part0010

The PER of a share is a measure of how much an investor is being asked to pay for future growth and how much (i.e. how many times EPS) other investors have paid in the past. It is far and away the most widely used measure of how cheap or expensive a growth share is despite the fact that it is only a one-dimensional measure. Far more meaningful, to my mind is the relationship between the PER of a company and its expected rate of EPS growth. I call this the price-earnings growth factor – PEG for short. As far as I am aware, the PEG is the first attempt to analyse systematically this important relationship and I am convinced that it is an invaluable investment tool.

This PEG is calculated by dividing the prospective PER of a share by the estimated future growth rate in EPS. Say a company is growing at 12% per annum and has a prospective PER of 12, its PEG is 12⁄12 = 1.0. If the growth rate were a much more attractive 24%, the PEG would be 12⁄24 = 0.5, and if the growth rate were a relatively poor 6%, the PEG would be 12⁄6 = 2.0.

Shares with PEGs of over one tend to be unappealing, shares with PEGs of about one are worthy of consideration and at well under one they are usually worth examining in much more detail with a view to purchase. In early 1996, the average PEG was about 1.5, having been nearer to 1.0 before the last bull market run. This historically high average PEG might be telling us that the bull market is nearing an end, but while I can find attractive stocks on PEGs of 0.75 or under, I am happy to continue investing in a highly selective way. However, I would not buy the market as a whole, because PEGs of 1.5 are too rich for my blood.

PEGs only apply to growth companies. The substantial EPS increases which are forecast by many cyclical companies are in most cases recoveries to previous levels or a step towards those levels. Applying the PEG concept in relation to EPS increases of this kind would produce absurd and meaningless statistics.

The attraction of buying shares with low PEGs is that they provide the element of safety normally associated with low-risk investments without sacrificing the upside potential of growth stocks in the early stages. Take a share on a PEG of 0.5 with a PER of 12 and anticipated EPS growth of 24% per annum. Even if next year’s profits disappoint a little, the share price in unlikely to fall much as the multiple is low and the shares are already very cheap in relation to the original forecast and the market as a whole.

The upside potential is far more interesting, even on the modest assumption that the forecast is just met. If EPS at the time of purchase were 10p and the shares at 120p were on a PER of 12, the expected 24% increase in earnings would lift EPS from 10p to 12.4p. Then, on a maintained PER of 12, the shares would rise by 24% to 148.8p (12 x 12.4p). However, it is extremely likely that the market would begin to wake up to the fact that the shares were undervalued. Investors would realise that the PER should be much higher and, in the weeks that followed the announcement of the results, the PER could easily improve to 18 times earnings and still leave the shares cheap relative to their growth rate.

On that basis, the shares would appreciate to 18 x 12.4p = 223.2p, a gain of 103.2p against the original purchase price of 120p. The interesting point is that only 28.8p (12 x 2.4p) of the profit of 103.2p arose from the increase in EPS, whereas 74.4p resulted from the status change in the PER. It is the possibility of a status change that investors should constantly be seeking when investing in companies with low PEGs.

Low PEGs work best when the PERs are in the 12-20 range and EPS growth rates are between 15% and 25%. Some of the best bargains are found amongst shares growing at about 25% per annum on a PER of 15, giving a PEG of 0.6. Exactly the same PEG would result for a growth rate of 50% and a PER of 30, but the important difference is that a growth rate of 50% is not sustainable in the longer term.

The PEG factor is used as a calibrated measure to give an accurate fix on the comparative attractiveness of each company relative to both the market as a whole and other companies. It is an invaluable sieve for reducing the whole universe of growth shares to a small number of candidates for further sieving. The lower the PEG, the fewer the number of companies that will manage to get through the sieve.

For comparisons between PEGs to be meaningful, it is crucially important that the data from which they are calculated (the shares’ PERs and growth rates) cover exactly the same period. It makes no sense to compare the growth statistics of a company with a March year end with another whose financial year ends three months earlier in December. Why? Because the business and economic world can change dramatically in a week, let alone three months. By adjusting the data for all companies to a common basis, comparisons are much more revealing and are always up to date and dynamic.

Think of it as the investment equivalent of a mother with an eighteen-month-old baby, Rosemary, and another with a nine-month-old one, Julie, comparing notes about their babies’ progress. Rosemary’s mother tells Julie’s mother that Rosemary has just learned to recite a poem. The fact that nine-month-old Julie is unable to do so would not in any way imply that she is backward. The only way to compare the relative progress of the two babies is to check exactly what each of them is capable of at exactly the same age. The principle is just the same with companies: the period during which their results are being compared must be identical.

To meet this important objective, REFS presents company statistics such as the dividend yield, PER, growth rate and PEG on a rolling twelve months ahead basis . To understand what this means, imagine a company, ABC, which has a December year end. Obviously, on 1 January 1996 the forecast that best describes ABC’s prospects in the year ahead is the one for the year to 31 December 1996. Equally, on 1 January 1997 it is the forecast for the year to 31 December 1997. But how do you value the company on 1 March 1996? On that date, the year ahead is covered mainly by the 1996 forecast supplemented by a small part of the 1997 estimate. To be exact, it is 10⁄12 of the 1996 figure and 2⁄12 of that for 1997.

To see how this works in practice, consider these forecasts for ABC at a time when its shares are trading at 200p:

The PER is calculated simply by dividing the share price of 200p by the rolling twelve months ahead EPS of 10.83p to give a twelve months ahead PER of 18.5.

The future EPS growth rate has to be calculated in two stages. First, it is necessary to establish as a base what EPS was for the twelve months preceding 1 March 1996. The calculation is made in this way:

The 8p in the first brackets is the historic EPS for the year ended 31 December 1995. For ten of the preceding twelve months this was the EPS base that was being improved upon. The 10p in the second set of brackets is the forecast EPS for 1996, which contains the two months needed to top up the ten in 1995 to cover a full twelve-month period preceding 1 March 1996.

The second stage is to calculate the percentage growth in EPS for the twelve months following 1 March 1996. Here is the calculation:

The rolling twelve months ahead PEG is calculated by dividing the prospective PER of 18.5 by the future growth rate of 30% to give 0.62.

The rolling twelve months ahead dividend is calculated in the same way as EPS: 10⁄12 of the 5p for 1996 plus 2⁄12 of the 7p for 1997, to give 5.33p.

To calculate the dividend yield, it is necessary to gross up the 5.33p dividend by adding back the ruling basic rate of tax – currently 20%. The resultant 6.66p is then divided by the share price of 200p to give a prospective yield of 3.33%.

The rolling twelve months ahead table then reads like this:

Armed with these statistics, and only then, can a growth investor meaningfully compare ABC with another company, which has an end of February year end or has been brought, by making the same kind of calculations, onto a rolling twelve months ahead basis starting on 1 March. The good news for REFs subscribers is that all of these calculations are made for them and shown every month (daily on TOPIC) on the shaded panel of key statistics in the company entry. Three examples from the June 1996 issue are given below:

The black moons show how each company statistic compares with market and sector averages – the blacker they are the better. The other key growth statistics will be explained more fully in Chapter 9.

There is another important advantage of using rolling twelve months ahead statistics. Fast-growing companies usually have high historic PERs, which can be very offputting to investors thinking in terms of the last financial year. Take, for example, a company that had grown at about 50% per annum compound and is forecast to continue doing so for the next few years. Its historic PER for the year ended 31 March 1996 could be about 30, which seems high in relation to the market as a whole. However, with further growth of 50%, the prospective PER for the year ending 31 March 1997 falls to 20 and for the year ending 31 March 1998 it drops to only 13.3. The REFS entry on 1 January 1997 would show the rolling PER for the twelve months immediately ahead, which would be calculated using 3⁄12 of the PER for the year ending 31 March 1997 and 9⁄12 of the PER for the year ending 31 March 1998. The REFS prospective PER would therefore be a much more attractive 15 times EPS and obviously a bargain in relation to a 50% growth rate, highlighted by a PEG of only 0.3.

The REFS approach enables investors to see clearly the kind of multiple they are really being asked to pay for such astonishing future growth. Also, every other company they compare it with will be portrayed on a similar basis. In REFS, chalk is always being compared with chalk.

There is no doubt that the rolling twelve months ahead basis helps to detect anomalous bargains very quickly, especially among companies with exceptional EPS growth rates. It is strange that no one else seems to work on forward statistics in quite the same way as REFS. In the many presentations I have given to private investors and to institutions, there has never been anything but praise for this approach. Everyone agrees that the idea is eminently sensible, so it has always surprised me that working with rolling twelve months ahead statistics has not become common practice. I suppose the reluctance of the institutions to change their ways should be welcomed. While big money continues to appraise companies using old methods, anomalies remain and, as a result, investment opportunities abound.

Comparing companies on the basis of their estimated future results is, of course, highly dependent on the accuracy of brokers’ forecasts. The past few years’ results may be historic, but at least they are factual, whereas forecasts are at best informed guesses. Sam Goldwyn, the American film magnate, put it very well: ‘Forecasts are dangerous, particularly those about the future.’

The risk of being wrong about the consensus forecast can be lessened by taking a few elementary precautions. First, check the annual report and interim statement to find out exactly what the chairman has said about future prospects. Also, check press cuttings to see if anything has been added at the AGM or in interviews with press. The REFS company entry always shows the last few price-sensitive forward-looking comments by the chairman starting from the most recent annual report. The tone and trend of them and of the newsflow is often just as important as the exact words that are used.

In their excellent book Interpreting Company Reports and Accounts , Geoffrey Holmes and Alan Sugden taken an imaginary company, Polygon Holdings, in a range of industries and industrial climates and give their suggestions for estimating the year’s profits based upon the chairman’s comments on current trading:

The authors also draw attention to the necessity of keeping an eye open for any details of discontinued loss-makers (usually an excellent sign), of judging the chairman’s previous record of forecasting (a helpful guide to this year’s accuracy) and of being very wary of vague statements like: ‘Unforeseen difficulties have occurred.’

Another useful indicator is the level of dividend paid and/or forecast. If dividends have been steadily rising and are then simply maintained, that can be a meaningful sign that there could be trouble ahead. Conversely, if the rate of increase in dividends is accelerated, it is obviously a bullish and confident gesture.

Further indicators can be of a more general nature. For example, the retail sales trends that are announced each month can be a useful pointer to the credibility of a retailer’s forecast and so can the performance of competitors. You might notice in the press that the government of a foreign country has become disenchanted with the UK because the British Government failed to extradite someone or offended it in some other way. As a result, major contracts in the country in question might become more difficult to obtain by UK tenderers. Any UK company with a very substantial dependence on business in that country would be bound to suffer.

Like Eisenhower trying to judge the right place and time to invade German-occupied Europe, you must take note of all the little signs and pieces of information you can put together. It is essential to monitor every share in your portfolio by keeping an eye on the validity of the brokers’ consensus forecast that, more than anything else, underpins the share price.

Precise details of brokers’ consensus details are shown in monthly publications like The Estimate Directory . They are now also shown in REFS each month as an essential part of the overall financial picture of a company.

The following is an extract from a typical REFS company entry, that of Medeva in June 1996:

A number of points deserve special attention when studying details of the brokers’ consensus forecast:

Brokers are often slow to update their forecasts, so always keep an eye on the relative strength of the share price against the market. It is often an early indicator that a company’s situation is changing for the better or worse. This subject is dealt with much more fully in Chapter 7 on Relative Strength and Chapter 15 on Portfolio Management.

There is always a risk that the brokers’ consensus forecast will not be met. However, by studying the detailed brokers’ forecasts and their recommendations set out in REFS, reading the chairman’s comments and press announcements and noting articles about general conditions in the industry, you can reduce the risk to an acceptable level.

2. The PEG is calculated by dividing the prospective PER of a company by its estimated future growth rate in EPS. Both the prospective PER and the growth rate should be calculated on a rolling twelve months ahead basis.

3. The average PEG in the market in early 1996 was about 1.5. PEGs of over 1.5 are unattractive, not very attractive between 1.0 and 1.5 and of great interest well below 1.0.

4. REFS awards PEGs only to growth companies which have at least four years of consecutive growth whether it be historic or forecast or a combination of the two. They must also meet a number of other criteria set out in Chapter 3.

5. Low PEGs work best in the 12-20 range of PERs with substantial annual growth rates in EPS of between 15% and 30%.

6. Shares with low PEGs combine aggressiveness with a safety factor. The aggression comes from the likelihood of an upward status change in the PER, which often has a greater impact on the share price than the increase in EPS. The safety comes from the capacity of shares with low PEGs to handle minor disappointments in actual EPS growth rates against consensus forecasts. If, for example, the EPS growth of a company turned out to be only 20% instead of an anticipated 25%, the shares would still be relatively cheap if they were on a prospective multiple of, say, 15.

7. Growth stocks can only be compared with each other when they are first reduced to a like-for-like basis. Expressing their growth statistics (PERs, PEGs, growth rates and dividend yields) on a rolling twelve months ahead basis achieves this objective and has the added attraction of being up-to-date and dynamic.

8. To lessen the risk of brokers’ forecasts misleading you, double-check the annual report and interim statement to find out exactly what has been said by the chairman about future prospects. Also check press cuttings and the newsflow in the REFS company entry.

Other indicators include the level of dividends and the general trend of sales in the industry as a whole, especially with retailing companies and any other major developments that could have a particularly adverse or beneficial effect on the company in question.