John Price

(Reprinted with permission from Investor Journal, November 1997,
the magazine of  Investors Alliance.)

The P/E ratio

After finding the price of a particular stock, usually the next number everyone looks at is the p/e ratio, the current price of the stock divided by the total earnings of the company per share for the past year. Often we read comments that you should avoid stocks with a low p/e, below 6 say, or with a high p/e, above 25 say. But, as we will see, this only tells a small part of the story.

Some analysts tell us that the p/e ratio is the number of years required for the earnings to cover the price of the stock. For example, if the p/e is 12, since this means that the price of the stock is 12 times its earnings (per share), it will take 12 years for the earnings to equal the value of the stock. There is a big ‘if’ here, namely that this is the case if the earnings remain constant. Clearly, if the growth rate is 10 percent, the price of the stock will be covered in less than 12 years. In fact, in this case it will be covered in approximately 8 years.

There are various rules of thumb to make allowances for the growth rate when considering the p/e of a stock. In One Up On Wall Street, Peter Lynch suggests that a p/e ratio that is half the growth rate is a good sign. Even with our current bull market, Power Investor makes it easy to find stocks satisfying this criterion. For example, the screen

[P/E] > 6, [EPS Growth-1] > 2 * [P/E], [EPS Growth-2] > 10

gave 351 stocks. (See the footnote for an explanation of these commands.) Here’s an extreme case: Technitrol, Inc TNL. It has a current p/e of 26.1 along with an earnings growth rate of 118.7 percent. Something more moderate: Advanta Corporation ADVNA. It’s current p/e is 8.35 and its earnings growth rate is 21.6 percent. It is certainly worth taking a closer look at stocks with these numbers.

Other analysts express a similar idea when they talk about a stock with a p/e that is at a discount to its growth rate. For example, Advanta is selling with its p/e ratio at a discount of 13.25 percent to its growth rate (13.25 = 21.6 – 8.35).

In the other direction, according to Peter Lynch, stocks with a p/e greater than twice their growth rate are to be avoided. As to be expected, there are many more of these in the current market. One example is H.J. Heinz HNZ which has a p/e of 24.9 and an earnings growth rate of 10.1 percent. This stock is selling with its p/e ratio at a premium of 14.8 percent to its growth rate (14.8 = 24.9 – 10.1).

One weakness of these rules of thumb for deciding when a stock has a favorable price or not is that they do not make any sense when the growth rate is close to zero. For example, the buy rule would tell us that if a stock has a zero growth rate, we can only buy it if its p/e ratio is zero. Considering stocks with positive earnings, this means that we only buy zero growth stocks if their price is also zero!

There is another complication when thinking of the p/e ratio as describing the number of years it will take for the earnings to cover the price of the stock. It arises from the fact that to do this properly, you need to take into account the requirement that the stock has to be paid for in today’s dollars, whereas the earnings are being realized in future dollars which will be worth less. For example, taking the treasury rate as 7 percent, receiving one dollar in a year’s time is the same as receiving 93.5 cents now (0.935 = 1.00/1.07), receiving one dollar in two years is the same as receiving 87.3 cents now, and so on. In the next section, I will describe a new rule that builds on the methods just described but which also incorporates all growth rates¾ positive, negative, and zero¾ as well as the treasury rate.

Covering Period

In the previous section I described a stock selection rule which was to seek out stocks if their p/e ratio is less than half their growth rate. On the other side, the rule was to avoid stocks selling with a p/e ratio greater than twice their growth rate. The first weakness of these rules is that they do not make sense when the growth rate is close to zero. Secondly, they do not take into account the fact that the stock is to be paid for with today’s dollars, but the earnings are realized in a stream of payments of future dollars which are worth less.

Suppose that we know the growth rate of the earnings per share of a company. Now discount these future earnings back to present time using the treasury rate. It is possible to calculate the number of years necessary for the sum of these present values to equal the price of the stock. I call this the covering period. Clearly the smaller the covering period, the more attractive is the stock. Conversely, the larger the covering period, the more the stock should be avoided. Calculation of the covering period combines both the p/e of the stock and its growth rate.

Believe it or not, in many cases, the sum of these present values never cover the price of the stock. In other words, the covering period is infinite. One example is Crestar Financial Corporation CF. The growth in its earnings over last year and the year before was 1.6 percent and -0.5 percent. Summing the present values described above gives a value of approximately 18, yet it is trading at value which gives it a p/e ratio of around 22. Of course, people are hoping that it is going to return to the growth levels of 1994 and earlier which were 20 percent and up. But even then, with such a high p/e it could be a long time before the share price will be pushed up from its current levels.

Of course, in all this a major problem arises in estimating the future growth of earnings. One approach is simply to plug in one of the estimates that can be obtained from various web sites. Another is to base the estimate on past growth rates. Care has to be taken in this latter method when the earnings are very small since minor variations in the earnings can produce abnormally large growth rates which cannot be sustained. One simple approach is to consider growth in earnings over a number of years and to give extra weight to the recent years.

In the first table I analyze major computer companies by using the average of the 5-year growth rate and the 1-year growth rate. The result is a ranking of these companies by the lengths of their covering periods. 

The shortest covering period is 3.66 years in the case of Storage Technology and is mainly due to the extremely large growth rate in the last year. The largest covering period is 46.66 years in the case of old ‘Big Blue.’ Even though it has a low p/e, its growth rate is even lower. In the terminology of the previous article, its growth rate is at a discount of approximately 17 with respect to it p/e.

In the second table I give the same ranking for restaurants. In this group, the p/e ratios are generally lower than those in the first group, as are the growth rates. There are, however, two stocks with infinite covering periods. Apart from these and IBM in the first group, the distribution of the covering periods in each group are approximately the same.

In general, stocks which are worth examining more closely are those with growth rates at levels which can be sustained but which also have short covering periods. Running through the examples considered by Peter Lynch, a pattern begins to emerge that he favors stocks with covering periods of 8 years and less. Within that group he also seems to favor stocks with lower that average p/e ratios and growth rates.

I am starting a long term statistical study to look at the returns of stocks with short covering periods compared to those with long ones. When its completed, I’ll let you know the outcome.

Footnotes 1. The commands [P/E] > 6, [EPS Growth-1] > 2 * [P/E], [EPS Growth-2] > 10 are for use in PowerInvestor  by Investors Alliance, Inc. They act as a screen to select only the stocks with (i) a p/e ratio exceeding 6, and (ii) earnings growth over the past year exceeding both twice the p/e ratio and 10.

2. A module for calculating the covering period of stocks is included in the  Valuesoft Investment System.