No description of P/E analysis is complete without mentioning some of its pitfalls. First, consider that the denominator in the P/E ratio is accounting earnings, which are influenced by somewhat arbitrary accounting rules such as the use of historical cost in depreciation and inventory valuation. In times of high inflation, historic cost depreciation and inventory costs will tend to underrepresent true economic values, because the replacement cost of both goods and capital equipment will rise with the general level of prices. As Figure 18.3 demonstrates, P/E ratios have tended to be lower when inflation has been higher. This reflects the market's assessment that earnings in these periods are of "lower quality," artificially distorted by inflation, and warranting lower P/E ratios.
Another confounding factor in the use of P/E ratios is related to the business cycle. We were careful in deriving the DDM to define earnings as being net of economic depreciation, that is, the maximum flow of income that the firm could pay out without depleting its productive capacity. But reported earnings are computed in accordance with generally accepted accounting principles and need not correspond to economic earnings. Beyond this, however, notions of a normal or justified P/E ratio, as in equations 18.7 or 18.8, assume implicitly that earnings rise at a constant rate, or, put another way, on a smooth trend line. In contrast, reported earnings can fluctuate dramatically around a trend line over the course of the business cycle.
Another way to make this point is to note that the "normal" P/E ratio predicted by equation 18.8 is the ratio of today's price to the trend value of future earnings, E1. The P/E ratio reported in the financial pages of the newspaper, by contrast, is the ratio of price to the most recent past accounting earnings. Current accounting earnings can differ considerably from future economic earnings. Because ownership of stock conveys the right to future as well as current earnings, the ratio of price to most recent earnings can vary substantially over the
582 PART V Security Analysis business cycle, as accounting earnings and the trend value of economic earnings diverge by greater and lesser amounts.
As an example, Figure 18.4 graphs the earnings per share of Sun Microsystems and Consolidated Edison since 1986. Note that Sun's EPS fluctuate considerably. This reflects the company's relatively high degree of sensitivity to the business cycle. Value Line estimates its beta at 1.25. Con Ed, by contrast, shows much less variation in earnings per share around a smoother and flatter trend line. Its beta was only .75.
Because the market values the entire stream of future dividends generated by the company, when earnings are temporarily depressed, the P/E ratio should tend to be high—that is, the denominator of the ratio responds more sensitively to the business cycle than the numerator. This pattern is borne out well.
Figure 18.5 graphs the Sun and Con Ed P/E ratios. Sun, with the more volatile earnings profile, also has a more volatile P/E profile. For example, in 1989 and 1993, when earnings were below the trend line (Figure 18.4), the P/E ratio correspondingly jumped (Figure 18.5). The market clearly recognized that earnings were depressed only temporarily.
This example shows why analysts must be careful in using P/E ratios. There is no way to say P/E ratio is overly high or low without referring to the company's long-run growth prospects, as well as to current earnings per share relative to the long-run trend line.
Nevertheless, Figures 18.4 and 18.5 demonstrate a clear relationship between P/E ratios and growth. Despite considerable short-run fluctuations, Sun's EPS clearly trended upward over the period. Its compound rate of growth in the decade ending 1999 was 30%. Consolidated Edison's earnings grew less rapidly, with an average growth rate of 2.3%. Sun's growth prospects are reflected in its consistently higher P/E multiple.
This analysis suggests that P/E ratios should vary across industries, and in fact they do. Figure 18.6 shows P/E ratios in mid-2000 for a sample of industries. P/E ratios for each industry are computed in two ways: by taking the ratio of price to previous year (i.e., 2000) earnings, and projected next-year earnings. Notice that although the ratios based on 2000 earnings appear quite high, the ratios are more moderate when prices are compared to forecasted 2001 earnings. This should not surprise you, because stock market prices are based on firms' future earnings prospects.
Was this article helpful?