eview of the literature studying the PE ratio
Molodovsky (1953) is one of the first studies that investigate the theory of the PE ratios. By applying a concept called the estimated future earnings power, which is the ten-year moving average annual earnings, the study uses it as economic true income or basic earnings power. The current earnings number that is used to calculate the PE ratio deviates from the underlining true earnings. As a result, we may observe an opposite movements of earnings and the PE ratio: when current earnings number is below the true earnings, the PE ratio is higher, and vice versa. Essentially, the PE ratio is a multiplier for capitalizing the earnings power.
Beaver and Morse (1978) studies the time series property of the PE ratio and finds there is a long-term persistency in the portfolios’ PE ratios. Three years after portfolio formation, there is convergence of the PE ratios between highest and lowest portfolio and there is a pattern of reversion toward a central value for the PE ratios.
Beaver, Lambert and Morse (1980) expands the previous literature regarding the earnings process and introduces a compound earnings process: Xt = ?t + ?t, which specifies earnings process as ?t and ?t. Xt is the garbled earnings. The first process, ?t, the ungarbled earnings and an IMA(1,1) process, is the earnings that reflects events also affect prices. The second process is independent of price or price change. More importantly, the study introduces a valuation assumption regarding the relationship between the price changes and earnings changes. By applying this valuation assumption, the study implicitly assumes an earnings capitalization formula. Based on these two assumptions, the study shows that we may use the information contained in the stock price to predict the garbled earnings. The study also reinterprets the PE ratio and its positive correlation to the subsequent earnings growth rate: a high PE ratio exists because the garbled earnings are lower than the true economic earnings due to a negative transitory component. Further, the common denominator for earnings growth rate and price-earnings ratio leads to the positive correlation between the PE ratio and future earnings growth rate, although there may exist other reasons. Instead of the grouping technique used by the Beaver, Lambert and Morse (1980) to invert the price-earnings relationship and use the price as a predictor of earnings, Beaver, Lambert and Ryan (1987) directly reverses the regression by changing the price to independent variable and earnings as dependent variable and confirms the previous results.
One of the main contributions of the Beaver, Lambert and Morse (1980) is the valuation model they propose to capitalize the ungarbled earnings, that is, earnings as an economic variable and associated with firms’ value. Although the study forgoes the explanation regarding the capitalization factor, as later shown by Ohlson (1988), the formula is actually a special case that requires a restricted dividend payout rate, which imposes a limitation to the study. Ohlson (1988) specifically points out how the Garman- Ohlson (1980) framework can be used to solve the problem because Beaver, Lambert and Morse (1980) model is just a special case of the Garman-Ohlson model.
Another line of literature investigating the PE ratio considers the excessive market return that can be earned by adopting a trading strategy based on firms’ PE ratio, which is referred as the earnings-price ratio anomaly (Basu (1977), Brown, Kleidon, and Marsh (1983), and Jaffe, keim, and Westerfield (1989), among others). Two alternative explanations exist for the PE anomaly: market inefficiency, and misspecification or incorrect calculation of the risk adjusted market return (intrinsic value explanation). Both explanations have been proven in the literature (Ball (1992)) with no satisfactory reconciliation.
2.2 A review of the PE ratio and the earnings growth rate literature
Existing literature testing the relationship between the price earnings ratio and the earnings growth rate uses different measurements as for the PE ratio and the earnings growth rate. Earlier literature defines PE as current period closing stock price divided by same period reported earnings (contemporaneous PE) and the earnings growth rate as the current period or following periods realized earnings growth rate to study the correlation between PE and the earnings growth rate (Murphy and Stevenson (1967), Beaver and Morse (1978), Fairfield (1994), and Chan et al. (2003), etc). Later literature uses current period closing stock price divided by the next period expected earnings as PE (referred as the forward PE ratio) and calculate the expected short-term earnings growth rate from analysts’ forecast data (Thomas and Zhang (2006), Wu (2009) working paper). Empirically, after calculating the PE ratio, existing literature either tests the correlation between the PE ratio and the following period earnings growth rate, or builds portfolios according to the PE ratio (or the EP ratio) to test if higher PE ratio portfolio has higher following period earnings growth rate. Murphy and Stevenson (1967) is among the first studies to investigate the correlation between PE and the earnings growth rate. They find that there is no systematic relationship between these two variables, and
eview of the literature studying the PE ratio Molodovsky
eview of the literature studying the PE ratio