paradigm that allows value creation to be analyzed [CHA 06]. In light of our problem, they make it possible to analyze the role of a PEF in improving (or deteriorating) the efficiency of alliances formed, as well as the impact of a PEF on the alliance’s value creation. An organizational system (including the alliances in which we are interested here) is considered to be efficient if it maximizes organizational rent (or surplus) and if, on average, there is no alternative mechanism to achieve better results for all stakeholders [MIL 92]. This efficiency is static in design within contractual theories. The goal is to maximize the organizational rent at a given time [CHA 06]. These theories do not concern the creation of value itself but rather the limitation of value losses, at a given point in time, that result from the presence of transaction and agency costs [LAN 99, pp. 201–202]. Knowledge-based theories are based on a dynamic notion of efficiency [CHA 06]. They question the productive origin of long-term value creation. Elements that are derived from sociological network theories, but which are compatible with the efficiency paradigm (especially emanating from the notion of social capital), complete the arguments of the two frameworks.

In each of the three parts of our theoretical analysis, we have included a section that begins with a brief introduction to the main points of the theory that are relevant to our particular problem. We then apply the theory to our topic. This includes a systematic analysis from the perspective of both the SMEs that form an alliance and the PEF. More precisely, the features of SMEs supported by PE are put forward in light of key concepts that are associated with the various theoretical frameworks. They form the basis for analyzing the difficulties that SMEs may face in forming alliances. The question then arises as to whether the presence of a PEF makes it possible to overcome these obstacles. Subsequently, we consider the benefit for PEFs in forming alliances for the companies they support.

2.1. The role of PEFs from the perspective of contractual theories

In this section, we intend to shed light on our problem by turning to the theoretical streams that make up the contractual theories of organizations (hereinafter denoted TCO): TCT [COA 37; WIL 91a] and PAT [JEN 76]. They appear because of the inability of traditional neoclassical theories to explain organizational phenomena. These theories therefore make it possible to, on the one hand, understand strategic alliances and, on the other hand, analyze the role played by PEFs.

We begin by introducing some theoretical bases that will be useful for exploring our subsequent problem (section 2.1.1). We then give an overview in section 2.1.2. In section 2.13, we apply the theory to our problem. We conclude with a review of the role of PEFs in alliance formation from the perspective of contractual theories in section 2.1.4.

2.1.1. Some theoretical foundations

Contractual theories include the TCT [WIL 91a] and the PAT, which is itself based on the property rights theory [ALC 72] and on the agency relationship, which arises from the principal–agent approach, and thus from the normative agency theory. According to these theories, different organizational forms – including alliances – are defined as nexus of contracts. They are said to be efficient if they minimize transaction and agency costs at any given time. Transaction costs result from the costs of finding the counterparty to the transaction, from information costs in the ex ante phase to the establishment of the transaction and from the costs of negotiating, deciding, monitoring and executing contracts in the ex post phase to the transaction itself. Agency costs arise from conflicts of interest between the agents involved in a transaction.

Let us begin by looking at the contributions brought about by TCT (section 2.1.1.1), followed by those from PAT (section 2.1.1.2).

2.1.1.1. Contributions of TCT to contractual theories

The initial questioning of TCT arises from the existence of the firm in relation to the market. In neoclassical theory, the only mode of coordination considered is the market through prices. This raises the question of whether a firm’s existence is justified, which was initially asked by Coase [COA 37]. If a company exists, it is because there are imperfections in the markets, which creates costs. If the actors require coordination via hierarchical structures within a company, it is because this makes it possible to reduce the costs linked to market inefficiencies. The use of the market or a firm as a coordination method therefore results from an arbitration between the associated (transaction and production) costs.

Williamson extends and generalizes Coase’s analysis by constructing a general theory of markets and organizational forms: TCT. There are three key notions within this theory: transaction as a unit of analysis [WIL 91a, p. 79], the limited rationality of agents and the (potential) opportunism of actors. Limited rationality refers to the fact that individuals make decisions in a computational manner (in other words, they evaluate the consequences of different opportunities presented to them with reference to a set objective). But individuals may not seize all the alternatives presented to them and may also make evaluation errors on the consequences of choosing one of these alternatives, or may even make reasoning errors.

Since the unit of analysis within the theory is the transaction, turning to the market or a firm depends on the respective transaction costs generated by the two modes of coordination [WIL 91b, p. 269]. By defining a transaction as the transfer of the right to use an asset (good or service), one can distinguish between the costs of finding the counterparty and providing information (ex ante to the transaction) and the costs of negotiating, deciding, monitoring and executing contracts (ex post to the transaction).

In order to make the analysis functional, Williamson defined three critical dimensions of the transaction as follows [WIL 79, p. 239]:

• – asset specificity: an asset is specific if it cannot be redeployed, in other words if it loses value by being reassigned to other uses. An asset will be highly specific if it is idiosyncratic to the transaction. It is the source of a quasi-rent [WIL 79, p. 241];
• – the frequency and duration of transactions;
• – the degree of uncertainty: the greater the uncertainty, the more difficult it will be to predict and determine all possible outcomes. Drafting full contracts becomes difficult and costly.

Due to the uncertainty and potential opportunistic nature of the actors, the contracting parties to the transaction run the risk of being ripped off. This risk of being ripped off increases with specificity of assets. Since specific assets lose their value if they are reassigned to uses other than the original transaction, the seller of the asset in question can act opportunistically and increase the price of the asset once an agreement is reached. This postcontractual risk of non-performance is known as moral hazard, if it is intentional [BRO 93, p. 14]. In the precontractual phase, a contract may not be concluded because of the risk of adverse selection [AKE 70], in other words, mistrust about the quality of the potential transaction partner or the object of the transaction. The uncertainty and the limited rationality of agents make it impossible to resolve contracts in such a way as to define all possible contingencies ex ante. In order to insure against these risks and limit the associated costs, one solution is to internalize the transaction. Having to turn to the market or hierarchy then results from a production and transaction costs’ optimization problem, thus minimizing the sum [WIL 79, p. 245]. Schematically, having to turn to the hierarchy as a mode of governance (internalization) is justified for transactions that involve highly specific assets and frequent and sustainable (and therefore long-term) transactions. Conversely, turning to the market (total outsourcing) is justified for short-term, infrequent transactions that involve assets that are not very specific.

Following this reasoning, strategic alliances can be described as a hybrid mode of coordination, falling somewhere between the two extremes of market and enterprise (hierarchy) [WIL 91b]. This hybrid contract covers recurring and long-term transactions including average to highly specific assets. Due to its average to highly specific nature, turning to the market can be risky and therefore costly. On the other hand, the problem with opportunism of actors does not justify the costs of internalizing the transaction. Alliances therefore make it possible to reduce transaction costs and uncertainty while allowing companies to remain autonomous. Consequently, the contract requires a special governance mechanism [WIL 91b, p. 271]. In what follows, we discuss the role that a PEF can play.

TCT provides a preliminary justification for the existence of organizational structures. It makes it possible to gain an initial approach to alliances and then to study the role of PEFs in alliance formation. In order to take the analysis further, let us present the PAT. This theory focuses on agency relations between agents, whereas the TCT uses the transaction as the unit of analysis. Unlike in TCT, organizational structures, including alliances, are no longer seen as a black box. PAT makes it possible to analyze internal organizational structures and account for conflicts of interest between contracting parties. It makes it possible to explicitly consider the preferences or attitudes of the contracting parties toward the transaction.

2.1.1.2. Contributions of the agency theory to contractual theories

Our introduction to the contributions of the PAT is split into four sections. Section 2.1.1.2.1 presents concepts from Berle and Means, The Modern Corporation and Private Property (1932; [BER 91]). These authors analyzed the managerial enterprise, which thrived in the United States, and they highlighted the problem of emergence of conflicts of interest that resulted from a dismemberment of ownership. Section 2.1.1.2.2 highlights key concepts from Alchian and Demsetz [ALC 72]. Starting from the definition of a company as a nexus of contracts, they see companies as production teams and thus deliver, beyond the justification of their existence, an analysis to explain the different forms of organizational structures. In section 2.1.1.2.3, we look at key concepts from Jensen and Meckling [MEC 76]. Starting from an entrepreneurial company and combining elements of agency theory, property rights theory and financial theory, they attempt to develop a positive theory of the ownership structure of the firm. Section 2.1.1.2.4 briefly discusses a firm’s organizational architecture that is made possible by the agency theory.

2.1.1.2.1. Analysis of Berle and Means [BER 91]

For a company in its simplest form – an individual entrepreneur – the person who provides the capital and assumes the financial risk of the investment also has residual decision-making power. That is, he or she has decision-making authority over any matter that is not governed by contracts or by law [CHA 02a]. This same person therefore manages the company both at the operational level (providing, coordinating and exploiting production factors) and at the strategic level (identifying, even creating and implementing opportunities). This person then simultaneously fulfills three functions [CHA 02a]:

• – the risk and uncertainty assumption function;
• – the company’s leadership or management function (providing, coordinating and exploiting production factors);
• – the function of identification, even creation and implementation of growth opportunities.

In their book The Modern Corporation and Private Property, Berle and Means [BER 91] discussed the distribution of ownership within a managerial type of enterprise, which was developed in the 20th Century mainly in the United States. Contrary to sole proprietorship, ownership is so dispersed that no single owner has sufficient interests to appropriate residual decision-making power, and therefore management of the business [BER 91, p. 78]. For example, in some cases, the majority shareholder holds less than 1% of the capital [BER 91, p. 47]. The management of the company is then delegated to a group of people: the management.

A managerial type of company therefore features, on the one hand, a dispersed ownership structure such that no single shareholder holds a significant share of the capital. On the other hand, a restricted group of people – management – has the power to take decisions and manage the company [BER 91, in particular p. XXIII, p. 5, pp. 83–84, pp. 244–246, p. 297, p. 300 sq.]. Usually, this management holds little or no capital [BER 91, p. 53, p. 83].

In this type of company, the three functions mentioned above in the example of the individual enterprise are then separated. They are divided between shareholders and management. There is a dismemberment of “passive” and “active” ownership [BER 91, p. IX, p. XXIII, p. XXXV, p. 9, p. 244, p. 297, p. 300, p. 304]. Passive ownership is held by shareholders who have the sole function of providing capital and assuming risk and uncertainty. They become mere providers of capital [BER 91, pp. 245–246]. Active ownership belongs to management, which assumes the functions of leadership as well as the identification, creation and implementation of opportunities. It holds the residual decision-making rights and manages the company at operational and strategic level.

Berle and Means were particularly interested in this type of company because it appeared to be the dominant organizational structure of the 20th Century in the United States. On an economic level, the emergence of this managerial type of company gave rise to a concentration of economic power that – at least in the United States – could compete with the political power of the State [BER 91, p. 309, p. 313]. Indeed, it placed the savings of countless individuals under the centralized control of a limited number of people: management [BER 91, p. 5]. As a result, these companies required the interconnection of a wide variety of interests [BER 91, p. 310]. The depersonalization of ownership, which landed into the hands of multiple actors, and which was accompanied by the dismemberment of its functions, thus led Berle and Means to assimilate this managerial type of enterprise to an institution for which the characteristics resemble those of the State [BER 91, p. XVII, p. XXXVIII, p. 5, p. 309]. Consequently, the authors believed that this type of enterprise must be considered not as a company but as a social organization. As a result, they concluded that the consequences of decisions taken within these companies must be taken into account, not just for the owners (shareholders that have passive ownership and management that have active ownership), but for a broader category of actors including employees, customers and suppliers, and even the interests of society as a whole [BER 91, pp. 312–313]. Their approach is therefore stakeholder oriented.

The question then raised is who has the status of residual claimant and, consequently, the right to profit [BER 91, p. 293]. According to Berle and Means, profit, as a counterpart to performance, fulfills two incentive functions [BER 91, p. 300]:

• – it encourages individuals to take risks by placing their savings within the company;
• – it encourages the individual who holds the residual decision-making rights to do everything possible to make the company profitable.

Profit therefore rewards both passive and active ownership.

This distinction between the two incentive profit functions is not important in sole proprietor companies. However, it becomes significant if ownership is divided into its passive and active components, with each component being in the hands of different actors. Berle and Means concluded that, on the one hand, from an economic point of view, the status of residual claimant must not go to passive owners (shareholders). Shareholders should receive sufficient remuneration for them to keep (and continue to have) an interest in risking their savings for the company. The authors compared shareholder remuneration logic to employee remuneration logic, the latter of which must be paid in such a way that the employee is willing to provide (and to continue to provide) his or her labor. Berle and Means perceived no social efficiency benefit in allocating the organizational rent (surplus) to the shareholder since the latter has renounced active ownership – the residual decision-making power – and thus any responsibility in managing affairs [BER 91, p. 301]. On the other hand, they did not conclude that the profit should only go to active owners, in other words to management [BER 91, p. 312]. Two justifications are put forward. First, the separation of passive and active ownership creates the problem that management, having a discretionary space, might not act in line with the interests of passive owners (shareholders) [BER 91, pp. XIV–XV, pp. 113–114]. If management was granted residual claimant status and given the organizational rent, it would then have no incentive (unless it was intrinsic in nature) to guide the company’s actions to be in line with the interests of passive owners. Second, the profit generated by a company (at least in the United States) does not solely result from transactions that are initiated by active ownership and financed by passive ownership. It also comes, in part, from the company’s market position and from subsidies financed by taxpayers through the State. In this sense, the U.S. state is an investor in nearly every U.S. company. There is growing recognition that, at least for very large companies, their actions can be defined as collective transactions that are similar to transactions done by the State [BER 91, p. XXXVIII]. There is therefore no justification for granting shareholders residual claimant status [BER 91, p. XXVII].

However, the authors did point out that this conclusion needs to be taken with a pinch of salt since the separation of active and passive ownership exists to varying degrees [BER 91, p. 5, p. 6]. In reality, and particularly in European countries including France (the country of interest in this study), this separation is rarely complete. It is even the exception [CHA 06]. In these cases, shareholders also take on part of the active ownership and may be entitled to part of the profit.

In our problem, PEFs bring a lot of capital to the companies they support. However, in addition to their financial contribution, they possess skills from which their portfolio companies can benefit. PEFs are active shareholders, which is one of their key features, as we presented in our introduction to PE.

In order to resolve conflicts of interest between shareholders and management, Berle and Means proposed to implement incentive solutions [BER 91, pp. XII–XIII]. Following this realization, a whole stream developed in the literature which, from a normative perspective and based on the notion of agency relationships, sought to find solutions that encourage management (active ownership) to act in the interests of shareholders (passive ownership). Later, notably after the articles published by Alchian and Demsetz [ALC 72] and, more particularly, the article by Jensen and Meckling [JEN 76], a positive perspective developed, which sought to explain why there were different organizational configurations [JEN 76, p. 310; JEN 83, p. 334; CHA 00]. Adopting a positive approach, we are part of this second perspective. Let us now present the two key articles.

2.1.1.2.2. The firm as seen by Alchian and Demsetz [ALC 72]

For Alchian and Demsetz [ALC 72, p. 777], a theory of economic organization must be able to solve two questions:

• – it must highlight the factors that determine when hierarchical coordination (the firm) is superior to market coordination and vice versa;
• – it must explain the structure of firms and their organizational architecture.

Starting from the view of a firm as a nexus of contracts [ALC 72, p. 778], the authors highlighted two determinants [ALC 72, p. 778, p. 783]:

• – the team use of production factors (inputs) held by different agents;
• – the management of this team by an agent with a central position within the nexus of contracts.

A company is therefore characterized by its management of production factors, both in teams and centralized. This joint management allows the company to benefit from synergies resulting from group work. The centralization of management must ensure effective teamwork. Alchian and Demsetz [ALC 72, p. 783] specified and insisted on the fact that this centralized management has the vocation of a common management, per team, and not an authoritarian management that allows the central agent to exercise disciplinary power. Each team member’s relationship with the central agent is reflected in a simple contract that is based on reciprocity (“quid pro quo”, which may consist of an exchange of input such as labor or capital contribution for remuneration. This could be linked or not, for example, to the performance or output of team work).

The role of the central agent is justified by the fact that team management, on the one hand, has the consequence of being able to benefit from synergies and, on the other hand, raises the problem of free riding. Because of the difficulties in observability of individual contributions of the different inputs to output and measurability of performance, the agents involved may, if they have an opportunistic character, seek to take advantage of this situation by benefiting from the work provided by their teammates. They would then appropriate all of the gains from their behavior, while the related costs (a reduction in production) are borne by all of the members of the team [ALC 72, p. 780]. Team production then poses the problem of how to encourage team members such that all agents work efficiently [ALC 72, p. 779]. According to Alchian and Demsetz [ALC 72, p. 783], the most effective organizational structure is one that allows the central agent to fulfill the function of a “typical” arbitrator, solving to the best of their ability the problems related to group production. Such an arbitrator should have the following rights:

• – the right to be the residual claimant that receives the remaining organizational surplus after remuneration of other production factors (inputs);
• – the right to observe the contribution (input) of actors;
• – the right to be the central intermediary, common to all actors;
• – the right to elect or remove parties (team members);
• – the right to transfer these rights.

As with Berle and Means [BER 91], organizational rent does not necessarily belong to the shareholder. According to the analysis proposed by Alchian and Demsetz, it belongs to the agent who manages the company’s production factors. The manager, or active owner according to the distinction by Berle and Means [BER 91], is then assigned the status of residual claimant to whom the profit accrues. If this person is paid once all of the other agents who are contractually bound to him have been paid, this solution would constitute an incentive for the central agent to effectively manage the productivity of the team [ALC 72, p. 785].

Thus, according to this view, strategic alliances, which until now were defined as hybrid contracts between the hierarchy and the market, constitute a cooperative relationship between at least two companies. This cooperation makes it possible to achieve synergy effects due to joint production [ALC 72, p. 779]. If a PEF is a common actor to all the alliance partners, it would be useful to determine whether it can possibly play the role of central actor. We discuss this in the section on applying the theory to our research question. Before that, let us present some contributions from the article by Jensen and Meckling [JEN 76].

2.1.1.2.3. Some key points from the article by Jensen and Meckling [JEN 76]

Jensen and Meckling [JEN 76] proposed a positive theory to explain the financing structure of firms. They transposed Berle and Means’ analysis of the managerial enterprise, and the agency problem that arises there, to an entrepreneurial firm in order to analyze, through a positive approach, the incentive consequences that result from the various forms of external financing that the manager and owner of a company uses in order to make its development possible.

They begin with an entrepreneurial firm where a person who owns 100% of the structure opens his or her capital to raise funds. This person chooses external equity or debt financing. Jensen and Meckling analyzed this situation from the perspective of the PAT. In both cases, the result is an agency relationship between the entrepreneur–manager (agent) and the new shareholders or creditors (principals).

According to Jensen and Meckling [JEN 76, p. 308], an agency relationship is “a contract under which one or more persons (the principal(s)) engage another person (the agent) to perform some service on their behalf which involves delegating some decision-making authority to the agent”. Due to differences of interest between the principal and the agent, agency conflicts can arise. Within contractual theories, these conflicts of interest can be resolved through contracts and incentive systems. However, a complete and inexpensive solution to the agency problem via contracts can only occur in an environment that is characterized by certainty. In the presence of uncertainty and information asymmetry, it is too costly (if not impossible) to determine all possible contingencies in advance, which makes contracts incomplete. Incomplete contracts are risky to make. Agents, who are supposedly of limited rationality and supposed to maximize their own wellbeing, may not perform a task delegated to them properly or may act strategically, even opportunistically, once the contract is concluded. In this case, they may try to take advantage of the situation and the flaws in the contract by exploiting the private information they hold in order to pursue their own goals. One solution to this problem may be the establishment of control mechanisms. But, because of the information asymmetry, control of agent behavior by the principal cannot be done without incurring costs.

The information asymmetry exists both before and after the contract is signed and distinguishes between pre- and postcontractual risks. In the first case, precontractual risks may prevent the formation of a contract even though it is advantageous for both parties, or it may lead to adverse selection. This adverse selection is linked to the fact that one of the contracting parties holds private information from which it can benefit because of the information asymmetry. In the postcontractual phase, a moral hazard problem sometimes arises, in other words agents may not respect their commitments. This can lead to the problem of free riding, which was mentioned in relation to the discussion on the work of Alchian and Demsetz [ALC 72].

The agency relationship thus generates agency costs in contrast with an ideal situation where there would be neither information asymmetry nor conflict of interest and thus an absence of costs. A comparison can be made between a reference situation, where a manager holds 100% of residual claims, and a situation where the owner opens his or her capital to external investors [JEN 76, p. 312]. In the first situation, the manager, who is supposed to maximize his utility, bears the totality of the consequences (monetary and non-monetary) of his choices. However, in the second situation, the manager is no longer the only one to hold the residual claims in the company. He then only bears a fraction of the costs associated with the consequences of decisions he makes to obtain non-monetary benefits. The investors, who are supposedly rational, anticipate the situation and propose a lower price than the intrinsic value of the title [JEN 76, p. 313; WIR 06]. This results in a loss of value that is rooted in the presence of agency costs related to the divergent interests of parties [JEN 76, p. 312; JEN 04, p. 21].

Agency costs can be broken down into three types of costs [JEN 76, p. 308, pp. 312–313]:

• – monitoring expenditures;
• – bonding expenditures;
• – residual loss.

The first two costs are explicit costs. Monitoring expenditures are costs that are incurred by the principal to induce the agent to act in the principal’s interest. This may involve the implementation of incentive or control systems, such as budget restrictions, operational rules or compensation policies. Bonding expenditures are linked to the self-discipline of agents. These are costs incurred by an agent in order to show the principal that if his actions go against his interests, he will be compensated. Residual loss are opportunity costs that arise from the fact that full protection of the principal’s interests, as guaranteed in an ideal situation and therefore in the absence of agency conflicts, cannot be ensured [JEN 76, p. 309]. They may be limited but cannot be entirely eliminated [JEN 76, p. 312].

In this context, value creation is achieved by minimizing all agency costs. This can be done through informal mechanisms, for example through trust, or formal mechanisms. In the latter case, the principal may put solutions in place to discipline the agent so that he acts in his interest.

Jensen and Meckling chose to illustrate their theory via agency costs that resulted from the relationship between shareholders and company management. However, their theory is not limited to this. Agency costs exist in any cooperative relationship. They do not necessarily require a subordinate relationship, as described by the principal–agent relationship [JEN 76, p. 309]. They can also arise in a dyadic relationship where both cooperation partners can simultaneously play the role of principal and agent, or in situations involving more than two actors. The different organizational structures are then justified if they are effective, in the sense that they make it possible to reduce conflicts of interest as much as possible, in other words, maintain the balance of interests.

The definition of strategic alliances, described so far as a cooperative relationship between at least two firms to achieve synergy effects due to joint production [ALC 72, p. 779], can therefore be supplemented. According to Jensen and Meckling [JEN 76], we can refine this description by stating that alliances constitute a dyadic relationship, where alliance partners both simultaneously play the role of principal and agent. This relationship is effective if it can reduce the losses in value that are associated with agency conflicts between contracting parties and the associated costs [JEN 76, p. 1992]. In light of our problem, of course, the question remains on the role of the presence of a PEF. This is discussed in the section on applying the theory to our research question (section 2.1.3).

Up to now, analyses have focused on relations between a company’s manager or management and its shareholders, and even between all of a company’s stakeholders. However, the PAT also allows agency relationships within the company itself to be considered. We then move from the field of governance to the field of organizational architecture.

2.1.1.2.4. The organizational architecture of a company

In defining a company as a nexus of contracts, its internal organization, i.e. its organizational architecture, can be defined as a set of contracts aimed at resolving conflicts resulting from the agency relationship and thus reducing the associated costs. These contracts are associated with three dimensions [JEN 92]: the allocation of decision-making rights, the evaluation of performance and the incentive systems to be designed such that they are all consistent.

The use of hierarchical coordination poses the problem of optimal allocation of decision-making rights between actors [JEN 92]. In order to allow good exploitation of knowledge, these are generally transferred to the agents that hold the specific knowledge [HAY 45]. This co-localization of specific knowledge and decision-making rights can be achieved in two ways: through the transfer of specific knowledge to the person holding the decision-making rights, or through the transfer of decision-making rights to the person holding the specific knowledge [JEN 92, p. 253]. The choice of allocation method depends on the transfer costs generated by the transfer of specific knowledge and decision-making rights [JEN 92, pp. 262–263]. However, the rights associated with the use of assets are generally not accompanied by the possibility of assigning those rights and appropriating the proceeds of that assignment [CHA 00, p. 199]. This lack of transferability means that agents are no longer encouraged to use their decision-making rights to act in the best interests of the organization [CHA 00, p. 199]. The delegation of decision-making rights then raises the problem of control of agents within the company [JEN 92, p. 251]. According to Fama and Jensen [FAM 83], delegation can then divide decision-making rights into two categories: rights related to decision-making and those related to control [JEN 92, p. 265]. The former include the initiative and implementation phases of a decision. The rights related to the control function include the ratification and monitoring phases [FAM 83, p. 303]. According to the theory of organizational architecture, the performance evaluation and the incentive system should be in line with the allocation of decision-making rights and in such a way that the costs of this arrangement do not exceed the efficiency gains.

2.1.2. Review of contractual theories

From a positive perspective, contractual theories include TCT and PAT. These are part of the efficiency paradigm. An organizational system (including the alliances we are interested in this study) is considered to be efficient if it maximizes the organizational rent (or surplus) and if there is no alternative mechanism to achieve, on average, better results for all stakeholders [MIL 92]. Within contractual theories, this efficiency is static in design. Agents are supposed to have limited computational rationality. This means that, when faced with a given set of opportunities known at an instant “t”, agents are supposed to be able to evaluate the consequences of their choices in probabilistic form but can commit computation or reasoning errors. Thus, we do not take into account a situation of radical uncertainty but rather a situation of weakened uncertainty, called risk, because it is possible to define this in probabilistic form.

Both theoretical frameworks seek to explain either the existence of a company or its internal configuration (its organizational architecture). According to the TCT, a company is justified because it makes it possible to reduce the costs that result from the imperfection of markets, which are characterized by the presence of information asymmetry, uncertainty and potential opportunism of actors. According to the Williamsonian analysis, hierarchical coordination (a company) can then reduce transaction costs in the presence of idiosyncratic assets and thus minimize value losses. According to the PAT, the unit of analysis is the agency relationship. This theory therefore makes it possible to consider the inside of a company, which was hitherto considered as a black box. Value creation is achieved by minimizing loss of value that results from agency costs due to conflicts of interest between the different agents involved in the transaction or in an agency relationship.

In both cases, a company is therefore efficient if, at a given moment, it can reduce losses in value as much as possible, thus getting as close as possible to the optimal situation in the absence of costs – also called the Nirvana economy [DEM 69] – or the first-order optimum (which is supposedly unachievable, meaning that one always strives for second-order optimum). The first-order optimum constitutes a reference situation or benchmark that cannot be reached, in theory [CHA 02b, pp. 19–20]. Ultimately, the focus is more on limiting value losses than on creating value [LAN 99, pp. 201–202]. Indeed, the question of the origin of given opportunities at a moment “t” is not asked. The whole point is to maximize the value created by acting on just one lever of action: the possibility of acting on the limitation of losses in value, linked to transaction and agency costs.

In Table 2.1, we list the key points.

Table 2.1. Contractual theories: key points

	Transaction cost theory	Positive agency theory
Core authors	– Coase [COA 37] – Williamson [WIL 75]	– Alchian and Demsetz [ALC 72] – Jensen and Meckling [JEN 76]
Angle and unit of analysis	– Microeconomic approach – Unit of analysis: the transaction, which has three characteristics: - idiosyncratic asset; - frequency and duration of transactions; - uncertainty	– Microeconomic approach and even macroeconomic approach (Berle and Means [BER 91]) – Unit of analysis: the agency relationship between the principal and the agent or the group constituted by the nexus of agency relationships
Type of efficiency	Static	Static
Type of rationality of agents	Limited computational rationality	Limited computational rationality (REMM model)
Vision of the company	The company as: – a nexus of contracts; – governed by hierarchical coordination; – efficient at a given moment if it allows, in comparison with the market, the totality of the costs generated by the transactions to be minimized	The company in its capacity as: – a nexus of contracts; – efficient at a given time if it minimizes transaction and agency costs
Handling of the environment	– Imperfect and static competitive environment – Consideration of uncertainty in probabilistic form (risk situation) – Set of given opportunities	– Imperfect and static competitive environment – Consideration of uncertainty in probabilistic form (risk situation) – Set of given opportunities
Central actors	– Approach of the company as a black box Key players: organizations versus the market; central players with decision-making power to act opportunistically	– Shareholder and stakeholder approaches Key players: the principals and the agents (or more generally the actors who cooperate)
Value creation	– Static – Minimization of losses in value related to the presence of transaction costs	– Static – Minimization of losses in value related to the presence of agency costs
Key concepts	– Information asymmetry, uncertainty, potential opportunism of actors – Transaction costs	– Information asymmetry, uncertainty, potential opportunism of actors – Conflicts of interest – Agency costs: monitoring costs, bonding costs, residual costs

2.1.3. Applying the theory to our research question

Staying in line with what has been presented on contractual theories in the previous sections, we begin by applying the theory of transaction costs to our problem (section 2.1.3.1) and following it up with PAT (section 2.1.3.2). In both cases, we begin our discussion from the perspective of an SME. The theories will help us, first, to identify the difficulties that companies supported by PE can encounter when forming alliances. Second, they will allow us to analyze the role that PEFs can play in solving the encountered problems. We will then turn to an analysis of the research question from the perspective of a PEF.

2.1.3.1. The role of PEFs in reducing transaction costs

We begin by looking at the role of PEFs through the TCT. According to the TCT, strategic alliances are defined as a hybrid mode of coordination, falling between the two extremes of market and firm (hierarchy) [WIL 91b]. So, what could the role of a PEF be in this perspective?

There are situations in which the implementation of governance systems by transaction partners may prove to be too costly relative to the potential gains. In such cases, the presence of a third party may be required. The role of the third party is then to be an arbitrator allowing contracts to be adapted more quickly than in its absence, which limits governance costs [WIL 79, pp. 249–250]. This may be the case if the involved asset or assets are moderately to highly specific, if the transaction is long term but occasional. Indeed, coordination costs in the presence of a specific asset are high and may not be covered when it is a non-recurring transaction. But, even in the case of recurring transactions, a third party can be useful.

Let us begin our discussion with a focus on SMEs (section 2.1.3.1.1) and follow it up with PEFs (section 2.1.3.1.2).

2.1.3.1.1. The SME perspective

PEFs typically invest in young, innovative, unlisted companies. As explained in the introduction to our problem, these companies present specific characteristics that will condition the role played by the PEF. We will therefore, first, detail these characteristics in the light of the TCT in order to explain the difficulties that these companies may encounter in forming alliances. This approach will then enable us to justify the presence of PEFs to overcome the problems encountered.

Difficulties encountered by companies supported by PE

Taking the effects of scale within transaction costs into account, Nooteboom [NOO 99, p. 20; NOO 93] argued that transaction costs increase in the presence of small unlisted firms compared to large firms. In light of the TCT, this can be explained in terms of information asymmetry, limited rationality, opportunism and uncertainty [NOO 93].

Information asymmetry

As presented in the introduction to our research question, due to their size and the fact that they are unlisted, small companies are not subject to the same disclosure constraints as listed and larger companies. This inevitably results in increased information asymmetry. The information asymmetry makes it difficult to assess these companies.

For alliance formation, this situation can generate additional costs, both for the company itself and for its future contracting partner. On the one hand, the detection of companies as future partners is more difficult and generates increased costs in terms of information research. On the other hand, this situation can induce costs for the company itself if it makes an effort to report or disclose information. Once detected, the examination and evaluation costs of these companies are generally higher due to the tacit nature of the knowledge and informal information.

Limited rationality

In small companies, the ability to process information is usually closely linked to the manager’s abilities. This also limits the scope for exploration and awareness of new possibilities for action. The problem in terms of intellectual capacity in information processing seems to be less significant for companies that are active in high-tech sectors, as the level of education of managers is assumed to be high. However, the lack of specialized internal staff (in finance, strategy, marketing, etc.) leads small businesses to resort to external experts. They build networks to obtain the necessary information. This networking is often informal and is costly in terms of seeking information and building relationships.

Opportunism

On the one hand, small companies are more vulnerable than larger firms to the potential opportunism of a contracting partner. On the other hand, large companies are more heavily subjected to the reputation mechanism, which reinforces the self-enforceability of contracts and thus reduces the risk of opportunism. This is due to the fact that the cost of cheating is higher if it is more likely that such a practice will be detected and the relevant information will be rapidly and widely disseminated. Large companies therefore seem to be more heavily subjected to this mechanism than small, unlisted companies. Building or maintaining reputation capital is more difficult and costly for a small, non-established company. As the markets in which small companies operate are less efficient, their information is disseminated more slowly and to a smaller audience than that of listed companies. Small businesses therefore have to make a greater effort when meeting with future partners to establish a situation of trust that will make their commitments trustworthy. This generates costs.

Uncertainty

Uncertainty only accentuates the contractual problems that we have mentioned so far [WIL 79, p. 254]. As defined by Knight [KNI 21, part I, Chapter I], it makes it impossible to model the various states of the world in probabilistic form. Companies financed by PE are typically active in high-tech sectors. These sectors are characterized by a high degree of uncertainty, which makes it even more difficult to draft contracts, so these therefore remain incomplete. The incompleteness of contracts and their informal nature are increasing in areas where the exchange of information and knowledge is important, such as R&D and marketing. Cooperation between companies in such areas gives rise to the creation of assets that are idiosyncratic to the transaction. They are the result of investments in human or physical capital that are specific to the transaction and which make it possible to achieve a rent if the contracts are executed correctly [WIL 79, pp. 240–241]. Investments in both intangible assets and human capital accentuate the effects and consequences of idiosyncratic investments [WIL 79, p. 242].

In summary, from a TCT analysis perspective, young, innovative and unlisted companies have certain specific features compared to large companies. These particularities are essentially linked to a more pronounced information asymmetry, a higher limited rationality and a strong uncertainty about the environment. Moreover, they appear to be less reliable ex ante than a large company with established reputation capital.

These features increase the transaction costs associated with seeking out information about potential transaction partners, evaluating them and setting up contracts. Contracts are essentially incomplete and informal in nature and controlling them ex post is also more costly. The formation of alliances in the precontractual phase is therefore more costly than for a large company, which can lead to the failure of such a transaction.

To ensure that the transaction does not fail, mechanisms must be put in place ex ante to reduce information asymmetry and set up guarantees or solutions that ensure the credibility of commitments and establish a situation of trust. Ex post contracts require a flexible mechanism so that they can be adapted [WIL 79, p. 242]. Because of this incompleteness of contracts, trilateral governance is then even more important, meaning a third actor playing the role of arbitrator [KRE 98, p. 133]. This is more effective than detailed contracts. Do PEFs reduce ex ante transaction costs that can prevent alliance formation? Ex post, can they assume the role of an arbitrator?

The role of PEFs in solving the problems encountered

Let us discuss if PEFs can reduce ex ante transaction costs, which can prevent the formation of an alliance and ensure an ex post arbitration role.

In our case, PEFs seem particularly well “predisposed” to limit transaction costs and act as arbitrators. Indeed, they are capable of reducing the information asymmetry that can explain, ex ante, why at least one of the future partners in the alliance gives up on forming the alliance. PEFs are able to produce the missing information at lower costs than the companies themselves (or potential alliance partners) because they already collect the necessary information for their own decision-making [JEN 76, p. 338]. Since they support the companies in question, usually they will have carefully analyzed the files before selecting a candidate for financing. In addition, their presence on the board of directors gives them direct access to information, particularly that of a strategic nature. They can thus facilitate the exchange of information and beliefs by monitoring information flows [HSU 06, pp. 206–207; LIN 08; GOM 09, p. 2]. PEFs are therefore able, ex ante, to reduce information asymmetry and thus increase the credibility of commitments, which limits transaction costs.

Ex post, PEFs can continue to play a favorable role in reducing the information asymmetry. They can also coordinate, if necessary, the actions of alliance partners and intervene in the event of conflicts. This last potential PEF intervention will be discussed in the section on applying the PAT to our research question. Before this, let us consider the perspective of PEFs.

2.1.3.1.2. The PEF perspective

Let us now turn to the point of view of PEFs. In the previous section, we argued that PEFs are able to reduce transaction costs that may prevent alliance formation and they can act as an arbitrator. But are these costs not borne by the PEFs? We distinguish between the cases of an intraportfolio alliance (formed within the investment portfolio of a PEF) and an extraportfolio alliance (formed between a company supported by PE and a company that is external to the investment portfolio of the PEF).

Intraportfolio alliances

From the perspective of a PEF, the costs associated with reducing the information asymmetry should be relatively low. As previously mentioned, the PEF collects a certain amount of information during its analysis of prospects for financing. Once a candidate has been chosen, they are regularly kept informed by the managers of the supported SMEs on the essential points. In addition, they usually sit on the board of directors or similar boards (strategic boards, etc.), which gives them access to information of a strategic nature.

Of course, PEFs are obliged to keep the information they have about their portfolio companies confidential. The provision of such information would in any case be unnecessary. It is sufficient for the contractor to know that there is a PEF involved in the capital of the future alliance partner to establish a certain level of trust. A contractor would know that PEFs carry out meticulous analyses before they invest in a company and that they require subsequent reports. This is all the more true if all of the companies in an alliance share a same PEF. The PEF is then a common actor that all of the partners are familiar with. Companies therefore know from experience what information PEFs have access to.

These conclusions are especially applicable to the formation of intraportfolio alliances, in other words alliances formed between companies supported by the same PEF. The question then arises as to whether this also applies to extraportfolio alliances, which include at least one company supported by PE and one company outside of the PEF’s investment portfolio.

Extraportfolio alliances

We propose distinguishing between two specific cases. In the first case, the company outside of the PEF’s investment portfolio is either a former PE-backed firm or a prospect that the PEF has analyzed but with which the PEF or the company renounced any cooperation. In this case, the PEF is aware of the company’s file and has had access to certain information about it. The argument can then be referred to the case of intraportfolio alliances.

In the second case, the company outside of the PEF’s investment portfolio is not known by the latter. In principle, the PEF has no informational advantage. Does this mean that the PEF cannot act as a mechanism to reduce transaction costs? We do not believe so. The fact remains that one of the companies is supported by PE. For the company outside of the PEF’s investment portfolio, this may reinforce the confidence it has in the SME supported by PE, because the presence of a PEF is a sign of the company’s quality compared to if it did not have this support. The company supported by PE has a strategic advisor with whom it discusses such steps and with whom it can seek advice. These different elements can reinforce an impression of quality and seriousness of companies and thus reduce the rapprochement costs for both companies.

The TCT makes it possible to highlight an initial role that a PEF could play in strategic alliances. In order to better study this relationship, let us use the PAT next. This will allow us to focus on the agency relationship, whereas the TCT uses the transaction as the unit of analysis. We can thus take into account potential conflicts of interest between contracting parties since the analysis allows the contracting parties’ preferences or attitudes toward the transaction to be explicitly taken into account.

2.1.3.2. The role of PEFs in reducing agency costs

Since the PAT focuses on agency relationships between two or more actors, it allows us to focus on the agency relationship that PEFs have in strategic alliances. According to the PAT, strategic alliances are a cooperative relationship between two or more companies. This cooperation makes it possible to achieve synergy effects due to collaborative production [ALC 72, p. 779]. It is effective at a given point in time if it can reduce the losses in value associated with agency conflicts between contracting parties and the associated costs [JEN 76; JEN 92].

Agency conflicts arise because of differences in interest between the contracting parties and they usually arise in the postcontractual phase, and therefore, in our case, after the formation of the alliance. However, potential differences of interests with the parties linked to the relationship may also pose problems in the precontractual phase, ex ante to the formation of the alliance.

In section 2.1.3.2.1, we discuss conflicts of interest in the precontractual phase and the solutions that PEFs can provide. Then, we proceed in the same way for the postcontractual phase (section 2.1.3.2.2). Although we are more interested in the formation of the alliance and not its life span, the anticipation of conflicts that may arise once the alliance has been formed can have the consequence, ex ante, that one of the potential partners refuses the cooperation. The alliance cannot then be formed. In a third point, we consider the possible interweaving of different governance mechanisms (section 2.1.3.2.3). Indeed, the presence of other mechanisms may weaken the respective weight of PEFs as a governance mechanism and thus their role in strategic alliances.

2.1.3.2.1. The role of PEFs in the precontractual phase

As before, let us begin our discussion from an SME perspective before continuing with the PEF perspective.

The SME perspective

Let us first consider the relationship between two companies that are backed by a PEF and that are forming an alliance, without taking the PEF into account. This will enable us to analyze the difficulties that this type of company may face in forming an alliance and to discuss, in a second step, the solutions that the PEF can provide.

Analysis of potential conflicts of interest between SMEs

Within the agency theory, actors are supposed to decide rationally – that is, they seek to maximize their own utility – and decide by evaluating the consequences of their choices. This computational rationality is however limited, as in the TCT. Individuals cannot predict everything and they make mistakes. As the information asymmetry is strong in the presence of young unlisted companies that are evolving in an innovative environment, agency conflicts linked to diverging interests are presumed to be likely.

The precontractual risk that arises in light of our problem is that a mutually advantageous alliance cannot be achieved. Because of the information asymmetry, it is possible that one party holds private information on the subject of the transaction and seeks to profit from it to the detriment of the other contracting party. The possibility of such behavior is sufficient to generate distrust of the future partner [BRO 93, pp. 13–15]. Since the relationship between the two companies forming the alliance can be considered to be dyadic, this risk of adverse selection arises simultaneously for both parties (for example [TEE 86]). In addition to these potential conflicts, given that PE-backed companies are in the precarious stages of their lifecycle and they operate in sectors with high uncertainty and often do not yet have any track record, potential partners may not join forces, first, because of doubts about their quality. Second, potential partners may refuse to cooperate with them for fear of possible financial instability [BAN 08].

As information asymmetry is at the root of potential conflicts, it must be minimized in order to avoid failure of the transaction. A classic solution to this problem is to insure against this risk on a contractual basis [ALC 78, pp. 302–307]. Depending on the design of the contract (for example the guarantees required), the contracting parties may need to disclose some information. Companies are sometimes led to send out signals that give credibility to their commitments. These signals are only credible if it is not possible to replicate them without bearing high costs. In the case of an alliance formation between young companies at the start of their activity and operating in an innovative sector, it is assumed that the investments of both parties linked to the transaction are mainly investments in human assets. Guarantees to give credibility to the parties’ commitments may be, for example, patents concerning know-how or diplomas guaranteeing the qualifications of the workforce [AKE 70, p. 494]. The request for guarantees and the issuing of signals by the contracting parties give rise to monitoring and bonding costs, respectively.

However, this solution presupposes the existence of appropriately formalized information. But, since PE-backed companies are often in the early stages of their lifecycles and operate in innovative sectors, their knowledge is usually quite tacit, and the information is rarely formalized. This complicates the situation, since commitments that are not based on written documents make it difficult to enforce contracts before the courts. Thus, they do not make the commitments credible. Formalizing the information can then be too costly, if not impossible, for these companies in the face of potential gains linked to the cooperation, which can lead to the failure of the transaction.

In the presence of informal contracts, mechanisms that facilitate the self-execution of contracts and thus reduce agency costs are therefore the solution. One example of such a mechanism is trust [AKE 70]. It facilitates investment in specific human capital, in particular by reducing the risk of adverse selection. This mechanism is important, especially since joint investment in human assets generates large flows of knowledge, which increases the risk of companies losing their own know-how [ALC 78, pp. 313–319].

However, trust is a bilateral mechanism [CHA 90]. It can only be established through the interaction of two agents, which requires either some time for it to develop or an ex ante guarantee. If the future contracting parties do not know each other before the transaction, a mechanism for establishing a situation of trust in the precontractual phase may be reputation. The latter is a complementary and necessary mechanism for the trust mechanism to work [CHA 90]. Reputation itself is only an effective mechanism if there is a high probability of detecting uncooperative behavior and if information about uncooperative behavior is disseminated quickly and widely. Reputation thus includes a shared dimension between several actors. However, when companies are young and unlisted, their reputation capital can be assumed to be low. The ex ante efforts made by these companies to demonstrate to their environment that they are reliable generate bonding costs that may exceed the anticipated gains from the relationship. The precontractual problem therefore remains.

This situation arises for SMEs that wish to form an alliance, whether it be an alliance formed within the portfolio of a PEF (intraportfolio alliance) or one including a company that is external to the PEF portfolio (extraportfolio alliance). However, the two cases are not similar with respect to potential PEF intervention. This is what we will discuss in the next point.

Solutions that PEFs can provide

Faced with this precontractual problem, we now seek to discuss the role that PEFs can play. As mentioned before, PEFs can reduce asymmetry and, consequently, transaction costs. Since information asymmetry is also the source of problems that cause agency costs, PEFs can also reduce then. They act as a mechanism to establish the necessary trust in the presence of informal and incomplete contracts and thereby give credibility to the long-term commitments of companies. But how do they do this? Once again, we distinguish between the two cases of intra- and extraportfolio alliance formation.

• A) Intraportfolio alliances

Let us begin with an alliance formed within the investment portfolio of a PEF (intraportfolio alliance). The PEF therefore supports both companies wishing to form the alliance. It can then act as a mechanism for building trust. But how does it do this?

First, since PEFs scrutinize the companies they support, this allows a relationship of trust to be established between the companies and the PEFs. The relationship they have with the companies they support serves as a basis for building mutual trust. This vertical trust between a PEF and the companies it finances can then serve as a basis for building horizontal trust between the two companies during the precontractual phase in the formation of a strategic alliance [AKE 70, p. 497; KRE 98, p. 133; CHU 00, p. 7]. Second, as PEFs are involved with several players and are usually listed, bad behavior on their part could be directly sanctioned. PEFs also have reputation capital to defend, which strengthens the trust mechanism. Because of their role as a trust mechanism, they reinforce the self-execution of contracts, which limits agency and transaction costs.

In the case of an alliance involving a company outside the PEF’s investment portfolio (extraportfolio alliance), the PEF can also play the role of a trust-building mechanism. However, in this case, the way in which it intervenes is slightly different.

• B) Extraportfolio alliances

Let us now consider the case of an extraportfolio alliance. Again, there are two scenarios. In the first case, the company external to the PEF’s investment portfolio is either a former portfolio company of the PEF, or a prospect that the PEF has previously considered but ultimately the PEF did not support. As before, our argument in this case refers to that of the intraportfolio alliance. In the second scenario, in principle, the PEF does not know about the company outside of its investment portfolio and vice versa. Unlike in the previous situation, however, the company outside of the PEF’s investment portfolio did not establish any prior relationship of trust either with its future alliance partner or with the PEF. As SMEs supported by PE are often, as already mentioned, at a precarious stage of their lifecycle, operating in highly uncertain sectors and often not yet having any historical performance to present, a potential partner that is external to the PEF’s portfolio may doubt its financial stability as well as its quality and may decide not to join forces. But does a PEF reduce these concerns?

Regarding the fear of financial instability, the presence of a PEF in the case of an extraportfolio alliance does have an effect. Mistrust toward the quality of a future partner leads parties to demand guarantees. A PEF, simply by its presence, can then play the role of “guarantor” and “certify” the financial stability of the SMEs it supports. If the latter face liquidity problems, PEFs are able to support them financially. This role seems all the more important given a PEF’s reputation on the markets. It should be noted that the terms “guarantor” and “certify” have been placed in quotation marks – these do not explicitly mean that PEFs formally guarantee or certify the financial stability of the companies they support. Rather, their mere presence in the SME’s capital allows them to play this role informally. Their presence, as well as their status as an investment company, helps to allay the mistrust of future partners and even to establish trust, compared to if there were no PEF at all.

Concerning the distrust of external companies toward the quality of an SME supported by PE, when it comes to associating within an alliance, a PEF can play a role in terms of certification of the quality of the supported companies. As one PEF in the United States stated in a 2006 press release: “Venture capitalists place a high value on strategic alliances and joint ventures as they provide an opportunity to demonstrate the validity of the science and its commercial potential” [MIT 06, p. 2]. In order to select the candidates they will support, PEFs undertake due diligence and assess, among other things, the specific knowledge, skills and resources of companies. They are able to identify and evaluate the know-how of SMEs, since they are specialized in these fields and provide the necessary skills. PEFs can then strengthen the credibility of these SMEs’ commitments to future alliance partners. They build trust through their expertise and reputation capital, which PE-backed companies could not do alone or at least not without incurring considerable costs.

In connection with the arguments we put forward, Megginson and Weiss [MEG 91], Stuart et al. [STU 99, p. 315], Hsu [HSU 04, pp. 1805–1806] and Nahata [NAH 08, p. 127] pointed out that when the quality of a young SME cannot be directly estimated, external actors rely on the quality of the actors that operate with the latter to assess its quality. As a PEF has more visibility and notoriety than the supported SME, it can then constitute such an actor. Indeed, in line with these ideas, Hsu [HSU 06; HSU 04] believes that PEFs can play a role in certifying the quality of the portfolio companies. In his 2004 article [HSU 04], he argued that the field of PE presents a near-ideal affiliation market with renowned agents for young SMEs. These companies are willing to accept a lower amount of liquidity in exchange for a percentage of their capital, if they value the certification role of the PEF [HSU 04, p. 1807; ALE 10]. Based on these results, in 2006 [HSU 06] he wrote that the presence of a PEF has a positive impact on the formation of cooperation between biotechnological SMEs that it backs, as well as on their likelihood of going public. Such cooperation takes the form of strategic alliances or licensing agreements. In addition, he shows that PEFs differ according to their reputation and that the latter has a favorable impact on both the cooperation activities of SMEs and their probability of going public. We can conclude that in the case of an alliance that includes a firm that is external to the PEF’s investment portfolio, the reputation of the PEF may play a favorable role in the establishment of trust.

The developments in this and previous sections lead us to propose the following hypothesis: the reputation capital of PEFs strengthens the trust mechanism, which makes it possible to reduce transaction costs (argumentation referring to the TCT) and agency costs (argumentation referring to the PAT), which has a positive impact on the formation of alliances for supported companies.

HYPOTHESIS 1.– All else being equal, the reputation capital of PEFs strengthens the trust mechanism and, therefore, has a positive impact on the formation of alliances for the companies they support.

The PEF perspective

So far, the issue has been presented from the point of view of SMEs. The question therefore remains about the interests and perspective of the PEF in the formation of alliances in the phase preceding their formation. In general, the interests of PEFs depend on their position in their lifecycle and the structure of their own investors.

According to Ozmel et al. [OZM 13], a PEF may have an interest in forming alliances between the companies it supports if its own reputation capital is weak. They studied the extent to which PEFs and alliance formation are complementary or substitutable mechanisms in the decisions of companies to go public as well as their impact on the probability of an exit via acquisition by another company or another PEF. In the case of an initial public offering (IPO), they found that the probability increases if the PE-backed SME forms an alliance. The PEF and the alliance are complementary mechanisms in this case. Nevertheless, this complementarity tends to become a substitution effect as the number of alliances formed increases.

Let us apply this reasoning to our own argument. In line with the developments in the previous section, a PEF plays a role in certifying the quality of the SME it supports if it has a certain reputation capital. This is built progressively through accumulation of experience and performance [SHA 83; HSU 04, p. 1807]. It can therefore be assumed that the younger a PEF is, the lower its reputation capital will be. In which case, it must build it. A faster way to achieve this reputation capital if it does not have enough, is to link its portfolio company with established partners, for example through a strategic alliance. Cooperation within an alliance with a partner with reputation capital or contact with other reputable agents can be either a complementary mechanism [CHA 04; OZM 13] (if the PEF has reputation capital but it is weak) or a substitute (if the PEF does not yet have reputation capital) for the role played by reputable PEFs. These developments lead to the following hypothesis:

HYPOTHESIS 2.– All else being equal, for a PEF with a low reputation capital, the formation of alliances for its portfolio companies with partners that have reputation capital constitutes a mechanism that is complementary to or even substitutable for the role of certification played by well-reputed PEFs.

Beyond the needs of PEFs according to the stage of their lifecycle, their interests will mainly depend on their own investors (the State, a region, private investors, a PEF owned by a company, etc.) as well as their purpose (focus on investments in certain sectors or regions, etc.). On the one hand, PEFs seek to make investments of their capital providers profitable. In connection with the objective of seeking rents, Kamath and Yan [KAM 10] pointed out that the presence of a PEF can have a negative effect during transactions between SMEs. They analyzed acquisitions between companies financed by PE and showed that PEFs can take advantage of their intermediary position by using their informational advantage to try to extract rents. On the other hand, it can be assumed that the interests of PEFs in forming alliances may differ depending on the nature of their ownership structure. Therefore, favorable effects can be expected if the State, a region or a competitiveness cluster are present in the capital of the PEF and they wish to promote regional development in general or in certain sectors. The influence of shareholders on the decision-making of PEFs, particularly in the formation of alliances, may be assumed to be greater if the PEF takes the form of a JSC with or without VC status, compared to if it takes the form of a VCMF. As presented in our introduction, PEFs in the form of VCMFs differ from those in the form of VCs in that investment decisions can be taken completely independently of the fund’s subscribers. Only venture capital funds with an investment committee are an exception to this rule if the committee consists of the fund’s main subscribers. The independence of the investment decisions undertaken by the management company may then be called into question.

In France, some PEFs are involved in schemes that are designed to promote the development of regional and/or sectoral activities and finance projects that structure Competitiveness Clusters (for example OSEO and the Fonds d’investissement stratégique). The objective of the clusters is to bring together companies, research laboratories and educational institutions within a given territory to develop synergies, cooperation and partnerships. We propose the following hypothesis:

HYPOTHESIS 3.– All else being equal, the presence of the State or a region in the capital of a PEF, or its involvement in a competitiveness cluster, positively affects the formation of alliances for companies supported by the PEF.

2.1.3.2.2. The role of PEFs in the postcontractual phase

Beyond the problem of adverse selection, potential partners may fear uncooperative behavior by the other party in the alliance once the alliance is formed. Although we are mainly interested in the role of PEFs in the formation of an alliance and not throughout the life span of the alliance, we must consider potential conflicts of interest after the formation of the alliance because their anticipation can lead, ex ante, to potential partners renouncing the alliance.

Information asymmetry makes it costly, if not impossible, to observe the other party’s behavior. The parties involved in a transaction may seek to take advantage of the situation. By taking advantage of the work provided by their partner (“free-riding”), they alone appropriate all of the gains from their behavior, while the related costs (a reduction in production) are borne by both parties [ALC 72, p. 780].

In light of our problem, there are two main agency relationships. The first is between two companies within an alliance, which we consider to be dyadic and bilateral in nature. This relationship is discussed in the section where we analyze the problem from the point of view of SMEs. The second main relationship, and the one that is of particular interest to us, is the agency relationship between the PEF and the alliance. This is further explored in the section on the PEF perspective. We distinguish, as we have done throughout this book so far, the relationship between companies that form an intraportfolio alliance and the relationship between companies that form an extraportfolio alliance, if this is relevant.

The SME perspective

To consider our problem from the point of view of SMEs, we break it down into three points. First, we discuss the different sources of conflict of interest. Then, we present the agency relationship between the SMEs that form the alliance based on these different sources of conflict. Finally, we present the solutions that PEFs can provide.

Different sources of conflicts of interest

Let us now turn to an analysis of potential conflicts of interest between actors. To do this, we must take the utility functions of the parties into account. Although the behavior of all actors cannot be accurately predicted, it is possible to make assumptions about their “typical” or “average” behavior. There are various reasons for conflicts of interest. By summarizing the models used in the literature on agency theory, the following sources could be identified:

• – the choice of level of effort that is a source of disutility for the manager or the consumption of non-monetary benefits by the manager, which opposes the interests of shareholders (for example [JEN 76, p. 312; JEN 86; BAL 03]);
• – different behavior due to different time horizons or costs of capital (for example [ROG 97; REI 97]);
• – differences in risk exposure: investors can be considered to have better access to financial markets than managers and, as a result, have better means of diversification. Consequently, assuming that individuals are risk averse, the premium asked by the manager should be higher than that called for by shareholders. In models, for reasons of simplicity, shareholders are generally considered to be risk neutral, whereas managers are assumed to be risk averse (for example [FEL 99]);
• – the problem of discretionary space and entrenchment.

This literature usually focuses on the agency relationship between shareholders and managers. The results therefore cannot be transposed to our problem, but they can serve as a “checklist” concerning potential sources of conflict.

Let us begin our analysis with potential conflicts of interest between two companies forming an alliance. First, we ignore the presence of the PEF. This will allow us to highlight any difficulties encountered by SMEs and, in a second step, to focus on the solutions that a PEF can provide.

The agency relationship between SMEs forming an alliance

The relationship between two companies forming an alliance is assumed to be dyadic in nature. Both companies thus simultaneously play the role of the principal, delegating the task to the agent, and the role of the agent who is delegated a task. The dyadic nature of the relationship, as we see it, does not mean that SMEs necessarily have the same weight in the alliance relationship. If one considers, for example, the particular case where an alliance takes the form of a joint venture, the percentage of capital held by the joint-venture companies is not necessarily identical. Nevertheless, a bilateral alliance is a collective effort. Like any group work, cooperation through an alliance encourages the transaction parties to take advantage of the partner’s work and appropriate a share of the rent. The problem is therefore about the choice of level of effort that one of the two companies puts in once the alliance has been concluded. Such behavior, known as free riding, is inherent in all group work [ALC 72, pp. 779–780].

This situation seems to become more prominent the more tactic and informal the purpose of the alliance is, and when the partners’ contribution is difficult to assess. This may occur if the alliance essentially involves the transfer of specific knowledge from one partner to another or if the alliance project is essentially based on the joint development of new knowledge. For young innovative companies, the specific knowledge they hold and can bring to the cooperative relationship within an alliance is often their key asset. However, their evaluation and protection require professional expertise that these companies often do not possess. Due to the tacit and informal nature of knowledge, the effort that one of the parties undertakes to enable the transfer of specific knowledge is difficult to assess. Similarly, when jointly developing new knowledge, it can be difficult to measure the involvement and quality of the effort made by each actor.

A different risk exposure is possible if the alliance is, for example, between a young unlisted company and a more mature company, since the latter could then have better access to the market and, consequently, better means of diversification. This seems to be more likely in extraportfolio alliances, which include a company that is external to the PEF’s investment portfolio. For intraportfolio alliances, it can be assumed that the majority of companies supported by PE are unlisted. In terms of the horizons of contracting companies with regard to the transaction, they can be assumed to be identical.

The solution proposed by the agency theory to such conflicts of interest is to discipline the agents. This raises the question of whether PEFs can be a disciplinary mechanism.

Solutions that PEFs can provide

PEFs usually sit on the board of directors or a similar strategic board within the companies they support and are able to act on the discretionary space of managers. They are therefore capable, in principle, of exercising a disciplinary role. We distinguish, again, the cases of an intraportfolio alliance and an extraportfolio alliance.

• A) Intraportfolio alliances

In the case of an alliance formed within the investment portfolio of a same PEF (intraportfolio alliance), it can be assumed that the PEF is indifferent to the companies forming the alliance and can exercise a neutral disciplinary role. It then plays the role of an arbitrator whose disciplinary power becomes all the more significant as it holds a majority stake in supported companies. In this case, as well as in the ultimate case where the alliance takes the form of a joint venture and the PEF directly holds shares in this joint venture, the PEF is even a “typical” arbitrator as described by Alchian and Demsetz [ALC 72]. This arbitrator reduces the problems associated with group production. According to Alchian and Demsetz [ALC 72, p. 783], such an arbitrator brings together all of the following rights:

• – the right to be the residual claimant;
• – the right to observe the contribution (input) of actors;
• – the right to be the central intermediary, common to all actors;
• – the right to elect or remove parties;
• – the right to transfer these rights.

The PEF is a residual claimant; it holds a majority stake in the companies forming the alliance or a significant number of shares in the joint venture and is thus encouraged to act within the objective of the companies forming the alliance, and therefore of the alliance itself or the joint venture.

• – Through its presence on strategic boards and its proximity to management, it is able to observe the contribution of all parties.
• – It is the point of contact between other alliance partners, especially since it is at the origin of the idea of the alliance.
• – It has the possibility of electing the CEO and the main executives of the companies that participate in the alliance or the joint venture.
• – The PEF is also in a position to determine the method of remuneration of the parties. Thus, it can have a disciplinary role on the holders of specific knowledge by having the possibility of linking, for example, their remuneration to the profitability of the alliance or joint venture [HAN 88, pp. 282–283].
• – It is able to directly expropriate by reselling shares [LIN 08, p. 1138] and appropriating the gains from the sale.

PEF representatives can monitor alliance partners and serve as an incentive mechanism to sanction uncooperative behavior. This disciplinary role becomes more important if a PEF holds a majority stake in the companies it supports.

PEFs are therefore in a position to exercise a disciplinary role in the case of an alliance that is formed within their investment portfolio. But what if the alliance includes at least one company supported by PE and another company that is not supported by PE?

• B) Extraportfolio alliances

If the alliance includes a company that is not supported by the PEF, the situation differs from before. The PEF then only has a link to the supported company. If it intervenes, it will do so to support the latter.

It will not then be able to play a disciplinary role as severely as for an intraportfolio alliance. On the other hand, it can advise the manager of the supported company if needed.

All these developments lead us to the following hypothesis:

HYPOTHESIS 4.– All else being equal, if a PEF holds a majority stake in the companies it supports, this increases its disciplinary power, which makes it possible to reduce agency conflicts. This reduces the ex ante distrust of future transaction partners, and thus has a positive impact on alliance formation.

The PEF perspective

In order to complete our analysis, we must also consider potential conflicts between the PEF and the alliance partners. Individuals working on behalf of the PEF are also expected to maximize their utility and act in their own interests. Generally speaking, any governance system that can act on the discretionary space of the top management is “two-faced” [CHA 98]. On the one hand, it can help to discipline the top management. On the other hand, it can also slow down their initiatives by imposing overly restrictive constraints.

A PEF primarily pursues financial objectives. It seeks to satisfy its clients, in other words to make the investments of its subscribers or shareholders profitable. The managers of SMEs that form an alliance pursue their own goals, which may not be financial. Discrepancies due to different horizons are also likely. PEFs usually have a time horizon of three to seven years to support companies inside their investment portfolio. The horizon of managers facing the alliance is one that is necessary to achieve the pursued objective. With regard to risk exposures, both PEFs and listed companies have access to the capital market and can diversify their risks, unlike unlisted companies.

Ozmel et al. [OZM 13] argued that a PEF may be reluctant to form an alliance for companies that it backs. In principle, an alliance partner has decision-making rights regarding the alliance project as well as control or monitoring rights. Depending on the weight of these decision-making and control rights, they may be the source of conflicts of interest between the alliance partner and the PEF. The alliance partner and the PEF could, for example, consider different exit arrangements [OZM 13]. From the point of view of a PEF, our problem, which is linked to possible conflicts of interest, becomes more significant with the number of alliances formed. Overall, the results suggest that PEFs may be reluctant to form too many alliances and thus abandon the formation of some alliances ex ante. These developments lead to the following hypothesis:

HYPOTHESIS 5a.– All else being equal, in the presence of a PEF, the number of alliances formed by a company negatively affects the formation of a new alliance (conflicts of interest).

In principle, therefore, PEFs seem to play a role in solving the problems that PE-backed companies encounter in the pre- and postcontractual phases of alliance formation. However, our analysis would be incomplete if the possible interweaving of several mechanisms, in a complementary or substitutable manner, were not taken into account in terms of governance [JEN 86; JEN 93; CHA 98].

2.1.3.2.3. The relative importance of the role of PEFs

Other mechanisms, particularly macroeconomic ones, that can also limit the discretionary space of the top management cannot be ignored. Consequently, through their involvement on the board of directors, PEFs may be just one governance mechanism among many others. The importance and impact of their governance roles depend on the interweaving of these different systems and their respective weights. The potential roles of PEFs in strategic alliances will be all the more important if other mechanisms are scarce and ineffective.

Jensen [JEN 93] put forward three macroeconomic mechanisms that should be taken into account alongside the internal control system with the board of directors at the top. These are capital markets, the legal system and the goods and services markets. The executive market can also be taken into consideration. In the company’s internal mechanisms, the board of directors is seen as the most important [CHA 98].

However, this analysis is based on a dominant view of governance, which corresponds to the shareholder’s vision and which essentially applies to managerial-type companies. These are characterized by a highly diffuse shareholding structure. Shareholders are considered as passive; their contribution is reduced to a simple financial provision. As for managers, they are not supposed to hold any significant share of the capital and, consequently, they do not bear the consequences of their decisions.

This view does not correspond to our problem, and therefore we must be careful when transposing the literature to our case (in other words, to small- or medium-sized unlisted companies) in the presence of a small number of majority and active shareholders (or even a single shareholder), thus a concentrated shareholding where managers generally hold a significant share of the capital. A small number of studies have looked into governance applied to SMEs [CHA 98; BRO 08] and showed that it is possible to use the work applied to listed companies as a guideline to analyze contexts that are more similar to ours.

As these companies are not listed, the financial market has no direct impact on their governance. This obviously changes with the listing of companies backed by PE. The disciplinary role of the executive market should have a relatively weak impact, given that managers in our context hold part of the capital. They thus bear, at least in part, the (financial) consequences of their actions. The goods and services market disciplines managers for listed and unlisted companies.

However, according to Jensen [JEN 86, p. 323], the disciplinary power of the goods and services markets appears to have a weaker impact in sectors that consist of new activities. La Porta et al. [LA 00] highlighted the importance of the legal system in corporate governance for investors. The investors involved in the legal system are essentially minority shareholders who need protection against the risks of expropriation by internal shareholders (majority shareholders and managers). According to Brouard and Di Vito [BRO 08]: “The legal investor protection system […] could apply to SMEs that use external financing such as venture capital investors. With well-established and enforced laws, companies must comply in order to avoid significant legal consequences”.

The role of PEFs in governance is particularly important for unlisted companies, as there are few alternative mechanisms. These other mechanisms are becoming more significant as PE-backed companies enter the financial market. The respective weight of PEFs and, consequently, their role in governance is diminishing. The role of PEFs in governance therefore depends on the ownership structure of the companies in which they invest. These developments lead us to the following hypothesis:

HYPOTHESIS 6.– All else being equal, the role of PEFs in strategic alliances is more significant in unlisted companies (and decreases with the opening up of company capital).

2.1.4. Review of the role of PEFs from the perspective of contractual theories

By applying contractual theories to our research question, we were able to provide an initial definition of strategic alliances and conduct an initial analysis of the role of PEFs in alliance formation.

According to the TCT, alliances are described as a hybrid mode of coordination, falling between the two extremes of market and firm. The PAT makes it possible to further develop this description. According to this theory, alliances can be seen as a nexus of contracts that takes the form of a cooperative relationship between two or more companies. This cooperation makes it possible to achieve synergy effects thanks to joint production. Alliances are effective if they can reduce losses in value that are associated with transaction and agency costs.

From this perspective, the literature suggests that these costs are higher for young unlisted companies, such as those supported by PE. These costs, which arise from information asymmetry and uncertainty, make it more difficult and too costly for these companies to form alliances. This raises the following questions: can PEFs have an impact on these costs? Do they reduce them? Do they bear them? Can they also generate costs?

The argumentation in the literature that is based on contractual theories emphasizes two main levers that reduce value losses linked to the presence of transaction and agency costs in situations of information asymmetry and uncertainty, for companies supported by PE. Here, we speak of trust, an informal governance mechanism that allows for self-execution of contracts and is reinforced by the reputation mechanism and disciplinary lever. As PE-backed companies lead to a greater information asymmetry, potential alliance partners may be wary of the quality of the company and renounce any cooperation. PEFs can then help build trust between future partners by reducing the information asymmetry. Since they support at least one of the companies in the alliance, they will usually have carefully analyzed the files before selecting a candidate for financing and hence they already collect information for their own decision-making [JEN 76, p. 338]. In addition, their presence on the board of directors (or an equivalent strategic board) gives them direct access to information of a strategic nature. They can therefore transmit this information inexpensively. If all future alliance partners are supported by a PEF, it can be assumed that these potential partners trust the information transmitted by it. And when alliances include partners that are external to a PEF’s investment portfolio, a PEF’s reputation capital strengthens trust and, as a result, has a positive impact on alliance formation for the companies they support (Hypothesis 1). Thus, Megginson and Weiss [MEG 91], Stuart et al. [STU 99, p. 315], Hsu [HSU 04, pp. 1805–1806] and Nahata [NAH 08, p. 127] point out that when the quality of a young SME cannot be directly valued, external actors base their assessment on the quality of the actors that operate with it. According to Hsu [HSU 04; HSU 06], reputable PEFs can play a role in certifying the quality of the companies they finance, which reduces distrust from external actors.

The second lever put forward by contractual theories is the disciplinary lever. Future alliance partners may also forego alliance formation for fear that the partner will behave uncooperatively once the alliance has been formed. In the case of an intraportfolio alliance, PEFs usually sit on the board of directors (or similar strategic boards) of the companies that form the alliance and are able to observe and monitor their contribution. They can facilitate the exchange of information and beliefs by governing information flows [HSU 06, pp. 206–207; LIN 08; GOM 08, p. 2]. They are able to control the behavior of alliance partners and discipline them in the event of non-cooperative behavior, which reduces the distrust ex ante of future transaction partners and thus has a positive impact on alliance formation. Since a PEF may hold either a minority or majority stake in the companies it finances, we assume that this disciplinary role of PEFs will be more significant if they hold a majority stake and that this majority stake will have a positive impact on the formation of alliances (Hypothesis 4). In the case of an extraportfolio alliance, a PEF only has a link to the supported company. If it intervenes, it will do so to support the latter. It will then not be able to play a disciplinary role as in the case of an intraportfolio alliance. On the other hand, it can advise the manager of a supported company if required.

Our analysis would be incomplete without taking into account the fact that other agents or mechanisms may play a complementary or alternative role to PEFs in alliance formation. The mechanisms that can influence the discretionary space of company directors and, in light of our problem, their decision to form alliances, generally appear with the opening of companies’ capital and their listing on the stock exchange [JEN 93]. The role of PEFs in strategic alliances is more significant for unlisted companies (Hypothesis 6).

We can consider our research question from the point of view of PEFs, in other words through their interest in intervening in the formation of alliances for their portfolio companies. Under contractual theories, the interests of PEFs will usually depend on the interests of their own investors.

In France, certain investors such as the State, a region or a competitiveness cluster have set up investment vehicles with the objective 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 positively influence the PEF in such a way that it becomes involved in the formation of alliances, at least in the case where the latter takes the legal form of a VC, thus leaving the investors the possibility of influencing decisions made by the PEF (Hypothesis 3).

As for the possibility of conflicts of interest between the PEF and an external alliance partner, for an extraportfolio alliance the current literature on the role of the PEF in alliance formation suggests that a PEF might be reluctant to have its portfolio companies form too many alliances. Since the alliance partner usually has decision-making rights under the alliance, this may result in a conflict of interest between the PEF and the alliance partner. In the presence of a PEF, the number of alliances formed by a company can negatively affect the formation of a new alliance (Hypothesis 5).

Finally, we argued that PEFs can play a role in certifying the quality of a company it supports. However, a PEF can only play this role if it has a certain amount of reputation capital. This is built progressively through accumulation of experience and performance [SHA 83; HSU 04, p. 1807]. It can therefore be assumed that the younger a PEF is, the lower its reputation capital will be, in which case it must build it.

If a company does not have enough reputation capital, a faster way to acquire it is for the PEF to link the company it is financing with established partners, for example through a strategic alliance. Thus, for a PEF with low reputation capital, the formation of alliances for SMEs that they support with partners that have reputation capital constitutes a mechanism of substitution for the certification role played by reputable PEFs (Hypothesis 2).

These points are summarized in Table 2.2.

	Transaction cost theory	Positive agency theory
Vision of alliances	The alliance as: – a hybrid mode of coordination; – between the two extremes of market and firm.	The alliance as: – a cooperative relationship between at least two companies; – a cooperation which makes it possible to achieve synergy effects due to joint production; – a dyadic relationship if formed between two companies; the two companies simultaneously play the role of principal and agent.
Difficulties raised	– SMEs supported by PE: - poor disclosure of information → information asymmetry; - low track record → low visibility, low transparency. – This results in high transaction costs when forming alliances.	– Ex ante alliance formation: problem of adverse selection. – Ex-post alliance formation: free-riding problem. – This results in agency costs.
Questions asked	– Can PEFs reduce these transaction costs? – Are the costs then not borne by the PEFs?	– Can PEFs reduce these agency costs? – Can there be agency costs between the PEF and future alliance partners? – Are there complementary or alternative mechanisms to the role played by PEFs?

Table 2.2. Review of the role of PEFs in strategic alliances from the perspective of contractual theories

		SME perspective	PEF perspective	SME perspective	PEF perspective
Responses received	Intraportfolio alliances	PEFs act as a trust mechanism. They: – give credibility to SMEs’ commitment; – reduce information asymmetry and transaction costs.	PEFs can reduce transaction costs without incurring them themselves, as they already collect this information for their own decision-making processes.	PEFs act as a trust mechanism. They: – ex ante, give credibility to SMEs’ commitments; – ex ante, reduce information asymmetry and agency costs; – ex post, can play a disciplinary and arbitrator role if necessary.	Positive impact of the presence of the State or a region in the PEFs capital on alliance formation.
	Extraportfolio alliances	Same as for an intraportfolio alliance, except that the PEFs only give credibility to the commitments of the supported company.		PEFs, by their reputation, play the role of trust mechanism: – ex ante, they provide credibility for the SME’s commitments and agency costs; – ex post, they can play an advisory role for the supported SME.	– PEFs with low reputation capital: linking their portfolio companies in alliances with reputable agents can be a complementary/alternative mechanism to the role played by established and reputable PEFs on the markets. – Reluctance to form too many alliances due to potential agency conflicts.
	Roles of PEFs decrease with the opening of the capital of supported companies.

So far, our research question has been discussed in light of contractual theories. On the one hand, PEFs make it possible to establish a situation of trust that prevents the eventual failure of an alliance. On the other hand, they can intervene in an alliance’s value creation by disciplining actors taking part in the alliance. They limit the occurrence of uncooperative behavior and the resulting loss of value.

Using them has resulted in an elucidation of the roles of PEFs in a single “negative” vision, limiting the losses in value that are linked to costs. These roles remain passive with regard to the formation of the alliance, the detection of potential partners. The sole use of contractual theories prevents the contribution of PEFs in the formation of alliances to be analyzed if the very idea of an alliance or the search for potential alliance partners emanates from them. Nor do they make it possible to value the contribution of PEFs in terms of skills, of their involvement in the strategic orientation of backed companies and, thus, of the alliance formed, or to understand how companies become aware of new growth opportunities that justify the alliance formation. Contractual theories thus dismiss the reasons and modalities for forming strategic alliances, as well as the analysis of the possible role of PEFs in the (productive) origin and creation of value through the formation of alliances. What about the potential roles of PEFs in creating value by forming alliances for the companies they support? In order to further discuss these points, we must turn to knowledge-based theories.

2.2. The role of PEFs in light of knowledge-based theories

Knowledge-based theories enrich our analysis by exploring the role of PEFs in creating value through the formation of strategic alliances using channels other than contractual. There are five such routes: analysis of the roles of PEFs in the identification and building of investment opportunities for backed companies, their strategic orientation, their growth, the generation and protection of new knowledge, and the coordination of individuals with divergent cognitive patterns [CHA 06].

Let us proceed with the following outline: in section 2.2.1, we present the theoretical foundations that are necessary to understand the argumentation on which knowledge-based theories are based. In section 2.2.2, we apply the theory to our research question. We carry out a review in section 2.2.3. Knowledge-based theories are then positioned against contractual theories in section 2.2.4. We conclude with a discussion on the complementarity of the two theoretical streams (knowledge-based and contractual) in section 2.2.5.

2.2.1. Some theoretical foundations

Knowledge-based theories – also known as KBV [CON 96, p. 477] – are underpinned by four major theoretical streams. These are the resource-based view (RBV), behavioral theory of the firm, evolutionary economic theory and the competence-based view (CBV). These theoretical streams are not independent. They are intertwined to a greater or lesser extent.

Let us begin by presenting some key points that are a common basis for knowledge-based theories (section 2.2.1.1). The components that we will discuss (RBV and CBV – because they are the most relevant to our subsequent developments) are based on these concepts (section 2.2.1.2).

2.2.1.1. Key concepts

We present some key concepts from Penrose’s work (section 2.2.1.1.1), then behavioral theory (section 2.2.1.1.2), followed by evolutionary economic theory (section 2.2.1.1.3).

2.2.1.1.1. Penrose’s work

In her analysis of the theory of growth of a firm [PEN 59], Penrose attempted to answer the following question [PEN 95, p. 7]: assuming that a business can grow, what principles will govern that growth and how quickly and how long can it grow for? In other words, assuming that there are growth opportunities within the economy, what determines the type of business that will benefit from it and to what extent? [PEN 95, Foreword, p. XI]. The core of her analysis on growth focused on a company’s internal resources and, more specifically, on the productive services that management can obtain from these resources, depending on its skills and experience. In order to carry out this analysis, we must define the firm in such a way as to take into account its internal organization as well bear adequate consideration of the environment. Keeping in mind that the economic function of an industrial company is “to acquire and organize human and other resources in order to profitably offer goods and services to the market”, Penrose defined the company as “a collection of resources bound together in an administrative framework, the boundaries of which are determined by the ‘area of administrative coordination’ and ‘authoritative communication’” [PEN 95, Foreword, p. XI]. A company is therefore both an administrative organization and a collection of productive resources [PEN 95, p. 24, p. 31, p. 77, pp. 149–151].

The unit of analysis that stems from this definition is the resource, and more precisely, the potential productive services that management can draw from these resources [PEN 95, p. 25]. The focus lies in management’s knowledge and skills, which accumulate with experience and enable them to perceive and use a resource’s potential services [PEN 95, p. 5 and p. 76]. The same resource may be used differently in alternative circumstances [PEN 95, pp. 67–68].

This accumulation of knowledge and skills is a dynamic and temporal process that is based on learning. As productive services are dependent on management, they are difficult to replicate, which makes them heterogeneous in nature. It is therefore the heterogeneity of productive services that can potentially be obtained from a resource, and that is a function of management skills, which gives a company its unique nature [PEN 95, p. 75, p. 199]. It determines the direction of growth [PEN 95, pp. 65–87].

No management can perceive the full range of productive services that can be exploited for a company. This mainly depends on the preexisting ideas that management puts forward in terms of possible combinations of the internal resources at its disposal. As a result, there are always opportunities for growth; it “just” requires having the entrepreneurial skills and experience to identify and implement them [PEN 95, p. 85].

Penrose distinguished between two types of management skills: managerial skills and entrepreneurial skills. Managerial skills are the services implemented by management to run the business and enable it to build and execute growth plans [PEN 95, p. 183]. These are qualities that enable good administrative coordination of the company. However, these managerial skills do not reflect any initiative to seek new investment opportunities [PEN 95, pp. 34–36] or a desire to acquire new knowledge in order to increase the existing stock [PEN 95, pp. 78–79]. Such initiative requires entrepreneurial skills. These skills are required for a business to grow¹ [PEN 95, pp. 34–36]. They include the skills required to create or accept innovation proposals and to take initiatives and decisions on growth proposals [PEN 95, p. 183].

The environment is considered as a reflection of management’s perception – having entrepreneurial skills – of the investment and growth opportunities at its disposal [PEN 95, Foreword, p. XIII and p. 5]. It is therefore not objective but is treated as a subjective construction that is dependent on management’s perception [PEN 95, p. 215]. The environment thus differs from one company to another, as it depends on management’s perceptions and a company’s internal resources. The latter is therefore not fixed; an entrepreneur can influence it through their perceptions and actions.

The markets in which companies operate are characterized by competition of a Schumpeterian nature. This competition involves innovators that introduce new combinations into markets and imitators trying to replicate these new combinations. Ultimately, the supply of these new combinations increases in the markets, which, all else being equal, reduces the rents received by innovators. Similarly, following the introduction of new combinations (innovations) on the markets, the old ones tend to become obsolete and disappear. This situation is characterized by a process of “creative destruction” [SCH 75, pp. 134–142]. This dynamic environment limits business growth and the ability of innovators to collect rents over the longer term [PEN 95, pp. 131–134].

Following Penrose’s analysis, strategic alliances can, in a first step, be defined as a collection of resources that are jointly governed and administered by the companies participating in the alliance. Companies governing the alliance can derive productive services from it. This allows them to increase the resources to which they have access as well as the quality and amount of managerial and entrepreneurial services. They can thus take advantage of a certain amount of growth opportunities that they would not have been able to seize had they gone it alone. This then raises the question of what effect the presence of a PEF has in light of knowledge-based theories. In order to answer this, let us continue our introduction to these theories. Penrose’s work is only a basis for knowledge-based theories and, in particular, the RBV and CBV on which our main arguments of the intentional roles of PEFs in alliance formation will be based. In the next two sections, we detail notions of behavioral and evolutionary economic theories.

2.2.1.1.2. Some essential contributions from behavioral theory of the firm

We present the essential contributions from behavioral theory in two points. First, we discuss Simon’s concept of limited procedural rationality, then we continue with the firm as a political coalition according to Cyert and March.

Simon’s concept of limited procedural rationality

Knowledge-based theories are based on Simon’s concept of limited procedural rationality. According to Simon, once we accept the knowledge-based and computational limits of agents, we must then distinguish between the real world as it presents itself to the individual and the individual’s perception of it. This will justify a theory on decision-making processes that can be empirically validated [SIM 82, pp. 270–271, p. 278, p. 368].

Limited procedural rationality is primarily focused on decision-making processes and not their outcomes [SIM 78, p. 498]. Individuals are no longer given a set of alternatives; they must obtain them. The consequences of choosing one of the alternatives are not necessarily known, either for certain or probabilistically [SIM 78, pp. 501–502]. Thus, we must explicitly take into account that anticipating the future can be linked to uncertainty. Agents’ expectations may be based, for example, on past experience and knowledge [SIM 61, p. 251]. As a result, individuals no longer optimize but instead choose a satisfactory alternative [SIM 56, p. 129; SIM 82, p. 291, p. 295] relative to a previously defined level of satisfaction that may have been revised during the decision-making process [SIM 78, p. 503]. The knowledge and experience of individuals determine a level of aspiration. As soon as they have found an alternative that meets this level, individuals stop the process of seeking new alternatives. As the decision-making process progresses, in particular with the arrival of new information and learning, the level of aspiration that was previously set can change [SIM 78, p. 503; BÉJ 01]. Similarly, the environment perceived by individuals is not fixed. It changes over time based on their experiences, their state of knowledge and the information available to them [SIM 82, p. 325].

While Simon was interested in how individuals make decisions (within companies), Cyert and March were more focused on the decisions made by a company.

The firm as a political coalition

In A Behavioral Theory of the Firm (1963), Cyert and March [CYE 63] studied how firms make decisions and the effects that organizational structure and business routines have on their behavior. They were particularly interested in the evolution of firms’ objectives, the formation of their expectations and the way in which they make their choices [CYE 92, p. 1]. In doing so, they developed the behavioral theory of the firm [CYE 92, p. 4].

Their analysis was based on the observation that two firms in the same environment and using the same variables in their decision-making process may nevertheless behave very differently if their organizational structures are different [CYE 55, p. 130]. They combined this idea by applying, on the one hand, Simon’s concept of limited procedural rationality [CYE 92, p. 19, p. 202, p. 214], and, on the other, elements from political science streams. At the time of their writing (1963), this political literature was increasingly used in the processing of organizational problems [MAR 62, pp. 665–666]. Cyert and March thus defined the firm as a political coalition where the leader is a political arbitrator. He is the intermediary between different parties who have an interest in the organization (investors, customers, suppliers, any type of employee, the State, etc.). This is a stakeholder approach that not only takes shareholders into account but also other agents or groups of agents with an interest in or affected by the organization. The focus is no longer on the shareholders and their objectives but on the agent that organizes the coalition [MAR 62, p. 674]. This agent, who is usually the manager, may nevertheless differ from one case to another. Various stakeholders may agree to participate in a coalition/company if the returns obtained by the coalition (usually salaries, dividends, or other types of remuneration) are high enough to keep them in the coalition.

The various actors within or in connection with a firm each have their own interests and objectives. They negotiate their opinions and actions in decision-making processes. It is up to the manager to organize them. Different parties that influence decision-making processes have different weights. This weight depends on their proximity to the problem that is to be solved, their hierarchical position within the organization and their personal ability to influence decision-making [CYE 92, p. 34]. A firm’s objectives and behavior then result from a process of negotiation between these different parties and are not given prior [MAR 62, p. 672; CYE 92, p. 31, p. 205]. The goals of a firm, or a coalition, are themselves the result of decision-making processes as determined by negotiations between the various parties. They can change and can be revised and adjusted in response to organizational changes [CYE 92, p. 33, p. 50]. These processes can be challenged. The firm/coalition learns and adapts to its environment [CYE 92, p. 99, p. 120, p. 215]. The latter is not perceived objectively. A firm responds to what the coalition observes, perceives and believes about the world around it [MAR 62, p. 673; CYE 92, p. 52; PFE 03, pp. 88–89]. It is therefore also negotiated between the various parties. Their observations and interpretations depend on the past, including their experiences, the state of their knowledge and the problems to be solved [CYE 92, p. 163].

Behavioral theory of the firm allows us to focus our interest on the decision-making process and the agents and factors that can influence it. In view of our research question, the central decision-making agent is the manager (often the owner) of an SME that is backed by PE and that forms the alliance. Since we are interested in the role of PEFs in alliance formation for the companies they back, the most decisive actor that interests us and that can influence the manager’s decision-making process is the investment manager who backs the SME. We must then ask ourselves the following question: to what extent can the investment manager intervene in the decision-making process of the manager of the SME he is backing? In combination with Penrose’s view of alliances, one might consider the investment manager’s influence on the strategic direction of the company being backed.

2.2.1.1.3. Some essential contributions from evolutionary economic theory

We break down our presentation of the essential contributions of evolutionary economic theory into two points. First, we present the influence of Schumpeter’s works which describe the competitive situation in which companies operate within the framework of the analysis of knowledge-based theories. Next, we present the company as a hierarchy of routines as defined by Nelson and Winter [NEL 82].

The influence of works by J. Schumpeter

Through his works Theorie der wirtschaftlichen Entwicklung (1911) and Capitalism, Socialism and Democracy (1942), Schumpeter can be seen as a precursor of evolutionary economic theory. The influence of his works within the streams underlying knowledge-based theories is felt at three levels: evolutionary argumentation and the processing of innovation as a source of economic change, the environment, characterized by a Schumpeterian competitive situation (as we have already mentioned for Penrose’s work), and the concept of an entrepreneurial function.

Evolutionary argumentation and innovation as a source of economic change

Schumpeter stressed the importance of an evolutionary economic approach. According to him, capitalism is by nature a process of change and not a stationary state. Nevertheless, economists often neglect this dynamic dimension in their models [DEM 69, p. 12; SCH 75, p. 136 and pp. 138–139]. Indeed, standard microeconomy models are based on a static design of efficiency. They do not, by nature, allow for a truly dynamic temporal effect.

Schumpeter’s argument for a dynamic design of economic models is a foundation of knowledge-based theories. All the theories underlying the latter are based on a dynamic design of efficiency, which can be described as “Schumpeterian” [CHA 06].

Schumpeter [SCH 35, pp. 316–317] considered that evolution is a process in which one moves from one position of equilibrium to another. There is an evolution after an innovation. This occurs if a new combination for the existing means of production is executed and this execution takes place discontinuously [SCH 35, p. 318]. Innovation can be described as a deviation from usual, routine behavior that transforms the equilibrium position [SCH 28, p. 483; NEL 82, p. 41].

Three factors appear in Schumpeter’s analysis of the theory of economic development [SCH 11]. The first is the “execution of new combinations” as a source of innovation. It is the core of the economic development analysis. The second is “credit” – at least, in a capitalist economy. It is how this execution can be achieved through financing it. The third is the “entrepreneur” or “entrepreneurial function”. A person with entrepreneurial characteristics allows new combinations to be identified and executed [SCH 35, p. 330, pp. 422–425 and p. 473]. This agent constitutes the fundamental factor of economic evolution [SCH 35, p. 330].

The creation of new combinations (innovation) includes the following five scenarios [SCH 28, p. 483; SCH 35, p. 319]:

• – the manufacturing of a new product;
• – the introduction of a new production method, usually reducing production costs;
• – the creation of a new market for an existing product or its introduction into a market that has not yet been exploited;
• – the new use of previously unused inputs;
• – the creation of a new organization.

The inputs used for the “old combination” are then used to generate the new combination [SCH 35, p. 321]. The latter is then juxtaposed with the first [SCH 35, p. 319]. A simultaneous process of creating a new combination and destroying an old one occurs.

The concept of innovation as adopted by Penrose is based on that of Schumpeter. Penrose takes up Schumpeter’s idea that an innovation takes place if there is a new combination of existing factors and if it is the basis of a firm’s growth².

Schumpeter’s view of innovation (largely technological in nature) as the basis for economic change, on the other hand, predates the argument and key points of the theory developed by Nelson and Winter [NEL 82, p. 277], which we present in the following section.

Schumpeterian competition

The set of publications positioning itself within one of the theoretical frameworks underlying knowledge-based theories places its analyses in a situation of Schumpeterian competition (for example [PEN 59; NEL 82; TEE 97, p. 509]). Schumpeter’s influence is particularly strong in the evolutionary theory proposed by Nelson and Winter [NEL 82, p. 39]. They even devote part of their work to the development of models under the conditions of Schumpeterian competition [NEL 82, Part 5 “Schumpeterian competition”, pp. 273–351].

Schumpeterian competition in a capitalist economy is characterized by a competitive situation that is dynamic. Innovation, the basis of evolution, generates a process of “creative destruction”. According to this process, when innovations are introduced, the old ones tend to disappear [SCH 75, pp. 134–142]. Innovations replace old combinations of productive factors for which the profitability decreases until they become obsolete. New combinations (innovations) are created and old ones are destroyed.

A situation of Schumpeterian competition is characterized by the intervention of two types of actors. On the one hand, there are entrepreneurs called “innovators”. In their quest for profit, they discreetly introduce new combinations of existing factors. On the other hand, “imitators” can appear. These attempt to imitate the innovators’ performance. The profit the entrepreneur receives is temporary in nature. When introducing an innovation, the entrepreneur is usually put in a monopolistic position. Over time, the profit tends to decrease, following the appearance of imitators on the market. Supply increases, which, at equal demand, leads to lower prices and, consequently, lower perceived profits.

The entrepreneurial function

The entrepreneurial function is based on the identification and execution of possibilities for new combinations of existing productive factors [SCH 28, pp. 483–485]. The agents who identify and execute these opportunities are entrepreneurs [SCH 35, p. 330, pp. 422–425 and 473]. This new combination is a temporary source of profit, which accrues to the entrepreneur. We give particular importance to the terms “identify” and “execute”. The entrepreneur is the one who allows the introduction and implementation of new combinations and not the one who invents them, even if these two agents can be one and the same [SCH 28, p. 484 and p. 486; SCH 35, p. 350]. The entrepreneur approaches an agent who takes the role of a leader to enable the economy to be carried forward [SCH 28, pp. 482–483].

Penrose also distinguished between the entrepreneurial function and the managerial function. The first is evidence of initiatives taken during the search and implementation of growth opportunities. However, Penrose’s entrepreneur does not quite correspond to the one defined by Schumpeter. Schumpeter’s is an innovator from the point of view of the whole economy since Schumpeter is interested in economic development. The entrepreneur in Penrose’s sense, on the other hand, is an innovator from the firm’s perspective, since she is interested in a firm’s growth [PEN 95, p. 36, footnote 1]³.

The firm as a hierarchy of routines

In their 1982 book, An Evolutionary Theory of Economic Change, Nelson and Winter presented an evolutionary theory of economic change. It is a dynamic and evolutionary analysis of the behavior and capabilities of firms that operate in a market environment that is characterized by Schumpeterian competition [NEL 82, p. 3]. Nelson and Winter [NEL 82, p. 52] argued that modeling economic change at an aggregate level (for industry or the economy) should be based on a plausible theory that best explains the diversity of a firm’s capabilities/skills and behaviors observed at the microeconomic level.

Through their thoughts, they provide answers to the following specific questions: what makes a company capable of achieving something? How does a company remember its capabilities or skills? [NEL 82, p. 52]. Where does business knowledge reside? What does the decision-making process involve? What skills are required to make a decision? [NEL 82, p. 53].

Nelson and Winter analyzed companies as hierarchies of routines. These constitute the core organizational capabilities where the organizational knowledge of companies resides. These routines define what a company is capable of doing well [NEL 91, p. 68]. They determine the behavior of firms (at least in part, as this is also a function of the environment they are in) [NEL 82, p. 14 and 128]. These routines can be seen as a set of habitual reactions, linking members of the organization to each other and to the environment [NEL 82, pp. 14–15, p. 142]. For the firm, they constitute what individual skills represent for individuals [NEL 82, p. 73 and p. 124]. They are composed of individual competencies carried by the agents who make up the organization, as well as their interactions and common experiences.

Defining the firm as a hierarchy of routines comprises three levels as follows [NEL 82, pp. 16–18]:

• – routines that guide short-term actions (operating characteristics);
• – medium-term routines that affect the variation of a firm’s production factors;
• – routines that allow companies to make major changes in the way they act in the short term.

The latter therefore act on the routines that guide short-term actions (according to Nelson and Winter, the “routine guided routine changing” processes). They follow on from new opportunity-seeking processes. On the one hand, they include evaluation processes (adequacy control and questioning), which are procedures or current operational routines of a firm. On the other hand, they include the search for new procedures and routines that can lead to drastic modifications, changes or replacements of existing routines [NEL 82, p. 400]. Since Nelson and Winter focused on technical change (technological innovation), this involves all of a company’s activities for improving their current technology [NEL 82, p. 210]. This hierarchy of routines, with its three levels, defines, on the one hand, the organizational competencies of a firm at an operational level and the way in which they are coordinated. On the other hand, it specifies decision-making procedures for making choices about the organizational competencies to be adopted and implemented [NEL 91, p. 68].

According to evolutionary theory, strategic alliances can be defined as the result of pooling the knowledge and skills of companies that form an alliance, with these appearing as organizational routines. The result is a new combination of operational routines, themselves the result of a learning process between companies involved in an alliance. In view of our research question, we then ask whether PEFs can intervene somewhere in the hierarchy of routines. One could be driven to think that by providing advice on the strategic direction of backed companies, they can intervene at the top of this hierarchy by questioning the general behavior of companies and therefore of the alliance, thus influencing the routines that allow companies to carry out major changes concerning their way of acting in the short term.

In the following section, we present the contributions of RBV and CBV to knowledge-based theories before applying the theory to our research question. These two theories, and in particular the CBV, are based on the key concepts that we have already discussed.

2.2.1.2. Contributions of RBV and CBV to knowledge-based theories

RBV has had and still gets many contributions in terms of academic publications. These articles usually have the common thread that, like Penrose [PEN 59], they adopt the definition of a firm as a set of resources. Nevertheless, although Penrose’s work is a dynamic analysis of a firm’s growth process, most of the work within the RBV (which is based on Penrose’s publication) places less emphasis on this temporal dimension. Instead, they focus on the conditions that a resource must meet to qualify as strategic in the sense that it is the basis of a competitive advantage for the firm [FOS 97, p. 15].

In contrast, the CBV, also called the dynamic capabilities approach [FOS 97; TEE 97] or core competencies approach [PRA 90, FOS 97], can be considered as a development or branch of the RBV, taking into account dynamic elements [FOS 97, p. 13; FRE 04] and interactive elements [FOS 97, p. 356]. Helfat and Peteraf [HEL 03] use the term dynamic RBV for this. In Penrose’s words, in constrast to publications that are “purely” part of the RBV, the focus is more on the services that can be derived from a resource (interaction of material and human resources) than on the conditions that the latter must meet to ensure a competitive advantage for the company. Many central concepts can be found in Penrose’s work [PEN 59] but the CBV was developed more recently than the RBV. The approach thus includes more recent contributions, notably those from the literature on strategy [CHA 62; CHA 90; AND 71; POR 80].

For studies within the RBV framework, one can thus distinguish those that do not explicitly integrate the temporal dimension from those that place greater emphasis on the dynamic dimension of phenomena. The incidence on the formulation of hyptheses is not zero. Following this distinction, let us first present the work that ignores the temporal dimension and that deals with the conditions that a resource must meet in order to achieve sustainable rents (section 2.2.1.2.1). Next, let us discuss the key concepts of CBV (section 2.2.1.2.2), which explicitly take dynamic effects into account. Our own analysis, which will follow, is to be included in the literature that takes dynamic elements into consideration.

2.2.1.2.1. Resource-based view

As defined by Andrews [AND 71], a resource is anything that can be seen as a strength or weakness of the firm [WER 84, pp. 171–172]. Examples of resources are as follows: a brand, clients, equipment, employees with specific know-how, working procedures, technology, capital, etc. The following question then arises: under what conditions does a resource ensure long-term competitive advantage for a company [WER 84, p. 172]? According to the RBV, Peteraf [PET 93] outlined four theoretical conditions. These four conditions are as follows: heterogeneity of resources, imperfect mobility of resources, ex post and ex ante limits to competition.

As Penrose’s analysis has already highlighted, the heterogeneity of resources is the basis of a company’s competitive advantage. It is the source of abnormal rents, i.e. above average [PET 93, pp. 180–182 and p. 185]. Unlike a homogeneous resource, it is difficult for a heterogeneous resource to undertake a relationship between the unit of entry and the service rendered (output) [PEN 95, p. 75]. Its heterogeneous nature gives it uniqueness and differentiability. Often, the services that are removed from a heterogeneous resource are also unique in the sense that they cannot be repeated or reproduced [PEN 95, p. 75]. As a result, these resources are scarce [BAR 91] and are generally limited in supply [LOC 09, p. 11].

However, the concept of heterogeneity alone is not enough for a resource to generate long-term rents. The conditions of imperfect mobility, ex ante and ex post limits to competition, must also be met.

There are four main explanations for the imperfect mobility of a resource [PET 93, pp. 183–184]:

• – it cannot be traded on a market [DIE 89, p. 1509]. These are, for example, idiosyncratic resources that have no use outside the transaction for which they were originally intended [WIL 79, p. 241];
• – it is negotiable, but loses value when used for other purposes;
• – it includes co-specialized assets, in other words its value is greater when used in conjunction;
• – its negotiation generates high transaction costs.

Ex ante limits to competition allow a firm to develop its competitive advantage while limiting its cost [PET 93, p. 185]. A firm can only generate positive rents if the costs of developing, implementing and maintaining competitive advantage are less than the gains generated [BAR 86, p. 1232].

Nevertheless, firms only derive a long-term competitive advantage from a heterogeneous resource if it cannot be developed or copied without costs [LIP 82; DIE 89, p. 1507; PET 93, pp. 181–182], or if there are few (if any) substitutes for these resources. These include ex post limits on competition to ensure that the supply of resources remains limited over time. Different mechanisms (such as property rights, reputation, excessive costs, etc.) can be the basis for imperfect imitability. Dierickx and Cool [DIE 89, pp. 1507–1509] highlighted five properties that can characterize the degree of imitability of a resource⁴. Among these, a central concept is that of causal ambiguity [LIP 82]. This concept refers to a state of uncertainty in terms of the causes and sources of different firm performances. It makes imitation difficult or costly because the difficulty in identifying causes does not allow potential imitators to know what to imitate [LIP 82, p. 418; BAR 91, p. 109; PET 93, pp. 182–183].

Publications within the RBV have focused on resources or assets that meet these four conditions [DIE 89, pp. 1506–1507 and 1509–1510]. These assets are usually characterized by a strong tacit dimension and result from an organizational and collective learning process within a firm [PET 93, p. 183]. These resources or assets are then considered to be strategic [DIE 89, p. 1510].

2.2.1.2.2. Competence-based view

The CBV takes on a more dynamic perspective. In a changing world, a company only has a competitive advantage (in other words receives higher than normal rents over the long term) when it is able to develop strategic assets, as defined by the RBV, on a continuous, faster and less costly basis than its competitors. Within the framework of the CBV, it is then the core competencies [PRA 90] of a company that play the role of potential catalyst, enabling it either to build and accumulate these strategic assets/resources internally, or to adapt and integrate strategic assets obtained externally through acquisition, alliance formation or other forms of cooperation.

For Prahalad and Hamel [PRA 90], a firm’s long-term competitive advantage thus lies in the management’s ability to consolidate productive technologies and skills, to develop skills to respond flexibly and adapt quickly to a changing environment. The “dynamic” nature of skills refers to a company’s ability to renew its existing skills or operational capabilities, such that they continuously adapt to changing external conditions [TEE 97, p. 515 and 516]. These skills can only be acquired and developed by human resources. As a result, employees are considered to be the most central asset in creating firm value [PRA 90, p. 11]. For Teece et al. [TEE 97], these skills (or dynamic capabilities) are essentially based on inimitable processes shaped by a firm’s pathways⁵ and strategic market positioning [TEE 97, p. 524]. These processes include coordination methods or routines on which a firm’s skills and abilities are based [TEE 97, pp. 518–520] and which determine the interactions between individuals [KOG 92, p. 396] within the firm and with its environment. The key feature is that skills cannot be obtained through markets (or through a price coordination system) [TEE 97, p. 517]. They must therefore be constructed and developed in-house [TEE 97, p. 518]. In general, they are the result of organizational learning. This is a dynamic and collective process of continuous development and improvement, usually involving a number of people in various positions within an organization. It is often characterized by trial-and-error, evaluation and feedback procedures [CHA 92, p. 84; TEE 97, p. 523].

Kogut and Zander [KOG 92] added that these skills are determined by socially constructed knowledge held by individuals but are also embedded in the organizational principles according to which individuals cooperate within a firm [KOG 92, p. 383]. They are composed of an “information” part: “who knows what?” and a “know-how” part, for example: “how to organize a research team”. A company’s superiority over the market thus lies in better sharing and transfer of knowledge for individuals and groups within a company. Firms develop new skills and create new knowledge by recombining their existing skills and knowledge through a learning process [KOG 92, pp. 384–385]. This includes combinatorial skills that can be acquired either through in-house learning or through external learning through cooperation, such as alliances, joint ventures or new contacts.

According to Markides and Williamson [MAR 94, p. 164], only companies with strategic links, that is “relatedness”, will perform well over the long term. A strategic link exists when the core competencies of one strategic business unit (SBU) of a company can be redeployed to another SBU, in order to reduce the costs and time required to create new strategic assets or to increase the existing stock of these assets [MAR 94, pp. 153–154].

Winter [WIN 03] pointed out that skills development does not always give a company a long-term competitive advantage. In order to cope with changing external conditions, a company has two means of action: either it acts proactively by developing dynamic skills (the company thus acquires a certain amount of flexibility) or it adapts to the environment in a reactive way, once the change has occurred. The superiority of one of these two methods over the other depends on the gains and costs they generate, respectively.

In this context, Langlois [LAN 92] introduced the concept of dynamic transaction costs or dynamic governance costs. Over time, companies’ skills evolve as they learn. The competitive advantage of a firm in relation to the market then depends on the relative learning capacities of the firm in relation to the market. In the long term, following the learning process, a firm becomes more and more capable. Nevertheless, the market, which is represented by competing firms, also becomes so and techniques disseminated by a company can be imitated. A firm’s ability to learn will depend on its internal organization. The market’s learning capacity depends on technical and institutional factors as well as the learning capacities of the firms involved [LAN 92, pp. 111–112].

This learning or skills development generates informational and/or knowledge costs that Langlois described as dynamic governance costs. These costs reflect the lack of certain skills needed by a company. These skills, if they cannot be built in-house or are too costly, must be acquired outside the firm. Therefore, an analysis is carried out on the costs of transferring a firm’s skills to the market and vice versa [LAN 92, p. 123]. These skill transfers are based on the transfer of information and knowledge, and include costs of conviction, negotiation, (knowledge-based) coordination, and training in terms of learning [LAN 92, p. 113].

In this context, the term “coordination” refers to coordination that can be described as “knowledge-based”, in the sense of ensuring that the agents involved are “on the same wavelength” [LAN 92, pp. 120–121]. Individuals may have different cognitive patterns. During the learning process, the transfer of skills can generate more or less significant costs, if mutual misunderstandings resulting from the diversity of cognitive patterns must be resolved between agents. Nevertheless, it may be advantageous for a firm to bear these coordination costs, since it is precisely this diversity between cognitive patterns of agents that is the source of innovation [LAN 92, pp. 121–122].

From a CBV analysis perspective, strategic alliances can then be defined as modes of cooperation between companies that remain autonomous but pool together their resources and skills, which enables them to draw new services from the resources at their disposal through combination and interorganizational learning, with a view to generating synergies, generating new knowledge or enabling organizational growth that they could not have achieved individually. This raises a question on the role of the presence of a PEF, particularly according to Langlois’ analysis, in coordinating the costs of skills transfer between partners in alliances. Before we begin our discussion, Table 2.3 summarizes the key points of streams underlying knowledge-based theories. It also takes key concepts into account as contributions from the different authors presented in the previous section. We give the reader a reminder on the interweaving of these theoretical streams, which is difficult to represent in tabular form. After that, we apply the theory to our research question.

Table 2.3. Knowledge-based theories: key points

	Resource-based view	Behavioral theory	Evolutionary economic theory	Competence-based view
Authors of fundamental works	Penrose [PEN 59] “The Theory of the Growth of the Firm”	Simon [SIM 47] “Administrative Behavior. A Study of Decision-Making Processes in Administrative Organization”; Cyert and March [CYE 63] “A Behavioral Theory of the Firm”	Nelson and Winter [NEL 82] “An Evolutionary Theory of Economic Change”	Penrose [PEN 59] “The Theory of the Growth of the Firm” and the influence of works such as Andrews [AND 71], Porter [POR 80] and Chandler [CHA 62; CHA 90]
Angle and unit of analysis	– Microeconomic approach – Unit of analysis: the resource (and more precisely for Penrose, the potential productive services that management can draw from these resources. But this aspect will be exploited more by the CBV)	– Microeconomic approach – Unit of analysis: the decision-making process	– Microeconomic approach – Unit of analysis: routines as company-specific key competencies	– Microeconomic approach – Unit of analysis: the services that can be obtained from a resource/interaction between material and human resources
Type of efficiency	Dynamic	Dynamic	Dynamic	Dynamic
Type of rationality of agents	Limited procedural rationality [CYE 63; SIM 76]	Limited procedural rationality [CYE 63; SIM 76]	Limited procedural rationality [CYE 63; SIM 76]	Limited procedural rationality [CYE 63; SIM 76]
Vision of the company	Simultaneously: – administrative organization; – collection of productive resources	– Simon [SIM 47]: an administration, an administrative organization – Cyert and March [CYE 63]: a political coalition	Simultaneously a hierarchy of: – routines; – decision-making rules	The company as a repository of: – competences/capacities; – knowledge
Environmental handling	– Schumpeterian competitive environment – Consideration of uncertainty – Set of opportunities not given, depending on management’s perception and actions	– Schumpeterian competitive environment – Consideration of uncertainty – A set of opportunities not given, depending on the perception and actions of the different parties in the company [SIM 47] and for which the final representation results from negotiations between the different parties [CYE 63]	– Schumpeterian competitive environment – Consideration of uncertainty – Set of opportunities not given, depending on management’s perception and actions	– Schumpeterian competitive environment – Consideration of uncertainty – Set of opportunities not given, depending on management’s perception and actions
Creation of value	– Long term – Source of rent: non-imitable and non-transferable heterogeneous resources		– Long term – Source of rent: non-imitable and non-transferable heterogeneous resources – Need to be able to constantly take advantage of this resource for rents	– Long term – Source of rent: non-imitable and non-transferable heterogeneous resources
Central actors	– Stakeholder approach – Key players: companies and human resources, especially management with entrepreneurial skills	– Stakeholder approach – Key players: actors with decision-making powers or coalitions concerned by the problem in question as well as the manager, playing the role of broker	– Stakeholder approach – Key players: companies in a changing environment; agents with decision-making power within companies	– Stakeholder approach – Key players: companies and human resources in general, meaning agents with key skills or knowledge; employees [PRA 90]
Key concepts	– Conditions that a resource must meet in order to constitute a long-term competitive advantage for the company – Resource heterogeneity [PEN 59], ex ante and ex post limits to competition [LIP 82; DIE 89; BAR86; BAR 91], imperfect resource mobility [DIE 89]	– Decision-making processes – Uncertainty – Limited procedural rationality – Coalitions – Negotiations between different coalitions	– Routines, hierarchy of routines – Organizational skills – Organizational knowledge – Technological innovation – Changing environment	– Key competencies [PRA 90] – Organizational learning [PRA 90] – Managerial and organizational processes, path dependence and strategic position of the company [PEN 59; TEE 97] – Combinatorial skills, knowledge and learning [KOG 92] – Strategic assets, related diversification [MAR 94] – Dynamic transaction or governance costs [LAN 92]

2.2.2. Applying the theory to our research question

Let us now apply knowledge-based theories to our research question. Referring to these makes it possible to account for the fact that PEFs are active and not passive shareholders, endowed with skills that they can bring to the companies they back. They will allow us to analyze the intentionality of PEFs in forming alliances for their portfolio companies.

These theoretical streams start from a vision of the firm as a set of inimitable resources [PEN 59], a repertoire of key skills or knowledge, organizational routines [NEL 82] or productive services according to Penrose [PEN 59]. Strategic alliances can then be defined as modes of cooperation between companies that remain autonomous but pool together their resources, key skills, organizational routines or knowledge in order to develop an activity, generate synergies, create new knowledge or enable organizational growth, which they could not have achieved alone (for example [TEE 86; PER 01; MEN 03; COL 06, p. 1166; HOF 07, p. 829].

In light of our research question and from the perspective of our analysis on knowledge-based theories, we can therefore ask: Where do growth opportunities come from? Are PEFs able to perceive growth opportunities other than those detected by the SMEs they back? Notably, by suggesting that their companies form alliances?

Once again, we begin with a discussion from the point of view of SMEs (section 2.2.2.1) before continuing with that of PEFs (section 2.2.2.2). The distinction between intra- and extraportfolio alliances is made if we consider it relevant. If it does not seem so to us, a justification will be given.

2.2.2.1. The SME Perspective

First, let us analyze the role of PEFs in the formation of strategic alliances in light of knowledge-based theories from the perspective of SMEs. The argumentation will provide answers in areas where knowledge-based theories are involved. This involves [CHA 06]:

• – the identification and construction of growth opportunities, the strategic orientation adopted, the diversity of corporate behavior, corporate growth and the direction of corporate expansion, as well as the creation and protection of new knowledge;
• – coordination of individuals with divergent cognitive patterns.

Our analysis is based on these two areas of intervention and is carried out, respectively, in sections 2.2.2.1.1 and 2.2.2.1.2.

2.2.2.1.1. The role of PEFs in identifying and building growth opportunities and creating new knowledge

The approach adopted in the following argument is based on the same manner used when analyzing the role of PEFs in light of contractual theories. In order to consider the role of PEFs in identifying and building growth opportunities and creating new knowledge, we once again start by questioning the characteristics of young companies that are typically backed by PE. Then we will analyze how a PEF can intervene in the formation of alliances for these SMEs.

Characteristics of SMEs backed by PE in light of knowledge-based theories

SMEs backed by PE have certain specific characteristics that, in the light of knowledge-based theories, are put forward on the basis of their key concepts. These include the limited procedural rationality, concepts of knowledge, resources and competences that appear in organizational routines and learning, the environmental context that is characterized by the consideration of uncertainty and Shumpeterian competition, and the perception of the environment by the company.

Limited procedural rationality

We focus our attention on the decision-making process and the actors and factors that can influence it. The central actor with decision-making power in view of our problem and, from the point of view of SMEs, is the manager (often the owner) of the SME backed by PE. Since we are interested in the role of PEFs in forming alliances for the companies they back, the most decisive actor we are interested in and who can influence the manager’s decision-making process is the investment manager who backs the SME. This involves analyzing the extent to which the manager of investments can influence the decision-making process of the manager of the SME he/she is backing.

Knowledge, resources and competences

With regard to the knowledge and competences of SMEs backed by PE, it can be assumed that these SMEs are highly specialized in their field of activity and therefore possess the specific knowledge and competences required. On the other hand, they usually suffer from a lack of competences and resources in areas such as marketing, finance and even management [NOO 93]. They can thus develop these, which takes time, or acquire them externally. The latter option requires either financial resources to acquire the skills, for example by recruiting key actors, or cooperating with other companies (by forming alliances to obtain these skills through interorganizational learning).

As for PEFs that back these SMEs, in principle they are less competent than SMEs in the latter’s field of expertise, but they do have skills or contacts in other fields such as management (for example in marketing and finance), as well as sectoral [BER 10], regional or international experience [LOC 08; LUT 10, p. 19].

Environmental context

Knowledge-based theories adopt an evolutionary perspective. They consider that an environment is characterized by uncertainty and competition of a Schumpeterian nature. Companies are faced with a constantly changing competitive environment. In order to be profitable in the long term and not succumb to the natural selection process, they must be innovative and constantly benefit from their innovation [NEL 82; NEL 91]. How management perceive the environment depends on their competences and knowledge, their past experience and the resources available to them [PEN 95, pp. 44–45, p. 215]. Thus, the environment, and the strategies adopted, can differ from one firm to another [NEL 82, p. 276; NEL 91, p. 69].

Companies backed by PE are typically young companies in the start-up, creation or development phase. As they are unlisted, they operate in very unstable environments. In order to be able to react flexibly to this environment, SMEs need to develop specific skills, especially organizational skills, which they often do not possess (as described previously). Can they obtain them through organizational learning? In particular, by forming alliances to ensure interorganizational learning?

Perception of the environment

The perception of the environment depends on the knowledge, skills and experience of agents. As described previously, these differ between SMEs and PEFs. It can then be assumed that the outlook of the environment faced by SMEs is broader for PEFs and narrower for SMEs. Can PEFs perceive growth opportunities or threats that SMEs do not, and vice versa? Is there any justification for forming an alliance in this situation? Do PEFs play any role?

In order to answer these questions, let us look more concretely at what can justify the presence of a PEF in the formation of alliances for these SMEs.

Potential PEF interventions in alliance formation

Given the characteristics of PE-backed SMEs under the light of knowledge-based theories, let us now discuss possible PEF interventions in alliance formation. We explain these in two steps. First, we clarify the role of PEFs in identifying and building growth opportunities through alliance formation. Second, we discuss the role of PEFs in creating new knowledge by forming alliances for the companies they back.

The role of PEFs in identifying and building growth opportunities through alliance formation

Let us start by analyzing the role of PEFs in identifying and building growth opportunities for their portfolio companies through the formation of alliances. Because of their differences in expertise, knowledge, resources and competences and the experience of the PEFs, on the one hand, and the SMEs they back, on the other, these two categories of firms may have different perceptions of the environment in which the SME operates. As explained, it can be assumed that SMEs backed by PE, which are highly specialized in their field of activity, have a narrower outlook of their environment than a PEF. Because of a PEF’s experience and expertise, it can have a broader overall vision of the environment in which its portfolio company operates. This then has an almost inevitable impact on the perception that both actors have of the risks that SMEs backed by PE can incur, as well as the opportunities that may be offered to them.

As soon as the companies applying for financing are considered, PEFs identify their key competences and resources and perceive the potential benefits (or services that can be derived from these resources, as per Penrose). The experience of PEFs, since they have seen many candidates, and since they have backed and are backing several SMEs, means that they may be able to identify certain difficulties or opportunities that SMEs could face quite early on. They are thus able to act accordingly by discussing and advising the managers of these SMEs. PEFs may therefore perceive different environmental changes (opportunities and threats) than SMEs and vice versa, and therefore they may adopt a different strategic vision.

Nevertheless – and as already mentioned – in order to be able to react flexibly to their changing environment, SMEs need to develop specific skills, especially organizational skills, which they often do not possess. PEFs can then recruit key managers [HEL 02, p. 171] that have specific skills or are likely to develop them within the SME. However, often the skills required can only be acquired through collective organizational learning [PRA 90; CHA 92]. Through the direct formation of alliances or the establishment of meeting clubs (Demeter Entrepreneurs Club or the Siparex Club), PEFs can promote organizational learning, and thus confer a long-term competitive advantage to a company, as defined by Prahalad and Hamel [PRA 90]. They can facilitate the formation of alliances with candidates that have the required skills to, for example, reduce the environmental risks that they have detected and that their portfolio companies face.

In line with these developments, Wang et al. [WAN 12] argued that PEFs intentionally contribute to alliance formation. PEF investments are often exposed to environmental risks. If neither the companies themselves nor the PEFs have the resources and skills to mitigate these risks, the PEFs would have to seek out alliance partners who do. PEFs thus create new opportunities for action. In their empirical study, Wang et al. [WAN 12] argued that in the presence of a PEF, SMEs in environments characterized by external technical risk tend to form strategic alliances with partners that complement their resources in order to mitigate this type of risk. Similarly, in the presence of a PEF, SMEs that operate in environments characterized by external market risk are more inclined to form strategic alliances with partners that have the resources and skills to develop expertise for reducing market risk.

Thus, the idea of forming a strategic alliance between companies that it backs can come as a consequence of a PEF identifying a growth opportunity. The formation of an alliance then results from the PEF’s initiative in order to allow backed companies to seize an opportunity. In this case, the PEF offers entrepreneurial skills [SCH 28, pp. 483–486; SCH 35, pp. 330–425, pp. 422–425 and 473; PEN 95, pp. 34–36, p. 183] and is the basis for the growth of firms that are in the alliance [PEN 95]. Consequently, it is directly involved in value creation if the alliance is formed; it is legitimately entitled to a share of the profit [SCH 35, p. 350].

These ideas can be further explored by analyzing whether the growth opportunities detected by PEFs are a function of their experience and skills. In other words, one may expect to see more frequent formation of certain types of alliances depending on the area of specialization of the PEF. In this perspective, a PEF that specializes in backing companies that belong to a certain activity sector would tend to form intersectoral alliances that enable the joint development of new products and services. On the other hand, a PEF with investment funds in different countries could be at the basis of alliances that allow for the international development of portfolio companies.

These developments are consistent with the results of Colombo et al.’s study [COL 06]. They emphasized that PEFs differ in terms of competences. They considered different types of PEF and showed that, depending on their nature, they tend to intervene in the formation of different kinds of alliances. Thus, for example, a subsidiary PEF of a company would tend to be involved in the formation of exploratory alliances that are focused on research and development of new products and/or services. Other types of PEF, which do not belong to a specific group or company, would instead tend to form commercial alliances, making it possible to exploit the know-how of the alliance’s partner companies. For the companies involved, the different types of alliances then provide access to specific resources and competences [TEE 86] through the intervention of the PEF.

Let us continue with a discussion on the role of PEFs in forming alliances to create new knowledge.

The role of PEFs in creating new knowledge through alliance formation

Through the involvement of PEFs in forming strategic alliances by identifying and enabling them to seize growth opportunities, their potential role in creating new knowledge can be brought into question [GRA 96, p. 120]. PEFs are also able to identify good practices within their portfolio companies, whether in terms of work organization, coordination methods or levels of employee participation in decision-making [PRA 90; TEE 97, p. 518]. They may then initiate alliances in which firms create new knowledge by exchanging and recombining their existing knowledge [KOG 92, p. 385]. PEFs thus promote the redeployment of certain key skills from one company to another. Through the direct formation of alliances, they can promote organizational learning, and thus confer a long-term competitive advantage as defined by Prahalad and Hamel [PRA 90], in particular by reducing the costs and time required to develop the required skills [NEL 82, pp. 117–121; MAR 94, pp. 153–154; HIT 11, p. 64]. In this case, PEFs show dynamic competences in the sense of Teece et al. [TEE 97, pp. 515–516], because they call into question existing organizational competences or routines [NEL 82; WIN 03; HEL 03] in the firms they back.

This questioning of existing organizational routines by PEFs can involve various levels along the value chain of the companies they back. These may include practices in internal control system, the use of management tools, how to disclose information [STI 65, p. 149], methods for supply and delivery, experience with certain production methods, etc. It is not uncommon for PE-backed companies to be managed by people with specific expertise and skills in the company’s field of specialization, but who have not been trained specifically in areas such as management, marketing, etc. [NOO 93]. The exchange of expertise and experience concerning work procedures or the use of certain management tools between companies can contribute to the questioning of existing procedures and the adoption of more efficient processes [OLI 90, p. 256]. Other forms of cooperation may involve undertakings that are active in the same sector. The exchange of practices between them provides an opportunity to see that they can unite in negotiations with certain key suppliers or to source between them within customer–supplier cooperations, which can give them a competitive advantage.

The distinction between intra- and extraportfolio alliances does not seem to be decisive in the context of the last two points. One may be led to think that if the idea of an alliance partner comes from the PEF, the latter prioritizes the other companies it has as potential partners in the alliance. If none are suitable, it would then subsequently consider a partner that is outside of its investment portfolio. Nevertheless, this simple analysis seems to become less relevant if one considers that, above all, the potential partner in the alliance must correspond to its objective and to the needs of the SME that wishes to form the alliance. At the same time, this potential partner must need what the company wishing to form the alliance can offer in exchange, according to the principle of dual coincidence of needs and desires. In other words, it is especially necessary that the choice of potential partner makes sense. We therefore conclude that a PEF does not necessarily and as a matter of priority establish alliances between the companies it backs.

PEFs seem to play an enabling role in forming alliances for their portfolio companies by contributing to the identification and construction of growth opportunities, allowing them to seize the opportunities and enabling them to identify knowledge and skills and develop new ones. However, our analysis would be incomplete if we did not consider whether PEFs might also play a negative role. Indeed, PEFs can, in principle, play a restricting role, in the sense that everything that occurs in the strategic orientation of a company’s activity forces it to simultaneously pursue this path to the detriment of exploitating other possibilities [CHA 02b, p. 52]. This seems less likely, however, if one considers that PEFs often act as advisors to the managers of backed companies. Except in cases where they hold a majority stake in the backed companies (or where they create a holding company in which they hold a majority stake, and which acquires the companies in question), PEFs do not impose decisions. They influence the decision-making process of the backed company manager, but the final decision rests with the latter. If a PEF directly or indirectly (via the creation of a holding company) holds a majority stake, PEFs are in a position to take strategic decisions. However, the manager of the backed company (if he/she is still in place) will have agreed to this when he/she handed over the management of his/her company to the PEF.

The developments in the last few sections lead us to make the following hypotheses:

HYPOTHESIS 7.– All else being equal, PEFs identify and enable their portfolio companies to seize growth opportunities by promoting the formation of strategic alliances.

HYPOTHESIS 7a.– All else being equal, we expect increased formation of intra-sectoral alliances in the presence of regional or sector-specific PEFs.

HYPOTHESIS 7b.– All else being equal, we expect an increased formation of alliances that enable the international development of participating companies if the PEF owns investment funds in different countries.

2.2.2.1.2. The role of PEFs as a facilitator of knowledge exchange

Now let us discuss if PEFs can play a cognitive coordination role. We begin by asking ourselves whether SMEs backed by PE can encounter difficulties during exchanges with future alliance partners. We then examine whether PEFs can provide any solutions.

The difficulties that SMEs backed by PE may encounter

An alliance involves the transfer of resources, competences and know-how between companies [TEE 86; KOG 88]. This exchange may be more or less costly depending on the “cognitive distance” [WUY 05; COL 06, p. 1171; KOG 92, pp. 388–389; NOO 93]. According to Langlois [LAN 92, p. 113], skills transfer is based on the exchange of information and knowledge and can entail costs of conviction, negotiation, coordination and training in terms of learning. In this context, “coordination” can be described as “cognitive”, in the sense of ensuring that the agents involved are “on the same wavelength” [LAN 92, pp. 120–121]. However, knowledge-based theories are not about minimizing discrepancies and related cognitive costs, but about managing these discrepancies as well as possible. This is justified by the fact that it is precisely this diversity between cognitive patterns of agents that is at the root of new knowledge generation and innovation [LAN 92, pp. 121–122; BRU 07].

Charreaux [CHA 02a] proposed to define cognitive costs by analogy to “traditional” agency costs – which are broken down into monitoring costs, bonding costs and residual agency costs. These then include mentoring costs, conviction costs and residual cognitive costs [WIR 06]:

• – mentoring costs: these are the costs incurred to enable knowledge transfer between actors [FOS 96, p. 18]. Thus, the processing of certain information requires specific skills that not all actors necessarily possess beforehand. The “mentor” acts as an advisor and tries to reduce misunderstandings. Wirtz [WIR 06] cites as an example: “A venture capitalist that helps a high-tech start-up to comply with the uses of communication with the financial community, in order to facilitate capital raising”;
• – conviction costs: this is the time spent by one actor to persuade the other to take a new opportunity into account;
• – residual cognitive costs: these costs are opportunity costs. Cognitive costs cannot be reduced to zero relative to a situation without cognitive costs, where all actors would have the same cognitive patterns. There is still some misunderstanding between actors with different cognitive patterns.

In summary, the different parties involved in a transaction – and in our case, the alliance – may have the same goals and horizons, the same exposure to risk, but act differently because of differing cognitive patterns. On the one hand, cognitive differences can be advantageous in the sense that one actor may, for example, be able to detect different opportunities that are the basis of innovation from another actor. On the other hand, a cognitive divergence between two actors can generate costs, linked to mutual misunderstanding. The aim is not to reduce cognitive differences but to facilitate their coordination [WIR 06].

Cognitive conflicts between companies that form a strategic alliance result from mutual misunderstanding. For this reason, these conflicts may, supposedly, be higher if the alliance partner emanates from the PEF and not from the companies themselves. If the idea of an alliance comes from the companies, they know each other at the time of the formation. They have formed a picture of the competences of their future partners and agree on the alliance. The situation is different if the idea of the alliance comes from the PEF or if the PEF proposes a partner. It is then likely that the potential alliance partners do not know each other beforehand. They then have to get to know each other first. This is discussed further in the following section.

Solutions provided by PEFs

The cognitive coordination mechanism can, in principle, be provided by a PEF [CHU 00, p. 7]. Let us discuss this by distinguishing between intra- and extraportfolio alliances.

Intraportfolio alliances

Intuitively, the role of cognitive coordination seems more likely to occur in the case of the formation of an intraportfolio alliance, in other words between two companies backed by the same PEF. In this case, the PEF is used to working with all the managers of the SMEs forming the alliance. The PEF knows their behaviors and the problems they face [COL 06]. Moreover, as a common actor, the PEF benefits from an intermediary position between the alliance partners.

This role seems to be more significant if the idea of an alliance partner comes from the PEF. One may expect that companies generally do not know each other beforehand. In this case, the PEF can be expected to play the role of “cognitive leader” in the sense that it will devote time to persuade partners of the opportunity that is offered to them, which generates conviction costs. In a second phase, the PEF’s intervention will generate mentoring costs in order to alleviate coordination difficulties that may arise between alliance partners [FOS 96, p. 18]. Mentoring costs, which arise from cognitive differences between partners, appear to be more important when they jointly control intangible activities or assets (such as knowledge transfers) than when jointly exploiting and monitoring assets (implied: tangible) [WIR 06].

Extraportfolio alliances

In the case of an extraportfolio alliance, one formed between at least one company backed by PEF and a company that is external in the PEF’s investment portfolio, we argue that PEFs can also act as a mechanism for cognitive coordination.

Either the alliance partner that is external to the PEF’s investment portfolio has been proposed by the latter, they know each other, and the argument is similar to that presented above, or neither the PEF nor its portfolio company wishing to form the alliance know the potential alliance partner. In this case, the PEF can nevertheless play a role through its presence. The PEF can be present during the first meetings to reinforce the image of seriousness and professionalism of its portfolio company. It can also act as an advisor for its portfolio company. This may be all the more important if the potential alliance partner is an established and reputable company on the market.

Thus, PEFs appear to be able to play this cognitive coordination role. Nevertheless, in all cases, this role may be assumed to be more important during the first meetings between future alliance partners, and decrease over time, depending on the frequency and intensity of exchanges between alliance partners. A mutual misunderstanding during the first meetings and negotiations between potential partners could result in the failure of a transaction. Little by little, the partners get to know each other and the cognitive coordination role of the PEF, which aims to reduce the cognitive divergences of the alliance partners, seems to diminish. As a result, cognitive costs decrease over time, and so does the cognitive coordination role of the PEF.

PEFs can then reduce cognitive discrepancies that are assumed to exist in the initial meetings between alliance partners. Our analysis would be incomplete, however, if we did not consider whether PEFs can also have a negative impact on the formation of alliances for their portfolio companies. Future partners may, in principle, have ideas for alliances that the PEF does not understand or for which it does not grasp the potential [STÉ 05, pp. 24–26]. This situation could then generate conviction and mentoring costs for the alliance partners, aimed at convincing the PEF. However, this theoretically possible situation should rarely occur in practice. In principle, there is no reason why a PEF should prevent one of its portfolio companies from forming any alliance, so long as it is in phase with the strategic direction of the SME.

These latest developments lead us to make the following hypothesis:

HYPOTHESIS 8.– All else being equal, the presence of a PEF facilitates exchanges between potential partners in an alliance, including at least one SME backed by the PEF and thus has a positive impact on the formation of the alliance.

Having discussed the role of PEFs in alliance formation from the SME perspective, let us now turn to the PEF perspective.

2.2.2.2. The PEF perspective

Let us now look at the issue from the perspective of PEFs. How could a PEF benefit from the formation of alliances between companies it finances in terms of creating its own value, competitive advantage or strategic direction? The following discussion is based on the interest of PEFs in forming alliances in terms of strategic direction and the generation of new knowledge or skills.

Like the companies they back, PEFs themselves face a changing environment marked by increased competition. In order not to succumb to the natural selection process, they must themselves be profitable in the long term. One way to innovate is to stand out on the PE market. This can be done by providing the companies they back with services that go beyond their primary role, which is to provide equity capital.

This additional service can be done by providing companies with a whole network of contacts that enables them to form long-term alliances. Thus, from its foundation in 1977, Siparex set up the Siparex Club to bring together the managers of companies backed by the Siparex Group, as well as shareholders and subscribers, with the aim of creating value in Siparex’s portfolio companies. Other PEFs have followed this example, such as Demeter Partners.

This brief development leads us to make the following hypothesis:

HYPOTHESIS 9.– All else being equal, the contribution of contacts that enable the formation of alliances constitutes for the PEF a means of differentiation on the PE market.

By their nature, alliances are a source of access to complementary resources and competences for the companies involved in the transaction [TEE 86]. From the perspective of PEFs, the formation of alliances between companies they back can be a mechanism to provide companies with resources or competences and visibility that they themselves cannot provide. Dushnitsky and Lavie [DUS 10] observed such an effect for young PEFs. The younger these were, the less experience, visibility or resources they had in terms of skills. This observed effect was reduced with increasing experience and reputation of the PEF, size and maturity of the backed SME and the number of alliances formed. Wang et al. [WAN 12] argued that the intervention of PEFs in the formation of alliances for the companies they back may constitute for a PEF a complement, or even a substitute, to part of their capital contribution. In order to reduce the external environmental risks faced by their portfolio companies, PEFs can promote the formation of alliances with partners that have the necessary skills and experience to mitigate these risks instead of providing the companies they back with additional capital.

However, there may also be cases where PEFs would be reluctant to form alliances for the SMEs they finance, if the alliance partner would offer access to the same skills as the PEF or if both were competing for exclusive access to the same resource held by the SME [COL 06, p. 1173]. The alliance would no longer be a complementary mechanism but a substitute [DUS 10]. A feasible scenario could be a PEF subsidiary of a company investing in an innovative SME because of access to its know-how. These results are supported in the study by Hoehn-Weiss and LiPuma [HOE 08]. The latter tested the hypothesis of complementarity or substitutability between PEFs and alliances on a sample of 111 American companies backed by PE and found no support for the hypothesis of complementarity between the formation of an alliance and the activity of a PEF. Ultimately, the results remain ambiguous.

A complementary reflection can however be made in terms of the number of alliances formed by an SME backed by PE in connection with the contractual argumentation that we discussed previously. We mentioned that the current literature suggests that a PEF may be reluctant to form too many alliances because, since an alliance partner usually has decision-making rights within the alliance, this may result in a situation of conflict of interest between the PEF and the alliance partner.

By using knowledge-based theories, however, we can also assume the contrary effect, in other words that PEFs are not reluctant to form a large number of alliances. The more alliances they form, the more experience they gain in alliance building, visibility and credibility. This enables them, among other things, to build a reputation and legitimacy on the market, which makes it possible to attract new partners to alliances and facilitate their formation. As for the PEF, in principle, it should not oppose the formation of a supplementary alliance through its portfolio company, as long as this is in phase with the strategic direction chosen for the company.

Thus, we wish to supplement Hypothesis 5a from the section on contractual analysis of our research question by formulating the following Hypothesis 5b:

HYPOTHESIS 5b.– In the presence of a PEF, the number of alliances previously formed by a company positively affects the formation of a new alliance.

2.2.3. Review of the role of PEFs from the perspective of knowledge-based theories

According to knowledge-based theories, strategic alliances can be defined as a mode of cooperation between companies, which themselves represent repertoires of knowledge, remaining autonomous but pooling together their knowledge, their resources, their key competences, their organizational routines in order to develop an activity, generate synergies or enable organizational growth that could not have been achieved had they gone it alone. The alliance is effective if it generates long-term value.

Long-term value results from the creation of competitive advantages that generate sustainable organizational rents. Knowledge-based theories provide five avenues of analysis according to which PEFs can create value by contributing to the formation of strategic alliances. These are the detection and construction of growth opportunities, the strategic direction adopted, the diversity of firms’ behavior, the growth of firms and the direction of their expansion, as well as the creation and protection of new knowledge, the coordination of individuals with divergent cognitive patterns [CHA 06].

Let us apply this reasoning to our problem in order to know if PEFs are able to identify growth opportunities for their portfolio companies and help them to implement these through collective learning which passes, in view of our problem, through the formation of alliances (Hypothesis 7). Through the detection and construction of growth opportunities, we can also consider the role of PEFs in creating new knowledge for the companies they back through the formation of strategic alliances. PEFs are also able to identify good practices within their portfolio companies, whether in terms of work organization, coordination methods or degree of employee participation in decision-making [PRA 90; TEE 97, p. 518]. They can then initiate alliances, in which firms create new knowledge by exchanging and recombining existing knowledge through collective organizational learning [KOG 92, p. 385]. These developments apply both to intra- and extraportfolio alliances.

The organizational opportunities and practices that PEFs are able to identify depend on the skills and experience of the PEF. Thus, we expect there to be a link between the expertise of PEFs and the type of alliances it can form. We also expect increased formation of intra- and intersectoral alliances in the presence of PEFs that focus their investments in certain sectors (Hypothesis 7a). Similarly, we anticipate an increased formation of alliances that enable the international development of the participating companies if the PEF owns investment funds in different countries (Hypothesis 7b).

Once the idea of an alliance has been evoked, what remains is to bring together the managers of companies that are candidates for the alliance formation. An alliance involves the transfer of resources, competences and know-how between firms [TEE 86; KOG 88]. This exchange may be more or less costly depending on the “cognitive distance” [WUY 05; COL 06, p. 1171] of alliance partners [KOG 92, pp. 388–389; NOO 93]. Any mutual misunderstandings during the initial meetings and negotiations between potential partners could result in failure of the transaction. In the case of an intraportfolio alliance, PEFs are used to working with the managers of the SMEs forming the alliance, and they know their behaviors and the problems they face [COL 06]. Moreover, as a common partner, PEFs benefit from an intermediary position between the alliance partners. In the case of an extraportfolio alliance, the PEF is not a common partner. It can nevertheless be present during the initial meetings to reinforce the image of seriousness and professionalism of its portfolio company. It can also act as an advisor for its portfolio company. This may be all the more significant if the potential alliance partner is an established and reputable company on the market. Consequently, PEFs are able to intervene in facilitating the initial exchanges between future alliance partners, which has a positive effect on the formation of the alliance (Hypothesis 8).

The question then arises as to how a PEF benefits from intervening in the formation of alliances for its portfolio company. According to knowledge-based theories, one may think that by providing this linking service, a PEF can find a way to differentiate itself on the PE market, which allows it to have a competitive advantage (Hypothesis 9). Moreover, the cognitive argumentation allows a different interpretation of the impact of prior alliances formed by a company backed by PE on the inclination of the PEF to encourage the formation of an additional alliance. As an alternative to the contractual argument, a positive effect is expected (Hypothesis 5b).

Table 2.4 lists the key points.

Table 2.4. Review of the role of PEFs in strategic alliances from the perspective of knowledge-based theories

		Knowledge-based theories
Vision of alliances		Alliance as a mode of cooperation between companies that remain autonomous but pool together their resources, key competences, organizational routines or knowledge in order to develop an activity, generate synergies or enable organizational growth that they could not have achieved on their own.
Difficulties raised		– SMEs backed by PE: - limited resources; - different vision of the environment from that of a PEF → different perception of opportunities and threats – Potential mutual misunderstandings during the first exchanges between potential alliance partners.
Questions asked		– Do PEFs create long-term value through the formation of alliances for the companies they back? – Do they give them a competitive advantage? – Do they intervene in the construction of growth opportunities, the generation and protection of new knowledge, the strategic orientation, the direction of growth, the coordination of divergent knowledge-based patterns, constituting the basis of innovation? – Can PEFs also play a restrictive role? – What interests can PEFs have in forming alliances in terms of their own value creation or strategic direction?
		SME Perspective	PEF Perspective
Answers given	Intra	– By forming alliances, PEFs can: - detect and build growth opportunities; - encourage the creation of new knowledge. – Role of PEFs in facilitating initial exchanges between future alliance partners.	– The contribution of an alliance formation service for backed companies can be a way for PEFs to stand out on the PE market. – No reluctance from PEFs to form additional alliances through their portfolio companies.
	Extra

2.2.4. Where knowledge-based theories stand compared to contractual theories

After discussing and applying contractual and knowledge-based theories to our research question, the following two sections are, respectively, intended to position these two theoretical streams one against the other (section 2.2.4), as well as our work in the debate that argues for complementary use of these theoretical frameworks (section 2.2.5).

Both contractual theories and knowledge-based theories are part of the efficiency paradigm [CHA 06] and are interested in the same types of phenomena: the explanation of the existence, limits and internal organization of firms [LAN 99, p. 214]. However, the two approaches differ in their conception of efficiency. The notion of efficiency that underlies contractual theories is static in nature. This is an allocative concept, of Paraetian origin, where value creation is achieved by minimizing transaction and agency costs at a given time. This value creation also depends on the distribution of the value created, except if Coase’s theorem is valid, whereby in the absence of transaction costs, the question of value creation and that of its distribution are separable [CHA 06]. Knowledge-based theories, on the other hand, adopt a dynamic, adaptive and proactive view of efficiency [HOD 98, pp. 187–189]. Efficiency, which is inspired by the Schumpeterian approach, depends on a firm’s ability to create value over the long term in a sustainable manner (and not at a given moment) [CHA 06; CUR 06].

Apart from the different conceptions of the notion of efficiency underlying the two theoretical streams (contractual and knowledge-based), a main difference lies in the hypothesis on the concept of rationality of agents. Contractual theories are based on the assumption of a (limited) computational rationality of agents in assessing the consequences of decision-making. Knowledge-based theories are based on procedural rationality, focusing on the decision-making process itself.

Two differences appear in relation to the concept of rationality retained by the theoretical streams: the distinction or lack thereof between “information” and “knowledge” and whether or not uncertainty is taken into account. The difference between “information” and “knowledge” only occurs within knowledge-based theories and becomes significant there [COH 99]. Knowledge comes from the interpretation and processing of information that is available to agents. Agents may interpret and treat the same information differently. Thus, they create knowledge. It is therefore subjective, whereas information is objective, independent of the interpretations and processing of individuals. This distinction is absent in contractual theories. The concept of computational rationality does not allow for a diversity of cognitive patterns. The second difference, related to the concepts of retained rationality, concerns the concepts of risk and uncertainty. This distinction goes back to Knight [KNI 21]. Unlike in a situation of risk, uncertainty makes it impossible to model the various states of the world in probabilistic form [KNI 21, downloadable version Part I, Chapter I. p. 26]. The concept of computational rationality does not take radical uncertainty into account. The concept of procedural rationality, on the other hand, explicitly reflects this. It assumes that agents are not in a position to rationally assess the consequences of their choices and shifts its attention to the decision-making process itself.

For the sake of argument, the two streams differ according to their definition of the firm. Within contractual theories, a company is defined as a nexus of contracts. The resulting unit of analysis is either the transaction or the agency relationship. Within knowledge-based theories, a firm is a set of resources or a repository of knowledge or competences. The unit of analysis is then the resource or competence or knowledge.

Because of the differences between the two streams and their approaches to efficiency (static or dynamic), the hypothesis of retained agents’ limited rationality (computational or procedural) and the different definitions of a firm, there results two fundamentally different arguments [CHA 06]. For contractual theories, the argument is disciplinary in nature. Value maximization occurs at a given point in time and involves reducing agency and transaction costs that result from conflicts of interest between the different agents involved in the transaction or in an agency relationship. Ultimately, the focus lies more on limiting value losses than on creating value [LAN 99, pp. 201–202]. Indeed, we do not question the origin of the opportunities at a given moment “t”. The whole point is to maximize the value created by making the possibility of acting on the limitation of losses in value (linked to transaction and agency costs) as the only lever of action.

Knowledge-based theories, on the other hand, question the origin and composition of value creation in a sustainable way. A firm is represented as a set of resources or a repository of knowledge or competences. Value creation depends on the company’s ability to generate new knowledge [FOS 96]. The argumentation, which is cognitive in nature, is part of a dynamic vision of value creation. We are interested in detecting and constructing knowledge at the source of new growth opportunities. We do not maximize the value created within a given set of opportunities. We seek to create them. All the opportunities that a firm faces are not given. The set of opportunities is open and its composition depends on the management’s ability to detect and exploit new opportunities. The environment typically differs from one firm to another or from one moment to another for a given firm. Knowledge-based theories make it possible to consider the productive origin of the organizational rent created, as well as to consider the origin of these opportunities, how to build them, identify them, perceive them and exploit them in order to create long-term value. Sustainable value is achieved by creating knowledge (interaction of competences and resources).

Table 2.5 summarizes the key points.

Table 2.5. Contractual theories versus knowledge-based theories

	Contractual theories	Knowledge-based theories
Theories underlying these approaches	Transaction cost theory (Coase, Williamson); positive agency theory [JEN 76]	Resource-based view [PEN 59], behavioral theory of the firm [SIM 47; SIM 76; CYE 63], evolutionary economic theory [NEL 82], competence-based view
Unit of analysis	– Microeconomic approach – Units of analysis: the transaction and/or agency relationship between at least two agents	– Microeconomic approach – Units of analysis: resource (RBV), services that can be derived from the interaction of material and human resources/resource interactions, knowledge, competences (CBV)
Definition of a company	Nexus of contracts	The enterprise as a collection of productive resources; as a hierarchy of routines and decision-making rules; as a repository of knowledge or competences/capabilities
Efficiency criterion	Static, reactive and allocative efficiency, Paraetian type	Dynamic and adaptive (proactive) efficiency, of Schumpeterian origin
Source of efficiency	Disciplinary	Productive
Selected rationality concept	– Limited computational rationality – Agents act by assessing the consequences of their choices	– Limited procedural rationality – Importance given to the decision-making process
Environmental handling	– Risk situation – The environment cannot be influenced – Given and closed set of opportunities – The terms “information” and “knowledge” are confused	– Situation of uncertainty – The environment is unique to each company and its stage of development. This depends on management’s perception and actions – Open and non-given set of opportunities to seize – Distinction between “knowledge” and “information”. The notion of knowledge takes precedence over that of information
Concept of time	Static approach	Dynamic and evolutionary approach, explicit consideration of the concept of time
Creation of value	Disciplinary approach to the concept of value creation. A system or entity is more efficient than another if it minimizes the loss in value that is caused by transaction and agency costs	Cognitive approach to value creation. Long-term value is created by creating and exploiting growth opportunities
Examples of questions that the theory can raise and answer	– What justifies the existence of the company? – What organizational structure should be adopted to limit value losses related to the presence of transaction and agency costs? – How to align interests between agents in order to limit losses in value linked to the presence of transaction and agency costs?	– Where does the created rent come from? – Where do growth and investment opportunities come from? How can they be built, detected, perceived and exploited to create long-term value?

2.2.5. The complementarity of contractual and knowledge-based theories

There are as many arguments that favor complementarity between contractual and knowledge-based theories as there are arguments that knowledge-based theories are an alternative to contractual theories [FOS 96; FOS 00]. There is no real consensus to date and we do not claim to contribute to this complex debate in this book. However, it seems important to position our work in this regard.

In our analysis, we borrow arguments from three theoretical frameworks: essentially, contractual and knowledge-based theories. Sociological network theories – which we have not yet looked at – are also used but, as we will explain in section 2.3, have the sole aim of complementing contractual and knowledge-based arguments, in other words for remaining within the framework of efficiency. Since we integrate arguments from different theoretical frameworks into a single explanatory model, our analysis should be placed among studies that are part of a complementary interpretation of the theories used [FOS 96, p. 19]. In this section, we therefore aim to present arguments in favor of complementarity between contractual and knowledge-based theories.

The main difference between contractual and knowledge-based theories lies in their argumentation of the value creation process [CHA 06, p. 5]. While contractual theories focus on minimizing losses in value that arise from the presence of transaction and agency costs, knowledge-based theories focus on the gains that can be generated by a firm through creating long-term value [CON 91a, p. 139 and p. 143]. To do this, contractual theories start from the point of view of a firm as a nexus of contracts; knowledge-based theories start from the point of view of a firm as a productive unit, representing a repertoire of knowledge [GRA 96, p. 113].

Thus, one may be inclined to suggest that there is a fundamental difference between the two contractual and knowledge-based streams, in particular, because of their different definitions. However, as Penrose [PEN 95, p. 10] stated, a firm is not easy to define except in terms of what it does and what happens inside it. Those who are interested in the analysis of a company are therefore free to choose the features of the firm on which they wish to focus their attention and apprehend it according to these features. Nothing then prevents one from, for example, considering the firm as a productive unit holding a repertoire of knowledge; the knowledge residing and evolving within it via the agents and their interactions; these agents being contractually bound to a central agent, a coordinator who ensures that the actors behave in such a way as to achieve the objective of the firm. In this case, the two theoretical streams complement each other; contractual theories focus their attention on the contractual aspect of the firm definition, making it possible to analyze how to reduce this risk and the associated loss of value. Knowledge-based theories, on the other hand, focus on the knowledge part, making it possible to explain the productive function of a company, in particular how it manages to achieve synergies beyond the market. Contractual theories then allow the necessary conditions for knowledge exchange and organizational learning to be created. In this way, they make it possible to create a business environment that is favorable to the development of knowledge, organizational learning, exchanges and thus the achievement of synergies owed to team production, which could not have been achieved by going through the market alone (see also [FOS 00, p. 23]).

Referring back to the fundamental texts, however, we see that a cognitive dimension is already present in the definition of a firm, as put forward by Alchian and Demsetz [ALC 72]. They defined the company as a production team that makes it possible to achieve synergies due to joint production while being governed by a central actor. On page 777 and in footnote 18, Alchian and Demsetz [ALC 72] insisted on interpreting the firm as a sort of private market with reserved access for team use of inputs or production factors, allowing better use and generation of knowledge on heterogeneous resources (the basis of organizational rents according to Penrose or Barney). Resource holders increase their productivity through cooperation and the firm, through its specific coordination, facilitates this cooperation (p. 777). Demsetz [DEM 88] – who can also be considered as a co-founder of the resource approach alongside Penrose – later pointed out that for him, the justification of production by a firm rather than by the market lies in the specificity of the production it can offer in relation to the market, depending on the circumstances.

********

The existence of a firm in relation to the market, to which Coase gave an initial answer, can thus be explained not only on the basis of the contractual argument but also on the basis of the knowledge-based (cognitive) argument. In analogy to Coase’s answer to the question on the existence of a firm that it reduces transaction (and production) costs relative to the market⁶, the following question can be formulated [CON 91a, p. 143]: does a firm allow for long-term gains superior to the market⁷? Alchian and Demsetz [ALC 72, p. 794]⁸ also provided an answer to this question: combining knowledge in a team makes it possible to achieve group synergies that cannot be achieved by isolated individuals. A team, by merging its knowledge, can design products and services that could not have been achieved on the market by purchasing them through other actors. In simple terms, the problem then becomes one of comparing an individual action or the addition of individual actions and a collective action. The existence of a firm is then justified by the collective action, as it makes it possible to carry out higher production (or an equal amount but at a lower cost) than that obtained through coordination through the market thanks to the synergy effects of the team production. The addition of individual knowledge via the market does not make it possible to benefit from the effects of the conjunction of knowledge, from organizational learning beyond individual learning and from the combination of knowledge.

Demsetz [DEM 91, p. 170] then presented one of the first attempts to summarize contractual and knowledge-based theories [CHA 06]: a firm is a specialized unit of production for others or a “repository of specialized knowledge and of the specialized inputs required to put this knowledge into work” [DEM 91, pp. 171–172], which is characterized by a sustainable and hierarchically coordinated nexus of contracts to ensure that agents within a company act in the desired manner [DEM 91, p. 170]. The central agent, the coordinator, makes it possible to limit the transaction and agency costs that can hinder cognitive exchanges.

We have yet to mention that contractual and knowledge-based theories are built on different models of rationality. Contractual theories are based on the limited computational rationality model; knowledge-based theories are based on the limited procedural rationality model [FRA 94, p. 736]. It also follows that an analysis carried out with contractual theories is static, whereas it is dynamic in knowledge-based theories. However, publications from core authors on contractual theories may contain elements that do not exclude a dynamic analysis. Jensen and Meckling [JEN 94] discussed different human models and concluded that the REMM (Resourceful, Evaluative, Maximizing Model), which serves as a basis for agency theory applied to financial, organizational or governance problems, best describes how individuals act. However, the model does not exclude the ability of individuals to create new opportunities, to innovate [JEN 94, p. 5]. There is therefore an evolving notion of time, which opens the door to possibilities of bringing the two theoretical frameworks closer together. Nevertheless, the pursuit of static efficiency may run counter to the pursuit of dynamic efficiency and vice versa [DEM 69; FOS 96, p. 8].

In the context of our work, the statements put forward following the contractual and knowledge-based analyzes are effectively used in a complementary manner, providing reciprocal theoretical refinements [WIN 88, p. 192]. The contractual analysis has made it possible to define strategic alliances as a hybrid mode of coordination (situated between two extremes, which are the market and the firm), characterized by a nexus of contracts and making it possible to achieve synergy effects due to joint production. The contractual argument has led us to define the following roles for PEFs in alliance formation:

• – from the point of view of SMEs, the role of the PEF is to reduce transaction and agency costs by establishing a situation of trust ex ante to the alliance formation, and of discipline once the alliance is formed (potential disciplinary role);
• – from the perspective of PEFs, this role is justified by the interests they may have in forming alliances for their portfolio companies (the presence of certain investors in the capital of a PEF may encourage the formation of alliances/a PEF with a low reputation may tend to encourage backed companies to form alliances with reputable actors/there may be a reluctance on the part of PEFs to form too many alliances as a result of potential conflicts of interest between alliance partners and the PEF).

The knowledge-based analysis has allowed us to define strategic alliances as a method of coordination between companies that remain autonomous but pool together their resources, knowledge and key competences in order to develop an activity, generate synergies or enable organizational growth that they could not have achieved on their own. The argumentation that follows from this has led us to identify the following roles for SMEs:

• – the role of PEFs in identifying and implementing growth opportunities and creating new knowledge for backed companies through the formation of alliances;
• – the role of PEFs in facilitating initial exchanges between alliance partners.

These roles are justified from the PEFs’ point of view because of the strategic interests they may have in promoting the formation of alliances for companies (strategy of differentiation on the PE market/lack of reluctance from PEFs to have their portfolio companies form several alliances).

In our case, the arguments from the two theoretical frameworks complement each other. PEFs are able to identify growth opportunities that can be achieved through the formation of alliances for their portfolio companies (knowledge-based/cognitive argumentation). The roles played by a PEF in the alliance companies in the light of contractual theories provide the necessary environmental conditions for the alliance formation to proceed smoothly. Ex ante, the establishment of a situation of trust between the actors makes it possible to avoid failure of the transaction. Ex post, the presence of a PEF can reduce potential free-riding problems, a necessary condition for the successful development of a cooperative relationship within the alliance and thus for learning, knowledge exchange and long-term value generation.

Definitions of strategic alliances in the light of contractual and knowledge-based theories can be integrated in the same way as definitions of a firm (seen as a nexus of contracts or as a repository of knowledge). Like Demsetz’s attempt at summarizing (presented previously), strategic alliances can be concisely interpreted as a hybrid mode of coordination between companies that remain autonomous. However, these companies are linked together within relationships of dyadic agencies (they play both the role of principal and agent), which enables them to pool together their resources, knowledge and key competences in order to develop, through joint production, an activity, generate synergies or enable organizational growth that they could not have achieved alone. In particular, in the case of intraportfolio alliances, the PEF can play the role of coordinating agent in order to limit the appearance of transaction and agency costs that can hinder cognitive exchanges.

To conclude, the static and dynamic concepts of efficiency in contractual and knowledge-based theories, respectively, do not seem to contradict each other in our analysis. Thus, from the very moment an alliance is formed, the reduction in value losses that occurs, which allows the alliance to be viable, does not in principle infringe the long-term efficiency targeted by the alliance for the participating companies.

2.3. Sociological network theories: a supplement to the contractual and knowledge-based argumentation

Sociological theories sometimes propose analyses that are compatible with the efficiency paradigm. Arguments put forward when referring to contractual and knowledge-based theories can thus be reinforced. In this way, they help to better understand the role of PEFs in bringing otherwise disconnected companies together, the role of networks in building up reputation capital, or an analysis of the possibilities to exploit the informational advantages of PEFs that occupy an intermediary position in the alliances formed. They also provide explanations in terms of power. By proposing explanations that are based on the concept of power and not in terms of efficiency, they make it possible to understand the role of PEFs in forming alliances in terms of a quest for social legitimacy [MEY 77], for coercion or power via influence or dependence [CHA 06]. The use of these theories is already present both in the literature on alliances (for example [ROB 06; LIN 09; AHU 09]) and the literature on PE (for example [WAL 97; LIM 10]). A few studies that directly tackle the subject also refer to it [CHU 00; COL 06; HSU 06].

In this study, we limit ourselves to discussing arguments that complement the ideas put forward within contractual and knowledge-based theories, and which are therefore compatible with the efficiency paradigm. This presentation of sociological network theories is therefore rather incomplete but is justified in view of the goal of our study. The conclusion will, however, propose avenues of extension in order to take other dimensions of sociological theories into account.

We begin by presenting some theoretical foundations (section 2.3.1). First of all, we are required to show what sociological network theories consist of so that the reader is well aware that we are indeed only referring to a rather limited part of them (section 2.3.1.1). We will then only present the part of these theories that we apply to our research question in more detail (section 2.3.1.2). We then directly apply the theory to our research question (section 2.3.2). After that, we review this topic in section 2.3.3.

2.3.1. Some theoretical foundations

In this first section, let us give an overview of the different theories that we believe can be included in the sociological network theories (section 2.3.1.1). In the following section, we selectively present the key concepts of these theories that seem compatible with the efficiency paradigm (section 2.3.1.2) and which will then be applied to our problem in section 2.3.2.

2.3.1.1. The approaches underlying sociological network theories

Unlike knowledge-based theories, the literature does not clearly list all the streams that encompass sociological network theories. We will include different streams that have the following three points in common: they must come from the field of sociology, must adopt a “meta” level of analysis and take into account the context or the social network in which agents are embedded, and they must be part of the paradigm of power, even if some of these theoretical streams may experience many developments that are compatible with the stream of efficiency. We have three main approaches:

• – the theory of population ecology [HAN 77];
• – the theory of resource dependence [PFE 78];
• – theories revolving around the concept of social capital, essentially:
  • - the theory of closed networks [COL 88],
  • - the structural hole theory [BUR 92].

The concept of power refers to a decision-making problem and is based on the human political model. Any organization is therefore considered, according to Cyert and March [CYE 63], as a political coalition [PFE 81; p. 28, PFE 03, p. 24]. Our focus is on the formation of political, negotiation or coalition activities that are undertaken to gain bargaining power and to cope with resistance and opposition in order to influence decision-making [PFE 81, pp. 6–7]. This is not an absolute power, allowing the agent who holds that power to impose their decision [PFE 81, p. 70; BUR 95a, p. 7]. Instead, the power is used to influence decision-making processes [PFE 81, p. 49].

Agents who hold the power are therefore usually those who can control or significantly influence decision-making processes, because they either:

• – have informational advantages and can exploit them at will [PFE 81, p. 275; PFE 03, p. 235];
• – are able to reduce uncertainty [PFE 03, p. 45];
• – have a critical and scarce resource on which other agents rely [PFE 81, p. 103; PFE 03, p. 108 and p. 232].

An agent is in control if he/she is in a position to influence or even initiate the actions of other agents at his/her discretion [PFE 03, pp. 258–259]. In particular, agents who are in control (in the sense that they are able to influence or even generate the information from which the different alternatives to decision-making are formed) hold significant power [PFE 81, p. 115 and 121]. Theories revolving around concepts of power and control are therefore really theories of negotiation (to exercise control) and, to this end, theories about the possibilities that an agent or an organization has of increasing its negotiating power or its coercive capacities in order to reduce its dependence. Thus, it would be more appropriate to speak of theories of freedom rather than power [BUR 95a, p. 7].

Beyond an agent’s individual ability to use power to influence decision-making, there are several conditions for the emergence of political, negotiation or coalition-building activities to gain power [PFE 81, p. 135]. Power is above all a structural phenomenon. It emanates from a division of labor, which results from the specialization of tasks within or between firms. Three of the conditions are structural in nature. These are the interdependence or interconnectivity between different agents, scarcity or rarity of resources and dispersal of power relations between agents that have an impact on decision-making. In other words, the power is not centralized and concentrated in the hands of a single agent, who can then simply impose his/her own decision [PFE 81, pp. 68–69, 70; PFE 03, p. 68].

The other two (non-structural) conditions are the following: first, the decision to be made must be of some or “critical” importance to the company. If this is not so, the cost of effort required to influence decision-making may be too high relative to the expected gains. Second, the situation must involve a conflict situation, in other words there must be a heterogeneity of objectives that is pursued between agents, or objectives that are incompatible between themselves. They cannot be pursued and achieved simultaneously [MAR 62, p. 663]. The nature of a firm’s relationships with other agents in turn determines the environmental uncertainty it faces [PFE 03, p. 68].

In this sense, an agent/firm’s structural position has an impact on its negotiating power, its access to information, its ability to participate in certain decision-making processes, and the means it is given to deal with uncertainty [PFE 81, pp. 130–131]. The position of an agent within a structure is not fixed. There are various strategies that allow the agent in question to increase their negotiating power [PFE 81, p. 137]. In particular, the resource dependence theory [PFE 78] places importance on the fact that an agent’s power is linked to his ability to cope and deal with uncertainty [PFE 81, p. 109]. This theory is based on the fact that in order to survive, companies need resources. Since they cannot own or develop all these resources themselves, they must procure them externally. This inevitably means that they must come into contact and interact with other agents who have the resources. The more essential these resources are to a firm, the more dependent it will be on the agents who own them. And the scarcer these resources are (in the sense that few agents have them), the more power and control these agents will have over those who depend on them. A firm’s survival therefore depends on its ability to cope with environmental circumstances, and to gain negotiating power in order to assert its interests and obtain the necessary resources [PFE 03, pp. 258–259]. A firm is seen as a political coalition [PFE 81, p. 28; PFE 03, p. 24] and its boundaries can be defined by the relative control that the firm has over the actions of its participants compared to other firms [PFE 03, pp. 258–259].

Thus, we can link this theory to Burt’s [BUR 92] theory of structural holes, which then allows us to consider the means available to actors to escape domination, whether by market domination or from another actor [BUR 95a, p. 3 and p. 7]. Faced with defined environmental constraints, the management’s task is to direct the firm toward a more favorably negotiated environment [PFE 03, pp. 262–263], for example, by allowing the firm to carry out mergers/acquisitions or alliances [SIR 11, p. 1401] when cooperating or lobbying [PFE 03, p. 267]. Therefore, a crucial point of managerial action is to correctly detect and identify the social context within which a firm operates and the environmental constraints it faces [PFE 03, pp. 18–19].

The argument in terms of power therefore differs from the argument in terms of efficiency and goes beyond the scope of this study. However, one should not assume that theories from the different paradigms are incompatible, even if their foundations differ greatly. Indeed, as we will see throughout this study, it is quite possible to apply an argument in terms of power to our problem: actions in terms of power can go against decisions taken following an argument in terms of efficiency. However, any decision taken (even if it is based on a logic of power) has an impact on value creation and can either run counter to efficiency or promote it. It is therefore possible to proceed, for example, by first analyzing the problem in terms of power and, second, to evaluate its consequences in terms of efficiency, thus evaluating the effect of the result of the analysis on value creation. According to Pfeffer [PFE 81, p. x], power processes are beneficial rather than constraining for a firm and the people who work within it. On page 335 of his book Power in Organizations, Pfeffer wrote:

“Although no data of the type described above have been gathered, the available literature and evidence does not support the argument that power and politics leads to performance problems. Indeed, there is some evidence that the reverse may be true”.

Not only is it possible to link an analysis on power to an analysis on efficiency as described, but some analyses or approaches to sociological network theories are directly compatible with the efficiency paradigm. This is particularly the case with the concept of social capital as grasped, for example, in Burt [BUR 92] and Coleman’s [COL 88] studies. They analyzed different network structures and the benefits they could bring to agents depending on their position within the network and the nature of the links they maintain with other agents in the network. Starting from the view that social capital is a resource, Coleman [COL 88] linked the concept of social capital to the efficiency paradigm [COL 88, p. 95]. As we shall see, there is a link, on the one hand, to knowledge-based theories because of the role of social capital in the emergence of human capital, in particular, in learning. On the other hand, there is also a link to contractual theories. Coleman’s analysis shows that due to the specific properties of the relationships between agents, certain structures favor the establishment of relationships of trust, which makes it possible (as we have seen) to reduce transaction and agency costs. Burt’s structural hole analysis also provides a link to knowledge-based and contractual theories. Burt focused on structural holes within a social structure and argued that these are a source of heterogeneity, and thus they are a competitive advantage for those who can identify them and know how to use them [BUR 95a, p. 2]. This creates a link with knowledge-based theories that favor the heterogeneity of resources as a source of sustainable rents. Simultaneously, an agent who knows how to use structural holes has an informational advantage that he/she can use at his/her discretion. They can act on information asymmetry and uncertainty, which has a direct impact on transaction and agency costs. There is thus a direct link to contractual theories as Burt himself suggests [BUR 95a, p. 3].

In view of our objective to use arguments from sociological network theories to complement the contractual and knowledge-based argumentation, we will refer to the concept of social capital and Coleman and Burt’s related approaches.

2.3.1.2. The concept of social capital: some reminders

Let us begin by presenting the origins and concept of social capital. After that, we will present the two main analyses: the Coleman analysis and Burt’s structural hole analysis.

2.3.1.2.1. Origins and description of the concept

Despite some earlier studies, the concept of social capital is most often attributed to Pierre Bourdieu’s work in the late 1970s and early 1980s. His novelty consisted of considering the social relations or contacts of an individual (or an organization) as a third form of capital, alongside human or physical (economic) capital [BOU 79; BUR 95a, p. 8; BUR 95b, p. 600]. As the name suggests, social capital is considered to be a resource or form of capital alongside physical or economic capital and human capital. Physical or economic capital is based on assets that are tangible, material, etc. Human capital consists of agents’ skills and knowledge. These are less tangible goods than those that constitute economic capital. Social capital lies in the relationships between individuals and the resources they provide for themselves. These are even less tangible assets than those that constitute human capital [COL 88, p. 100]. All three forms of capital contribute to productive activity [COL 88, p. 101]. Non-monetary forms of capital can be important sources of power and influence [POR 98, p. 2] and can complement or substitute for the holding of human or economic capital.

Bourdieu [BOU 80] defined social capital as “the set of actual or potential resources that are linked to the possession of a sustainable network of more or less institutionalized relationships of interknowledge and interrecognition; or, in other words, belonging to a group, as a set of agents who are not only endowed with common properties (likely to be perceived by the observer, by others or by themselves) but are also united by permanent and useful ties” [BOU 80, p. 2, our own translation]. The amount of social capital owned by an agent depends on the extent of the contacts they can actually use and the resources they hold (economic, cultural, human) [BOU 80, p. 2].

What is important in Bourdieu’s definition is that two conditions must be met for an agent to hold social capital:

• – they must have contacts that give them access to resources that can be useful;
• – the agent must be assured that they can effectively obtain these resources when they need them (see also [COL 88, p. 98 and p. 101]). An analogy can therefore be drawn with the concept of an option in finance. From this perspective, social capital can be considered as a set of resource options that an agent can obtain if needed.

In simple terms, the idea is that individuals with “better” contacts or social relationships perform better [BUR 00, p. 348]. Thus, an individual may perform better than another, not only because of his or her intellectual abilities (human capital) but also because of the opportunities he or she has in relationships with other individuals. Social capital is thus a complement to human capital [BUR 00, p. 347]. While human capital can be built individually, social capital can only exist through the interrelationships between two or more individuals. The analysis of the performance of individuals is therefore contextualized. It takes into account the structure of the social context in which individuals are embedded [UZZ 97, p. 35]. The level of analysis is therefore at the meso level, at the intersection of the agent and their social environment.

Thus, to understand the behavior of an individual or a firm, one must understand the context in which he/she acts [GRA 92, p. 6; PFE 03, Foreword, p. 19 and p. 39; GRA 05]. This refers to the notion of “embeddedness”, which dates back to Granovetter [GRA 85; GRA 92, p. 6]. Granovetter considered that economic organizations and actions reside at a level of socialization between two extremes: “oversocialization” and “under-socialization” [GRA 85, p. 483]. The idea of oversocialization is based on a Marxist view that considers individuals to be completely subjected to their social environment. Their actions are guided by what social norms dictate. Undersocialization is based on the concept of homo-oeconomicus. Individuals act in a rational and calculated way, free from any influence from their social context. This concept makes it possible to consider that individuals do not reason and act solely by economic calculation but, also, because of social constraints or in search of legitimacy of power [MEY 77].

The analysis therefore focuses on both the resources held by an agent’s contacts and on the structure of contacts within a network, in other words how the agents are related to each other [BUR 95a, p. 12]. Granovetter [GRA 73] discussed the nature of these relationships. The nature of the relationship depends on the time, attention, emotional intensity, intimacy and reciprocal services exchanged between agents. Simply put, a link can be strong, weak or absent [GRA 73, p. 1361]. Authors such as Coleman [COL 88] and Burt [BUR 92] considered a structure of different types of specific links (strong, weak or absent) and the resources they could provide. These will be presented in more detail in the following sections.

Networks of links are not given. They must be built and maintained [BOU 80, p. 2]. They thus require an investment of time and effort. These investments are strategic in nature (what emerges is uncertain, non-transparent and long term).

In view of our problem, a company’s relationship with its partner(s) in an alliance therefore constitutes part of its social capital [HOF 07, p. 829]. On the one hand, this relationship presents an opportunity that gives access to resources beyond the company’s borders [UZZ 96, p. 675] and simultaneously allows it to attain a certain legitimacy in the face of its external environment. On the other hand, societal relations can act as a brake on company development [UZZ 97, p. 35; HOF 07, p. 830].

Let us now consider Coleman’s analysis and Burt’s structural holes.

2.3.1.2.2. Coleman’s analysis

Coleman looked more closely at three forms that social capital can take depending on the function it performs. These are the forms of (A) relationships of obligations and expectations between agents, knowing that these relationships require trust; (B) channels that provide access to information; (C) social norms. He then looked at the conditions that a social structure must meet in order to allow these three forms to emerge. He concluded that, in particular, the social network must consist of strong links between agents and be closed in on itself.

In the following, we present Coleman’s three forms of social capital, the conditions for the emergence of these three forms of social capital, and social capital as a prerequisite for the emergence of human capital. This will enable us, in the section where we apply the theory to our problem, to analyze on the one hand whether PEFs are able to provide the social capital required for the emergence of human capital (organizational learning, transfer of skills and knowledge) for the companies they back through the formation of alliances. On the other hand, it will allow us to consider the structure of the social network that constitutes the PEF and the companies taking part in the alliance in the cases of intra and extraportfolio alliances.

Coleman’s three forms of social capital

• A) Social capital in the form of obligation/expectation relationships

This form of social capital exists when an agent “A” renders a service to an agent “B” and trusts the agent to render a reciprocal service later. In this case, “A” has a claim on “B” and “B” has an obligation to “A”. This form of social capital depends on two factors: first, on the trust that one can have in the proper functioning of the social structure, and therefore on the trust that one has in the services rendered to the agent being “reimbursed” at a later date. Second, it depends on the felt extension of agent “B’s” obligation to agent “A” [COL 88, p. 102]. Agent “A” only has social capital if he/she holds a service debt with at least one other agent and can actually obtain reimbursement (in the form of another service) when it is needed. Thus, we find the definition of social capital given by Bourdieu [BOU 80].

• B) Social capital in the form of information channels

Another form of social capital is the social relations of an agent “A” with other agents who possess specific knowledge in certain fields. Social capital then no longer resides in the obligations that these agents acknowledge they have toward “A” because it has previously rendered them a service. It lies in the access to information [COL 88, p. 104].

• C) Social capital in the form of standards

A third form that social capital can take is social norms. These are rules or laws, often informal, that the agents belonging to a network are obliged to respect. Their actions and behavior are not without consequences. According to established standards, they can either be sanctioned or rewarded. If an agent engages in non-compliant behavior with one or more other members of the network, they may unite to collectively sanction it [COL 88, p. 107]. This type of structure fosters relationships of trust between agents. This form of social capital can thus reinforce other forms of social capital, such as the relationship of obligations – expectations between agents by reinforcing the trust that agents have in the proper functioning of the system.

Conditions for the emergence of the three forms of social capital

In order to emerge, these different forms of social capital require the prior existence of social relationships and structures. Although all forms of social relations and structures favor the constitution of social capital to some extent, Coleman [COL 88, p. 105] cited two particularly important structures: the fact that a network is closed (“closure of social networks”) and the presence of social organizations that allow for the very establishment of relations between agents.

A network is closed if each individual in the network has a connection with every other individual in the same network. Interactions between agents are usually numerous and intensive. They have strong ties and know each other “well”. They know who has what resource or what information. These can be disclosed through several channels to be made available to different agents. They circulate rapidly between them, regardless of whether they are tacit or not, and whether they are easy or difficult to transfer [UZZ 96, p. 677]. The fact that a network is closed is particularly important for the emergence of social norms and relationships of trust. Social norms that allow a collective sanction or reward for agents’ behavior cannot emerge in an open social structure. In an open relationship, an agent “A” who has rendered a service to an agent “B” has a claim against “B”. If “B” decides not to render a service later to “A”, the latter is left alone to sanction “B”. In a closed structure, information regarding the inappropriate behavior of “B” is disseminated to all members of the structure. They can unite to collectively sanction “B”.

Social capital as a prerequisite for the development of human capital

Social capital not only allows an agent to access resources (physical, human or social capital) that are owned by another agent, but also develops an agent’s own human capital [BUR 00, p. 347]. In other words, it promotes the transfer and development of knowledge and skills, i.e. learning. Coleman [COL 88] illustrated the importance of social capital in building human capital (measured by the level of education) for the next generation, whether within a family or a community [COL 88, p. 108]. He demonstrated that social capital is a necessary complement to the transmission and construction of human capital from one agent to another (learning).

This is illustrated by the field of education (whether it is the parents who educate their children or collective institutions such as schools or churches that contribute to education). Thus, although in principle it is favorable for the education of a child that his parents (or those around him) themselves possess human capital (measured through their level of education), it serves little if these persons do not devote sufficient time and attention to educate the child [COL 88, p. 108]. In this case, social capital takes the form of physical presence and dedicated attention [COL 88, p. 111] and is a complementary mechanism to the human capital held by parents in their child’s learning process [COL 88, p. 109]. At the extreme, children do not benefit in any way from the human capital of their parents if they do not devote the necessary time to pass on their knowledge. Learning therefore requires relationships between agents that is characterized by strong ties. Social capital can therefore play a crucial role in the transmission of human capital.

Concerning PEFs, let us consider the question of whether they can intervene in the formation of alliances by making it possible to establish the social capital that is required for the emergence of human capital for the companies they back through the formation of alliances. First though, let us look at Burt’s contributions to the concept of social capital.

Burt’s structural hole analysis

Burt [BUR 92] analyzed the functioning of competition between agents when they have established relationships between themselves. According to Burt, an imperfect competitive situation is linked to the existence of structural holes, which are characterized by a lack of contact between agents. Agents in third party positions who have access to these holes and are able to exploit them can then receive non-zero rents, above the market average in the medium to long term [BUR 95a, p. 1]. These structural holes are a source of inequality and heterogeneity between agents.

In other words, the structural hole theory makes it possible to analyze the means that agents possess to escape domination, be it domination by the market or another agent. It is a theory of negotiated control and not absolute control, a theory of freedom rather than power [BUR 95a, p. 7].

In the following sections, we discuss the notions of weak ties and structural holes, the advantages of structural holes and tertius strategies, based on the third party position of an agent that can exploit these structural holes.

Weak ties and structural holes

The focus on the absence of links between agents (“structural holes”) dates back to White et al. [WHI 76, p. 731 and p. 774]. A structural hole is characterized by the absence of links or the existence of a structural non-equivalence between agents. This means the agents are not directly or indirectly linked to each other. These contacts (if one can speak of a “contact”) are then non-redundant, in other words they provide access to additional or supplementary information as opposed to substitute information [BUR 95a, p. 18].

Contacts can be redundant by cohesion or structural equivalence. Two contacts are redundant by cohesion if they are directly linked by a strong tie [BUR 95a, p. 18]. They are empirically characterized by frequent contact and emotional closeness (for example father and son, sister and brother, husband and wife, etc.) [BUR 95a, p. 19]. Two contacts are redundant by structural equivalence if they are not directly linked but have the same contacts. They are then not redundant by cohesion but redundant by structural equivalence [BUR 95a, p. 19].

Burt claims that these holes (non-redundant contacts) constitute entrepreneurial opportunities [BUR 95b, p. 601]. An agent who detects them can exploit these opportunities by filling in/closing the structural hole. He/she can achieve this by positioning him/herself as an intermediary (broker), linking the otherwise disconnected agents. As long as he/she is the only means of connection, he/she acts as a bridge [GRA 73, p. 1364] and gains a special position conferring informational and control advantages [BUR 95, p. 2]. It should be noted, however, that this is particularly true for small networks: it is rare that in fairly large networks there is only one means by which information can circulate (an agent that constitutes a gateway). Rather, this phenomenon occurs locally [GRA 73, p. 1364; BUR 95a, p. 24].

Granovetter [GRA 73, p. 1376] demonstrated that only weak ties can constitute bridges, with one exception. In principle, within a closed network, there can be no gateways. Within a closed network that is solely composed of strong ties, each agent has the same informational advantages and will discover the same opportunities at the same time [BUR 95a, p. 17]. There is one exception, however, where a strong bond can play the role of a bridge: if two individuals have a strong tie and neither of them has a strong tie with another person. However, weak links are faster gateways. These are often the shortest paths through which information can pass. This does not mean that every weak tie is a gateway. Burt differed from Granovetter by shifting his attention to the structural hole rather than to the (weak) tie that forms a bridge between two contacts that are separated by a structural hole (absence of tie) [BUR 95a, p. 28].

The wider and more diversified an agent’s network of contacts is, including access to non-redundant contacts, the greater the entrepreneurial opportunities will be [BUR 95a, pp. 16–17]. An agent who exploits these opportunities can hope not only for informational, timing and reference benefits, but also for control [BUR 95a, p. 2]. Let us see what these benefits are.

Structural hole advantages

Information gains are related to the nature of ties that are inherent in a structure that is rich in structural holes. The more a structure is rich in structural holes and the weaker or more absent the ties, the more it allows access to information that is not redundant and therefore, to new and varied information that brings opportunities. The informational benefits are threefold: they concern access, synchronization and referral of opportunities [BUR 95b, p. 602]:

• – access benefits: increased access to a variety of information allows the individual to be better informed about opportunities or bad news;
• – synchronization benefits: the individual who makes it possible to establish a contact tie between the holders of non-redundant information occupies an intermediary position (broker). This position is favorable to him/her insofar as he/she is the first to detect the needs of an individual or a group that another individual can fill with his/her skills. He/she is then in a position to synchronize the needs and skills of different agents;
• – opportunity referral benefits: because of the intermediary agent’s various contacts, he/she is more likely to be taken into account when new opportunities arise.

The intermediary agent also benefits from control advantages: because of their position as intermediary between otherwise disconnected contacts, the broker benefits from an information asymmetry that can be exploited to their advantage. His/her informational advantage allows him/her to play the role of “tertius gaudens”, that is the role of an agent who adds value through his/her intervention as an intermediary [BUR 95b, p. 604]. The intermediary can benefit from his/her informational advantage and decide on the dissemination of information both in quantity and quality in a strategic manner [BUR 95b, p. 605]. Figure 2.1 illustrates the position of intermediary.

Figure 2.1. The position of intermediary

Tertius’ strategies

As mentioned in the previous section, the intermediary agent (or broker) benefits from an informational asymmetry to their own advantage. Having access to information, they are able to identify situations where there will be an advantage in connecting otherwise disconnected agents. They are also in a position to play off the interests of one party against those of another [BUR 95a, p. 33]. The intermediary therefore gains control benefits [BUR 95a, p. 34]. Unlike informational benefits, control benefits require active intervention by the intermediary. In order to qualify an intermediary as a tertius, the two disconnected agents must be in a situation of competition or conflict, and therefore there must be tension between them [BUR 95a, p. 32]. Successful exploitation of the tertius position lies in connecting agents who are willing to negotiate, have sufficient comparable resources to understand and mutually recognize their expectations but would not enter into contact/negotiation without the presence of the tertius [BUR 95a, p. 33]. Thus, “without uncertainty, no tension, no tertius”. If none of the agents can dominate the situation by holding absolute authority (by taking control of the situation), there is an opportunity for the tertius to position him/herself as an intermediary in the negotiation for control [BUR 95a, p. 33]. It therefore plays the role of “tertius gaudens” (a term from [SIM 55]) [BUR 95a, p. 30; BUR 95b, p. 604]. The tertius is an entrepreneur [GRA 92, p. 6], a person who generates profit by placing him/herself among others and strategically exploiting the information at his/her disposal [BUR 95a, p. 34]. Figure 2.2 illustrates the position of the tertius.

Figure 2.2. The position of the tertius

The tertius can exploit not only existing structural holes, but can also create new ones. This requires proactive action; they must have detected an opportunity they wish to exploit [BUR 95a, p. 230]. These actions then change the boundaries of their social network or structure [BUR 95a, p. 231]. There are therefore two tertius strategies: the tertius that exploits existing opportunities, and the tertius that creates non-existent opportunities.

• A) The tertius that exploits existing opportunities

In this case, the tertius is an intermediary between at least two agents that are competing for access to the same resource or between at least two agents with conflicting requests. For example, let us consider a situation where two customers compete with a seller for the same commodity. The seller is in a position to act as an intermediary and has the opportunity to take advantage of this situation by playing off the customers’ price offers against each other [BUR 95a, p. 31]. In addition, agents may disclose information to the tertius in their own interest for the tertius to act in their favor [BUR 95a, p. 33].

By controlling the situation, the tertius thus has the choice of extracting value by negotiating relationships or adding value by strengthening relationships for subsequent profit [BUR 95a, p. 34]. They thus have the power to decide on the distribution of generated profit (or the organizational rent) [BUR 95a, p. 35].

• B) The tertius that creates new opportunities

Agents who have a lack of structural holes on their side of the relationship and are aware of the presence of structural holes on the other side of the relationship are structurally autonomous. Agents are freer to act as they wish and are able to be less careful about conforming to their environment [BUR 95, p. 195, p. 217 and p. 218]. They are better placed in terms of negotiations [BUR 95a, p. 45]. In the presence of structural holes, an agent runs the risk that other agents, if they are aware of it, will take advantage of them to threaten to replace them with another contact that provides them with the same resources and benefits. If this is the case, the agent with a structural hole on his/her side of the relationship must act strategically to eliminate this hole, either by differentiating him/herself, or by allying him/herself with other agents that are likely to be able to replace him/her [BUR 95a, pp. 44–45]. Three strategies are possible: changing the structure of the network by excluding an existing contact, integrating a new contact, or embedding the relationship into a second one where there is more control [BUR 95a, p. 233], for example by forming an alliance.

2.3.2. Applying the theory to our research question

We begin the application of the theory to our problem by completing the argument presented in the section on analyzing the role of PEFs in light of contractual theories (section 2.3.2.1). We then continue by completing the argument put forward in the section on the problem in light of knowledge-based theories (section 2.3.2.2).

2.3.2.1. Complement to the argumentation of contractual theories

According to the arguments put forward for contractual theories, which we wish to supplement, PEFs make it possible to establish a situation of trust, thanks to their simple presence, to their specific position as common agents in the case of intraportfolio alliances, and to their simple presence and reputation capital in the context of extraportfolio alliances.

An alliance between two or more companies implies permanent reciprocal exchanges between the agents. This establishes relationships of expectations and obligations. In order to cooperate, companies must have the opportunity to build relationships and trust that mutual obligations are honored during the relationship. Thus described, alliances may refer to one of three types of social capital, as described by Coleman. For a reciprocal relationship between agents to emerge, if the agents do not know each other beforehand, a preliminary social structure that allows companies to establish links and, in the longer term, trust, is necessary. To use terms closer to Coleman’s analysis, the conditions required for the emergence of social capital must be met. This means, in his view, a closed social structure.

In view of our problem, a question then arises on the role that a PEF can play when forming an alliance. In our discussion, we again distinguish between the cases of an intraportfolio alliance and an extraportfolio alliance.

2.3.2.1.1. Intraportfolio alliance

In the case of an intraportfolio alliance, the presence of a PEF makes it possible to close the social structure of the alliance that is not yet formed, which otherwise remains open. Let us consider an alliance that is supposed to form between two agents (Figure 2.3).

Figure 2.3. Structure of an alliance without the presence of a PEF

Here we face a dyadic relationship. Let us now introduce a PEF which, in the case of an intraportfolio alliance, constitutes the common third between the two agents who wish to form an alliance (Figure 2.4).

Figure 2.4. Structure in the case of an intraportfolio alliance in the presence of a PEF

We observe that there is a triadization or contextualization of the dyadic relationship: reciprocal exchanges (dyads) are embedded in a wider social system (triadization). Such an analysis can be found in anthropological texts (for example [LEB 75, p. 559]). These texts then describe two possible effects that arise from the triadization that occurs in a triadic relationship. This can either strengthen and stabilize the initial relationship, or it can inhibit or break it [LEB 75, p. 559].

We believe that, through Coleman’s thinking, this type of structure is a source of advantages in view of our problems. In our case, through their presence, PEFs create the required social capital for the obligation–expectation relationships in order for them to be established. The triadic relationship reinforces the emergence of relationships of trust. Here we can refer to the argument already made in the section on contractual theories. In the case of an intraportfolio alliance, companies that wish to form an alliance are backed by the same PEF. They already have a relationship of trust with the PEF and vice versa. These vertical relationships between the PEF and backed companies then serve as a basis for the emergence of horizontal trust between future partners in the alliance. Moreover, this closed structure allows for the establishment of norms that can sanction inappropriate behavior, such as uncooperative behavior within the alliance. There is the possibility of triadic control and sanctions. This, in turn, strengthens the establishment of trusting relationships [BUR 00, pp. 351–352].

But there is another advantage to this structure. Throughout the section on contractual theories, we argued that PEFs are capable of reducing the information asymmetry and thus transaction and agency costs. Burt’s structural holes analysis can shed some light on this. In the section on Burt’s analysis, we saw that an agent who acts as an intermediary between two agents that could enter into a relationship can take advantage of his/her situation and, in particular, allow the agents to reveal information that they would have withheld from the other party. The agents agree to disclose the information to the intermediary in their own interests so that the intermediary can act in their favor [BUR 95a, p. 33]. This situation seems particularly likely in our case. Companies that wish to form an alliance know the PEF that constitutes the intermediary and trust it. In addition, the PEF backs the company to ensure its development. In the case of an intraportfolio alliance, all future partners therefore have an interest in disclosing information to the PEF so that its actions are profitable for them. In addition, future alliance partners know and trust the PEF; it is the agent that delivers reliable information to them.

Nevertheless, by controlling the situation, the PEF also has the possibility of extracting value by intervening in the negotiation of relations [BUR 95a, p. 34]. In particular, it has the power to decide on the distribution of information [BUR 95a, p. 35]. However, we believe that a PEF tends to use information in the interest of companies (and therefore not to extract value for itself) for two reasons. First, it has an interest in the companies (it is a shareholder). Second, the alliance project is long term. If companies suspect that the PEF is using the information or its control position for its own interest, they will be inclined to get out of the situation. This is easy for them: they just have to communicate directly with each other within the alliance without going through the PEF. The latter no longer holds any informational advantage. Again, therefore, one of the characteristics of the closed structure described by Coleman is the ability to control or even sanction unwanted behavior and to set standards. This tripartite relationship, in which all parties are linked in the longer term by strong ties, makes it possible to restrict behavior that is inappropriate for everyone.

Let us now consider the case of an extraportfolio alliance.

2.3.2.1.2. Extraportfolio alliance

In the case of an extraportfolio alliance formation, at least one of the companies forming the alliance is backed by PE and another is not. Previously, we distinguished between two cases. The first refers to our analysis of the intraportfolio alliance. This is when the company not backed by PE is nevertheless known to the PEF, either because it is a former portfolio company of the PEF, or because the PEF had considered it but ultimately did not back it. The second case is the one of interest in this section because it does not refer to the analysis carried out for the intra-portfolio alliance. This is the case where the company outside the PEF’s portfolio has no relationship with the latter. Applying Coleman’s analysis, the following diagram can be drawn (Figure 2.5).

Figure 2.5. Structure in the case of an extraportfolio alliance in the presence of a PEF

In this case, we see that the social structure is not closed. There is no triadization of the alliance’s dyadic relationship. Nevertheless, the PEF can play a role, but this brings us back to the arguments already discussed within the contractual theories. We argued that in a situation where future alliance partners lack information about each other, they then resort to general information such as the reputation of the future partner or, in the case of young companies not yet having reputational capital, the reputation of the agents that maintain relations with these companies. Thus, the reputational capital of the PEF can play a role in establishing a relationship of trust that is favorable to the formation of the alliance.

PEFs therefore seem to play a more important role in the case of an intra portfolio alliance compared to an extraportfolio alliance. In the first case, the relationship of trust is established independently of the reputational capital of the PEF on the market, quite simply because it is a common agent, which the companies wishing to form the alliance are already familiar with [GRA 85, pp. 490–491]. The reputational capital of a PEF can be a bonus but it does not appear to be necessary to establish a situation of trust between the future alliance partners, whereas in the case of an extraportfolio alliance, this is a prerequisite. Therefore, we make the following hypothesis:

HYPOTHESIS 10.– All else being equal, the role of PEFs in the formation of alliances is stronger if the latter are formed between companies backed by the same PEF.

2.3.2.2. Complement to the argumentation of knowledge-based theories

Sociological network theories also complement the argumentation stemming from knowledge-based theories where the emphasis is placed on the resources and competences that PEFs can bring to the companies they back, as well as on their impact on the formation of an alliance. In addition, we also highlighted the cognitive coordination role played by the PEF when bringing together companies in an alliance. We wish to enrich these arguments with those put forward by sociological network theories.

Let us begin by completing the argument of the role of PEFs in identifying and taking into account growth opportunities for the companies they back, through the formation of alliances (section 2.3.2.2.1). We then continue by deepening our analysis of the role of PEFs as facilitators of exchanges during the initial meetings between future alliance partners (section 2.3.2.2.2).

2.3.2.2.1. The role of PEFs in identifying and addressing opportunities for growth through alliance formation

In the section on applying knowledge-based theories to our problem, we argued that alliance formation can result from the detection and implementation of growth opportunities on the part of the PEF and that the detection of opportunities is a function of the resources and competences possessed by the PEF. Sociological network theories emphasize that this detection depends not only on the resources, competences and experience of PEFs but also on the information and knowledge held by other agents. In view of our problem, this means that growth opportunities detected by PEFs depend on:

• – their own knowledge, competences and experience;
• – the knowledge, competences and experience of agents to whom PEFs have access, who they know possess the required knowledge, competences and experience;
• – the manner in which PEFs combine this information and knowledge with their own.

We then argued that by intervening in alliance formation, PEFs can be the basis for creating new knowledge through organizational learning within the agency relationship. Here, too, sociological network theories provide an additional explanation. They draw attention to the fact that in order for learning to occur, the social conditions necessary for its emergence must be met. This raises the question of whether PEFs can be a source of social capital that is conducive to learning and can lead to human capital growth.

In the two sections that follow, we first complete the argument of the role of PEFs in exploiting growth opportunities through the use of sociological network theories. We focus on the channels through which PEFs can know who holds what knowledge and skills, and how to access these channels. Second, still applying sociological network theories, we consider part of Coleman’s analysis and highlight the role of PEFs in building the social capital required for organizational learning that is inherent in exchanges between alliance partners.

The role of PEFs in detecting opportunities

Sociological network theories emphasize, among other things, that certain agents may have privileged access to certain information and can use this information asymmetry to their advantage. Thus, we wish to complete the argument that PEFs can detect and exploit growth opportunities for the companies they back and enable them to exploit these by forming alliances. Thus, we question the main channels through which PEFs have access to this privileged information. The following arguments apply to both intra- and extraportfolio alliances.

In addition to the fact that PEFs can actively select suitable companies from among the companies that are already in their portfolio [LIN 08, p. 1138], these informational advantages involve analyzing potential future portfolio companies and subsequently providing privileged access to information that is strategic in nature via the seats that PEFs hold on the board of directors and supervisory boards (or similar strategic boards) of the backed companies. Moreover, their participation in associations revolving around the field of PE constitutes a privileged source of access to information and knowledge held by other agents.

Because of their expertise and informational advantage, PEFs are in a position, on the one hand, to detect potential candidates who could form a strategic alliance. On the other hand, they can detect potential alliance candidates if the company they are backing has informed them of its plan to form an alliance. They can also identify a latent desire on the part of the company manager to encourage the development of his/her company through the formation of an alliance and act accordingly. After analyzing the portfolio companies, PEFs have information about their strengths, weaknesses, opportunities and threats, and they themselves have an interest in guaranteeing their prosperity and encouraging their development.

These developments lead to the following hypotheses:

HYPOTHESIS 11.– All else being equal, a PEF’s access to strategic information has a positive impact on alliance formation.

HYPOTHESIS 11a.– All else being equal, the number of companies in which the PEF has one or more seats on the board of directors or supervisory board has a positive effect on the formation of alliances between the companies it backs.

HYPOTHESIS 11b.– All else being equal, a PEF’s participation in associations (Afic (Invest France), Evca (Invest Europe), Unicer, etc.) has a positive impact on the formation of alliances between the companies it backs.

In short, PEFs are in a position where they have privileged access to information about various companies. They can help the companies they back to benefit from their informational advantage. Thus, we find the three types of informational advantages, as defined by Burt [BUR 95b, p. 602]:

• – access benefits: a PEF’s social network allows it to seize more opportunities compared to the social networks of innovative start-ups;
• – synchronization benefits: PEFs are a link between the holders of non-redundant information, in other words the companies in their portfolio, and thus occupy an intermediary broker position. This position is favorable to them insofar as PEFs are the first to identify companies that can meet the needs of the companies they finance through their skills. PEFs are then in a position to synchronize the needs and skills of different companies;
• – opportunity referral benefits: the richer a PEF’s social network is in structural holes, the more likely it is to be taken into account when new opportunities arise.

Control benefits: PEFs benefit from an information asymmetry that can be exploited to their advantage. This position enables them to play the role of “tertius gaudens”, and to benefit from their informational advantage by strategically deciding on the quantity and quality of the information they disseminate.

In connection with the role of PEFs in identifying and creating growth opportunities for the companies they back through alliance formation, PEFs can also be the basis for organizational learning, knowledge exchange and the creation of new knowledge. Let us explore this further using Coleman’s analysis of the role of social capital in the emergence of human capital.

The role of PEFs in building social capital for organizational learning

As Coleman’s analysis concludes, the presence of social capital may be favorable to the emergence of human capital (see also [ZAH 09, p. 258]). In our case, this human capital lies in the exchange of knowledge between the companies that participate in an alliance, in the organizational learning that follows and that leads, ideally, to the production of new knowledge. By enabling the companies they back to form alliances with other companies, PEFs provide an ideal ground for organizational learning. Moreover, the opportunity to participate in networking groups or forums also allows companies to meet and get to know each other. PEFs thus provide the companies they back with structures that, according to Coleman, constitute a form of social capital that is essential to the development of human capital. This applies regardless of whether the alliance is intra- or extraportfolio.

These rather brief developments are linked to the following section where we complete (always by referring to sociological network theories) the knowledge-based argumentation brought forward on the role of a PEF in facilitating the initial exchanges between future alliance partners.

2.3.2.2.2. The role of PEFs as facilitators of exchanges

In Section 2.2.2.1.2, in the discussion on applying knowledge-based theories, we argued that an alliance implies the transfer of skills and knowledge between its partners and that this transfer can prove to be costly if there is any mutual misunderstanding between the partners. A PEF can then play a role of cognitive coordination, making it possible to facilitate the initial exchanges between partners. Indeed, this role played by PEFs seems more significant during the first meetings between companies wishing to form an alliance. It decreases as companies get to know each other better, thus according to the time and intensity of exchanges. We then assumed (Hypothesis 8) that PEFs are able to facilitate exchanges during the initial meetings between future partners, which has a positive effect on the formation of alliances. Let us now explore these conclusions further using Coleman’s closed network analysis.

The plan is as follows. First, we wish to further develop the argument that the intervention of PEFs as facilitators of exchanges takes place particularly during the first meetings and decreases with time and the intensity of exchanges. Second, we again take Coleman’s analysis of the three forms of social capital, assuming that PEFs can only play this role of facilitator of exchanges if they make the necessary social capital available – that is, if they devote time and attention to their portfolio companies, to their needs and projects, which can lead to the formation of alliances.

Complementing the role of PEFs as facilitators of exchanges

In this section, we elaborate on the argument that the role of PEFs in coordinating exchanges between the companies participating in an alliance is particularly important during the initial exchanges. It decreases with time and intensity of exchanges. The following developments apply both to intra- and extraportfolio alliance.

Little by little, the ties and, subsequently, the structure of the network between companies that form an alliance and the PEF evolve. Initially, the structure is characterized by a structural hole; a structural hole that the PEF can cross by acting as a bridge. There is no tie between the companies wishing to form the alliance. On the other hand, a more or less strong tie exists between the PEF and at least one of the alliance partners. Over time, as agents interact, the structure evolves into a closed network where trust builds between the agents. The ties between all the agents are then strong. Generally speaking, the interaction between alliance partners and the PEF enables the agents to gain experience. Agents get to know each other, which strengthens the ties they have with each other. These interactions result in the emergence of rules and norms that enable coordination among agents within the structure [HOD 98, pp. 189–191; KOG 00, p. 407]. On the one hand, these interactions allow for a mutual understanding of the cognitive patterns of the alliance partners, or even a rapprochement of the cognitive patterns of agents, which facilitates the circulation of tacit information and knowledge. This reduces cognitive costs and, as a result, the mentoring role of the PEF. On the other hand, contacts reduce opportunistic behavior and facilitate trust [WIL 79, p. 240]. The more companies interact and get to know each other, the less they require the intervention of a PEF, given that trust is established.

The role of PEFs under social capital constraints

Sociological network theories draw attention to the fact that PEFs can only facilitate exchanges between alliance partners if they are able to devote sufficient time and attention to them, which constitutes a form of social capital [COL 88, p. 108]. Indeed, the formation of alliances and the facilitation of exchanges during the initial meetings are not the primary vocation of PEFs, although they are able to provide these services. The investment manager who monitors the company must therefore pay attention to this and be able to devote part of their time to it. This time will depend, on the one hand, on the number of portfolio companies managed by this PEF employee. On the other hand, the geographical distance between managers and companies backed by the investment manager can also have an impact. If the distance is large, physical visits may be less frequent and exchanges are therefore less intensive. Therefore, we make the following subhypothesis:

HYPOTHESIS 8a.– All else being equal, (1) a high number of portfolio companies to be managed by an investment manager (lack of time) and (2) a large geographical distance between the participating companies and the investment manager in charge of their portfolio have a negative effect on the formation of alliances.

In terms of lack of time, this argument could also be part of the contractual approach: the higher the number of portfolio companies to be managed by an investment manager (lack of time), the higher the cost of disciplinary intervention, which has a negative effect on the formation of alliances.

An argument that is based purely on a knowledge-based analysis, however, also suggests a positive effect between the number of portfolio companies to be managed by an investment manager and the formation of alliances. The more portfolios are managed, the more ideas can be generated. It is therefore also a question of testing the following alternative impact:

“(1) A high number of portfolio companies managed by an investment manager (source of ideas) has a positive impact on alliance formation (Hypothesis 8a, knowledge-based/cognitive argument)”.

By testing Hypothesis 8a (later), we can test whether a lack of time and a certain geographical distance between agents can hinder the role played by PEFs in strategic alliances. However, this does not make it possible to see whether, when an investment manager has time because he/she monitors relatively few portfolio companies and the geographical distance that separates him/her from the manager of the backed company is small, it follows that he/she actually devotes this time to paying the required attention to the formation of alliances and to intervening as a facilitator of exchanges at the time of initial exchanges between future partners in an alliance. Consequently, let us complete our hypothesis by adding the following subhypothesis:

HYPOTHESIS 8b.– All else being equal, (1) a limited number of portfolio companies to be managed by an investment director and (2) a small geographical distance between the participating companies and the investment director who is in charge of their portfolios, have a positive effect on the formation of alliances.

It should be noted that we will have to take into account that the number of portfolio companies managed by various PEFs also depends on certain factors, notably the stage in the lifecycle of the company, so the category of the company (seed, development, turnaround, etc.) in which the PEF invests. This also depends on the type of stake the PEF holds (majority or minority). In general, PEFs that hold majority stakes back a smaller number of portfolio companies than those that hold minority stakes.

Before concluding, it is worth noting again that the argumentation derived from sociological network theories is not limited to supplementing that of knowledge-based theories, it also allows for a different explanation. It makes it possible to focus on the advantage conferred by a social structure on the agents who act within its framework. Thus, by enabling the companies they back to form an alliance, PEFs embed them in a new relationship. This then changes the boundaries of the social network (or structure) within which companies operate [BUR 95a, p. 231]. By uniting in an alliance, companies can, on the one hand, strengthen their bargaining power vis-à-vis other market players, for example, with key suppliers. On the other hand, they may source from customer–supplier relationships or jointly develop resources for which the cost and development time would have been too high if they had done it alone. Companies thus reduce their dependence on resources held by other market players. However, as we have just noted, the arguments discussed here refer to the theory of resource dependence and incorporate arguments that approximate the power paradigm which, as noted at the beginning of this section, go beyond the purpose of this study. We therefore bring the development of this perspective to a close and refer the interested reader to the extensions proposed at the end of this book.

2.3.3. Review of the additional arguments brought forward by the use of the concept of social capital

In relation to the concept of social capital, alliances can be seen as part of this capital for the companies involved. On the one hand, this relationship presents an opportunity for the company concerned, which provides access to resources beyond its borders and simultaneously enables it to achieve a certain legitimacy in the face of its external environment. On the other hand, relationships with other companies can hinder business development.

Burt and Coleman’s studies allow us to complete the following arguments:

• – in the application of contractual theories, PEFs can play a role in building trust;
• – in the application of knowledge-based theories, PEFs can:
  • - identify, exploit and implement growth opportunities for the companies they back by forming alliances. As a result, they can also be the basis for the creation of new knowledge;
  • - facilitate initial exchanges between future alliance partners.

Let us clarify our first point. Coleman’s analysis highlights that, in the case of an intra-portfolio alliance, the presence of a PEF makes it possible to create a closed structure with the SMEs that form the alliance, because of a PEF’s specific position as a common agent. The result is a structure that is conducive to the establishment of relationships of trust between agents. In the case of an extraportfolio alliance, however, the PEF’s relations with potential alliance partners do not allow the structure to be closed; the relationship between PEFs and the external partner being open. The potential alliance partner outside the PEF’s investment portfolio can nevertheless rely on the PEF’s reputation capital to reassure itself on the validity of the commitments of the SME backed by PE. The role of PEFs in strategic alliances is thus stronger in the case of intraportfolio alliances (Hypothesis 10).

With regard to the role of PEFs in identifying, building and implementing growth opportunities for the companies they back through the formation of alliances, the concept of social capital provides additional arguments. It emphasizes that the member companies in an alliance depend not only on the knowledge, skills and experience of PEFs, but also on the skills, knowledge and experience of the agents to which the PEFs have access, as well as on how PEFs combine them with their own know-how. This raises the question of identifying the channels through which PEFs can access this information. We believe that the seats that PEFs hold on the board of directors and supervisory boards (or similar strategic boards) constitute one of these channels (Hypothesis 11a), as does the participation of PEFs in associations linked to PE or with themes related to the activity of backed companies (Hypothesis 11b).

Through the detection, construction and implementation of growth opportunities for companies backed by the formation of alliances, PEFs can also be the basis for organizational learning and, thus, the creation of new knowledge for backed companies. This is consistent with Coleman’s analysis that emphasizes the need for social capital for the proper development of human capital. By encouraging the formation of alliances and by proposing forums or meeting clubs for their portfolio companies, PEFs make such social capital available to them, which is favorable to exchanges between companies.

Finally, as a last point, the concept of social capital has made it possible to focus on the fact that PEFs (according to the knowledge-based argument) can facilitate initial exchanges between SMEs that form an alliance but can only do so successfully if they devote the required time and attention to it. According to Coleman, time and attention are a form of social capital. We argue that the time that PEFs can devote to facilitating exchanges between future alliance partners depends on the number of portfolio companies to be managed by an investment manager, and the geographical distance that separates this investment manager from the managers of SMEs wishing to form the alliance (Hypotheses 8a and 8b). Table 2.6 summarizes these developments.

Table 2.6. Complementary arguments provided by applying the concept of social capital to the roles of PEFs from the perspective of contractual and knowledge-based theories

		Concept of social capital Coleman’s analysis and Burt’s structural hole theory
		Complement to the contractual argumentation		Complement to the knowledge-based argumentation
Vision of alliances		The alliance as: – a portion of the company’s social capital; – on the one hand, it presents an opportunity that provides access to resources beyond the company’s borders and simultaneously enables it to achieve a certain legitimacy in the face of its external environment; – may also present an obstacle to the development of the company.
Roles of PEF that theory complements		– Role as a mechanism establishing a relationship of trust. – Role of the reputational capital of PEFs in establishing a relationship of trust.		– Role of PEFs in detecting, constructing and implementing growth opportunities for companies they back through the formation of alliances. – Role of PEFs in the creation of new knowledge for the companies they back through the formation of alliances. – Role of PEFs in facilitating initial exchanges between future alliance partners.
Questions asked		– Analysis of the structure of an intraportfolio alliance. – Analysis of the structure of an extraportfolio alliance.		– Through what channels do PEFs have access to strategic information? – Can PEFs provide the social capital needed to develop new knowledge for the companies they back? – Do PEFs provide the necessary social capital so that they can play the role of facilitator of the initial exchanges between future partners in an alliance?
		SME perspective	PEF perspective	SME perspective	PEF perspective
Answers provided	Intra	Through its presence and position within the structure, the PEF establishes trust because it: – constitutes a common trustworthy agent; – allows the structure to be closed, which is favorable to the establishment of a relationship of trust; – allows information to be revealed.	Because of the closed structure it forms with alliance partners, it has no interest in using the information provided by SMEs to their detriment, in principle.	PEFs provide the social capital required for organizational learning and the creation of new knowledge for the companies they back by: – forming alliances; – making clubs and meeting forums available for their portfolio companies.	– Access to strategic information via: - seats on strategic boards; - participation in associations. – PEFs only facilitate exchanges if they offer the required social capital: time and attention. This depends on: - the number of portfolio companies to be managed by the investment manager; - the geographical distance between the investment manager and the backed company manager.
	Extra	– The PEF does not close the structure. – The reputational capital of the PEF makes it possible to establish a relationship of trust.

2.4. The role of PEFs: a summary

Let us now bring together the arguments from the different theoretical frameworks we have alluded to in order to provide an overall explanation of the role or, rather, the roles of PEFs in the formation of strategic alliances for the companies they back.

Let us again take two cases: intra- and extraportfolio alliances. Each discussion (intra- and extraportfolio alliance) includes four points. These provide an explanation for the numbers in the corresponding figures.

2.4.1. Intraportfolio alliances

Let us begin with the case where all alliance partners are backed by the same PEF. The idea of forming an alliance either comes from the PEF or from one of the companies that the PEF backs (see number 1 in Figure 2.6).

2.4.1.1. The idea stems from the SME

If the idea of an alliance comes from the SME, the company manager discusses it with the PEF. The latter then gives an opinion as to whether the alliance project seems to be in line with the pursued strategy or not. At this stage, either the manager of the SME backed by the PEF already has an idea of an alliance partner who also happens to be a portfolio company of the PEF and wishes to be put in contact through the PEF, or the PEF brings suggestions of potential partners, seeking these out within its investment portfolios. The idea of who could potentially work with the SME in question arises from knowledge of the companies with which it works. Moreover, during regular meetings or telephone conversations with the managers of backed companies, the PEF can discuss or suggest the idea. Once a potential partner has been found, and after having discussed it with the managers of the companies taking part in the alliance, the PEF obtains their agreement to make an initial appointment. During this initial meeting, the PEF can then choose to be present or not, according to its own time constraints and the wishes of the managers (Hypotheses 8, 8a and 8b). Once an alliance has been formed, the PEF can intervene ex post if necessary, and if the managers so wish. This intervention is all the more important if the PEF holds a majority stake in the backed companies (Hypothesis 4).

2.4.1.2. The idea stems from the PEF

A PEF can also come up with the idea of forming an alliance for its portfolio companies. As early as during its analysis of financing potential portfolio companies, PEFs can consider possible collaborations between future portfolio companies and already backed companies. There are multiple reasons that will push a PEF to consider such a networking activity right from the start of backing these companies:

• – investors in the capital of a PEF such as the State or a region, or its involvement in a competitiveness cluster, can promote the formation of alliances (Hypothesis 3);
• – it may include a service provided by the PEF in order to differentiate itself on the PE market (Hypothesis 9);
• – by analyzing investment candidates, a PEF can detect a growth opportunity for a company applying for financing and for a company already in its portfolio and suggest this alliance (Hypothesis 7).

PEFs are also able to identify good practices within their portfolio companies, whether in terms of work organization, coordination patterns or levels of employee participation in decision-making [PRA 90; TEE 97, p. 518].

They can then initiate alliances, within which companies create new knowledge by exchanging and recombining their existing knowledge through collective organizational learning.

A PEF may also have the idea of linking two or more companies that it already backs, for example, because the development of the activity of these companies are evolving in such a way that an alliance may be favorable, or because the PEF (which inquires about changes in the markets of the backed companies) can detect an opportunity or an environmental risk that could be seized/eliminated by forming an alliance (Hypotheses 11, 11a and 11).

Again, the company being backed could face an organizational problem that is similar to those that other SMEs in a PEF’s portfolio of backed companies have been able to solve. The PEF that holds this information can then put the managers in contact with each other.

In this case, the PEF must first share its idea of an alliance with the managers of the backed companies and then propose an appointment to bring the agents together. This can be done during regular meetings between the backed company managers (or within a networking club or forums) if the PEF proposes this, or a meeting initiated by the PEF for the sole purpose of bringing these managers together (Hypothesis 8).

As indicated by number 2 in Figure 2.6, depending on who suggested the idea, either the PEF shares it with the relevant portfolio companies or the SME shares it with the PEF. Both discuss it to ensure that the alliance is in line with the strategy pursued by the SME and to find a partner. Once the idea of a partner within the PEF portfolio is found, the PEF approaches this potential partner to see if it is tempted by the alliance.

As indicated by number 3 in Figure 2.6, the SME contacts a future external alliance partner. The PEF can assist its portfolio company and advise it, being present if it wishes. The PEF’s status as an investment company and its reputation can reinforce the seriousness of its portfolio company and thus reduce any ex ante distrust from the external partner. In doing so, the PEF reduces transaction costs (Hypothesis 1).

It can then intervene in the initial exchanges between companies wishing to form an alliance and play a role of cognitive coordination by advising and supporting its portfolio company (Hypothesis 8). This cognitive coordination role depends on the time and attention that the PEF can devote to it (Hypotheses 8a and 8b).

As indicated by number 4 in Figure 2.6, a triadic relationship is formed between the PEF – a common agent – and the companies forming the alliance. The structure is closed: all the agents end up with strong ties between each other that promote the circulation of information. This structure is favorable to the establishment of trust links between agents, which makes it possible to reduce transaction and agency costs (Hypothesis 10).

It remains to be noted that the role of PEFs in strategic alliances is more significant for unlisted companies (Hypothesis 6) and that, depending on the contractual or knowledge-based argumentation, PEFs may be reluctant or not to have their portfolio companies form too many alliances (Hypotheses 5a and 5b).

Figure 2.6 shows a diagram of the roles of PEFs in the formation of intraportfolio alliances.

Figure 2.6. Summary of the roles of a PEF in the formation of intraportfolio alliances

2.4.2. Extraportfolio alliances

Let us continue with the case where only one of the alliance partners is backed by the PEF. Again, the idea of forming an alliance either comes from the PEF or from one of the companies that the PEF backs (see number 1 in Figure 2.7).

2.4.2.1. The idea arises from the SME

If the idea of an alliance comes from the SME, the company manager discusses it with the PEF. The latter then gives its opinion: whether the alliance’s project seems to be in line with the pursued strategy or not. At this stage, either the manager of the SME backed by the PEF already has an idea of an alliance partner, or the PEF comes up with suggestions of potential partners. The idea of a partner that could potentially work with the SME in question may come from former potential portfolio companies that the PEF has considered but where the candidate company was ultimately not selected for financing. The idea may also come from the PEF having done a market study or because it is already following the market in which the SME is operating (Hypotheses 11 and 11b). Once a potential partner has been found, the PEF can encourage and advise the manager of the backed SME to contact the potential partner. During this contact, the PEF may or may not be present, depending on its time constraints and the wishes of the managers (Hypotheses 8 and 8a). It is also possible that the meeting will take place on its premises. However, there may be potential conflicts of interest between the PEF and the unknown alliance partner. It is likely that the PEF will be reluctant to have too many extraportfolio alliance formations if it did not come up with the idea of the alliance (Hypothesis 5a). Once an alliance is formed, the PEF can intervene ex post by advising its portfolio company.

2.4.2.2. The idea stems from the PEF

The PEF can also come up with the idea of forming an alliance for its portfolio companies. Various reasons can push a PEF to consider such a networking activity right from the start of backing a company: investors with PEF’s capital such as the State or a region or its possible involvement in a competitiveness cluster may promote the formation of alliances (Hypothesis 3); this may be a service provided by a PEF in order to differentiate itself on the PE market (Hypothesis 9); by analyzing the portfolios, a PEF may detect a growth opportunity for the company applying for financing and a company already in its portfolio and suggest the idea of an alliance (Hypothesis 7). The PEF may also have the idea of linking two or more companies that it already backs, for example because the development of the activity of these companies has evolved in such a way that an alliance can be favorable or because the PEF, which monitors changes in the markets of backed companies, can detect an opportunity (an environmental risk) that could be seized (reduced) by forming an alliance (Hypothesis 7b). PEFs are also able to identify good practices adopted within their portfolio companies, whether in terms of work organization, coordination methods, or levels of employee participation in decision-making. They can then initiate alliances in which companies create new knowledge by exchanging and recombining their existing knowledge through collective organizational learning.

In this case, the PEF must first share its idea with the managers of the backed companies and then propose an appointment to meet and bring the agents together. This can be done by meeting the backed company managers during regular meetings between managers (or within a networking club or forums) if the PEF proposes this, or during a meeting initiated by the PEF for the sole purpose of bringing these managers together (Hypothesis 8).

As indicated by number 2 in Figure 2.7, depending on who came up with the idea, either the PEF shares it with its portfolio company or the SME shares it with the PEF. Both discuss it to ensure that the alliance is in line with the strategy pursued by the SME and to find a partner.

As indicated by number 3 in Figure 2.7, the SME contacts a future external alliance partner. The PEF can assist its portfolio company and advise it, bringing its presence if it wishes. The PEF’s status as an investment company and its reputation can reinforce the seriousness of its portfolio company and thus reduce any ex ante mistrust from the external partner (Hypothesis 1). In doing so, it reduces transaction costs. The PEF can then intervene in the initial exchanges between companies wishing to form the alliance and play a cognitive coordination role by advising and supporting its portfolio company (Hypothesis 8). This cognitive coordination role depends on the time and attention that the PEF can devote to its portfolio company (Hypotheses 8a and 8b).

As indicated by number 4 in Figure 2.7, in this case, the PEF only has a strong relationship with one of the two companies forming the alliance. We then see that the social structure is not closed. This structure thus does not offer the social capital required for the emergence of norms, sanctions and the establishment of trust. However, if it is not possible to observe the quality and commitments of a company, external agents may rely on the reputation of agents who operate with the company in order to assess its quality. The reputational capital of the PEF can then help to establish a relationship of trust and thus have a positive impact on alliance formation (Hypothesis 1). However, a PEF only plays this quality certification role for the SME it finances if it has a certain reputation capital. This is built gradually through accumulation of experience and performance. A young PEF with little or no reputational capital can then link its portfolio companies with established and reputable partners on the markets (Hypothesis 2).

Again, it should be noted that the role of PEFs in strategic alliances is more significant for unlisted companies (Hypothesis 6) and that, depending on the contractual or knowledge-based argumentation, PEFs may be reluctant or not to have their portfolio companies form a large number of alliances (Hypotheses 5a and 5b).

Figure 2.7 shows a diagram of the roles of PEFs in forming extraportfolio alliances.

Figure 2.7. Summary diagram of the roles of a PEF in the formation of extraportfolio alliances

Figure 2.8 summarizes the arrangement of the different theoretical frameworks. Two main theoretical approaches are used: the contractual approach and the knowledge-based approach. The first allows us to consider the role of PEFs in alliance formation with a view to reducing losses in value (agency costs or transaction costs) at a given time. This approach is mainly based on the use of contractual theories (TCT and PAT), which highlight the role of PEFs in reducing transaction and agency costs that essentially result from a pronounced information asymmetry, the presence of uncertainty and potential conflicts of interest between agents. The concept of social capital has enabled us to complete our analysis by contextualizing, mainly, the role of PEFs in establishing trust between agents, by analyzing the type of tie that connects them according to the structures of intra- or extraportfolio alliances.

In summary, eight hypotheses derive from the contractual approach, six of which stem directly from the use of contractual theories (TCT and PAT: Hypotheses 1–6), and two from the additional analysis carried out using the concept of social capital (Hypotheses 8a and 10).

The second approach (knowledge-based) allows us to consider the role of PEFs in the formation of alliances from a long-term value creation perspective based on the concept of knowledge (as opposed to contractual theories, which are limited to the concept of information and costs that are directly related to information asymmetries). It is mainly based on the use of knowledge-based theories in a dynamic perspective and highlights the roles of PEFs in the creation of opportunities for growth, new knowledge creation, and facilitation of initial exchanges between future partners in alliances (cognitive coordination). Knowledge-based theories allow for the contribution of knowledge, resources and competences from PEFs to be taken into account. The concept of social capital recognizes that PEFs may also have privileged access to knowledge, resources and competences possessed by other agents. The combination and exploitation of their own knowledge, resources and competences and those held by other agents are then sources of future organizational rents. The concept of social capital again allows for further analysis by emphasizing that PEFs only offer these services (human capital) if they devote sufficient time and attention to the provision of this type of service (social capital). In total, seven hypotheses arise from this approach, including three from knowledge-based theories (Hypotheses 7a, 7b and 9) and four from the concept of social capital (Hypotheses 8a, 8b and 11a, 11b), which are used in a complementary way to bring out additional hypotheses while maintaining coherence with the knowledge-based argumentation.

As previously mentioned, Figure 2.8 outlines the connections between the two approaches and the different theoretical frameworks. This representation is followed by a table in which we summarize the various hypotheses in relation to the theoretical framework from which they derive and the related argument that allows them to be expressed (Table 2.7).

Figure 2.8. Summary of the connections between the two approaches and the different theoretical frameworks

Table 2.7. Summary of the hypotheses

	Nature of the argumentation	Theoretical framework	Argument	Summary of hypotheses	Expected sign
H1	Disciplinary	TCT / PAT	Role of trust in reducing transaction and agency costs	The reputational capital of PEFs strengthens the trust mechanism and thus has a positive impact on alliance formation.	+, especially for extraportfolio alliances
H2	Disciplinary	TCT / PAT	Role of PEFs’ interest and trust in reducing transaction and agency costs	For a PEF with a low reputation capital, the formation of alliances of SMEs that they back with partners that have a reputation capital constitutes a complementary or even substitutable mechanism to the role of certification played by a renowned PEF.	+ on linking with reputable partners in alliances
H3	Disciplinary	PAT	Investor’s interest in forming alliances	The presence of the State or a region in the capital of a PEF or its involvement in a competitiveness cluster positively affects the formation of alliances.	+ and more important role if the PEF is a JSC/VC
H4	Disciplinary	PAT	Monitoring, control to reduce agency costs	Holding of a majority stake by PEFs in the companies it backs has a positive impact on the formation of alliances.	+
H5a/b	Disciplinary (a)/knowledge-based (b)	PAT (a) / KBV (b)	Conflicts of interests (a)/learning (b)	In the presence of a PEF, the number of alliances formed by a company has (a) a negative impact (contractual argumentation)/(b) a positive impact (knowledge-based argumentation) on the formation of a new alliance.	–/+ (contractual / knowledge-based argumentation)
H6	Disciplinary	PAT / TCT	Relative importance of PEFs as a governance mechanism	The role of PEFs in strategic alliances is higher for unlisted companies.	+
H7	Knowledge-based	KBV	Contribution of competences	PEFs detect and enable their portfolio companies to seize growth opportunities by promoting the formation of strategic alliances.	+
H7a	Knowledge-based	KBV		We expect increased formation of intrasector alliances in the presence of regional or sector-specific PEFs.	+ on the formation of sectoral/regional alliances
H7b	Knowledge-based	KBV		We expect an increased formation of alliances that allow international development of the companies taking part if the PEF’s investment portfolios are in different countries.	+ on the formation of alliances with an international nature
H8	Knowledge-based	KBV	Provision of competences/cognitive coordination role	The presence of a PEF facilitates exchanges between potential partners in an alliance comprising at least one SME backed by this PEF, and, therefore has a positive impact on the formation of the alliance.	+
H8a	Disciplinary/Knowledge-based	Social capital – PAT/KBV	Role of social capital in human capital: lack of time, attention/access to information (source of ideas)	(1) A high number of portfolio companies to be managed by an investment manager (lack of time) and (2) a large geographical distance between the portfolio companies and the investment manager in charge of the portfolio have a negative effect on the formation of alliances. Alternatively, a high number of portfolio companies to be managed by an investment manager (source of ideas) can have a positive effect on alliance formation.	–/+ (contractual/knowledge-based argumentation)
H8b	Knowledge-based	Social capital	Role of social capital within human capital: the concept of time/attention	(1) A limited number of portfolio companies managed by an investment manager and (2) a short geographical distance between the portfolio companies and the investment manager in charge of the portfolio have a positive effect on the formation of alliances.	+
H9	Knowledge-based	KBV	Strategic direction	Providing contacts for alliance formation is a way for PEFs to differentiate themselves on the PE market.	+
H10	Disciplinary	Social capital – TCT/PAT	Role of a structure that consists of strong ties for building trust in order to reduce transaction and agency costs	The role of PEFs in forming alliances is stronger if the latter are formed between companies backed by the same PEF.	+
H11	Knowledge-based	Social capital	Privileged access to sources of strategic information	The PEF’s access to strategic information has a positive impact on alliance formation.	+
H11a	Knowledge-based	Social capital		The number of companies in which the PEF holds one or more seats on the board of directors or supervisory board has a positive effect on the formation of alliances between companies it backs.	+
H11b	Knowledge-based	Social capital		The PEF’s participation in associations (Afic/France Invest, Evca/Europe Invest, Unicer, etc.) has a positive effect on the formation of alliances between the companies it backs.	+

1 The term “enterprise” is linked to the verb “enterprising”, which means “to show initiative”. In this sense, Penrose defined the term “enterprise” as a psychological predisposition on the part of individuals to take advantage of an opportunity in the hope of making gains and, in particular, the willingness to invest effort and resources in speculative (uncertain) activities (Penrose, 1995, p. 33). In her book, Penrose assumed that a company’s management is ready to seek growth opportunities, that this management has both the managerial and entrepreneurial skills and is willing to implement them (Penrose, 1995, p. 32).

2 Penrose writes [PEN 95, p. 85]: “Unused productive services are, for the enterprising firm, at the same time a challenge to innovate, and an incentive to expand, and a source of competitive advantage. They facilitate the introduction of new combinations of resources – innovation – within the firm”. She adds (p. 86): “No firm ever perceives the complete range of services available from any resource, because the range of services recognized is for the most part confined by the management’s existing ideas as to possible combinations”. Her application of the entrepreneurial function differs, however, from that of Schumpeter in that it is applied at the enterprise level and not at the level of the whole economy.

3 Nelson and Winter did not consider an entrepreneur in their analysis, but a large managerial company. They are thus based more on Schumpeter’s work, Capitalism, Socialism and Democracy (1942), than on the one from 1911, Theory of Economic Development. Whereas in Schumpeter’s work on the theory of economic development, the aim of the analysis was the entrepreneur, and enterprises constitute the place of execution of new combinations of existing factors of production [SCH 35, p. 319], greater importance is given to the large enterprise in his 1942 publication. This goes hand in hand with economic progress [SCH 75, p. 176]. It represents the entity required to achieve economic progress [SCH 75, p. 166]. Nelson and Winter [NEL 82, p. 278] pointed out that more importance is given to industrial research laboratories (mainly in large firms) as a central element of the innovation process, than to the entrepreneur who played the determining role in Schumpeter’s 1911 publication.

4 These include the following properties: asset mass efficiencies, time compression diseconomies, interconnectedness, asset erosion and causal ambiguity.

5 Pathway dependence takes into account the fact that the direction in which a company can develop depends not only on its current position, but also on its track record. A company’s past thus becomes important and conditions its future behavior [TEE 97, pp. 522–524]: a company’s past experiences determine the opportunities perceived by the management according to the environment it faces [PEN 95, TEE 97, p. 524].

6 According to Coase’s analysis, whether one goes through the market or through the firm depends on a comparison of the costs generated by the two coordination methods for the same transaction. The costs of coordinating production factors within one firm must be compared with those resulting from market coordination or from the sum of production and coordination costs incurred by another firm [COA 91a, p. 47; COA 91b, p. 54 and 59; COA 91c, pp. 65–67 and 73]. Similarly, a firm grows/can be expanded until the (marginal) cost of processing an additional transaction within the firm exceeds the cost of processing the same transaction on the market or within another firm [COA 91b, p. 48]. It is therefore a question of comparing the total costs incurred to carry out a transaction via the market (transaction costs) with the total costs incurred by a firm (management and production costs) [DEM 88, pp. 161–166 and 174].

7 Coase clearly sees the firm as a nexus of contracts [COA 91b, p. 56]. He indicated that a type of coordination can be called “enterprise” if the central agent, coordinator or organizer maintains contact with a number of factors for the activities he coordinates [COA 91c, p. 68] However, he pointed out that his analysis does not explain the structure of a firm’s production function [COA 91c, pp 67 and 73] and that he has therefore confined himself to explaining a firm’s existence on the market without analyzing why the costs of production coordination may differ from one firm to another. Coase himself said that he still wishes to undertake this project [COA 91c, p. 73].

8 “…. the firm is a specialized substitute of a market for team use of inputs; it provides superior (i.e. cheaper) collections and collation of knowledge about heterogeneous resources”.