An Empirical Study Of Asset Size, Financial Leverage, And Stock Market Anomalies
Empirical research has confirmed that stock market anomalies, such as the small firm effect [1, 3, 4, 6, 7, 8, 9] and the price to earnings ratio (P/E) effect [2, 3, 4, 7], have been in existence for long periods of time. These anomalies, i.e., apparent violations of the efficient market hypothesis, seemingly allow investors in low market value stocks and low P/E stocks, respectively, to earn positive abnormal risk-adjusted returns. The results of these investigations suggest that stock markets are inefficient, asset pricing models are misspecified, or both. Secondly, it has been suggested that (1) the P/E effect is merely a proxy for the small firm effect , (2) the small firm effect is a proxy for the P/E effect , and that (3) the two effects are separate from each other . From the investor's perspective, this contradiction presents a problem since the investor is uncertain about which method to use to obtain maximum risk-adjusted returns. The implementation of a strategy that provides excess risk-adjusted returns would be greatly simplified if one anomaly is clearly demonstrated to be a proxy, or substitute, for the other anomaly since the benefits of both strategies could be obtained from use of only one strategy.
Unlike previous investigations, this study examines the mutual independence of the P/E effect and small firm effect based upon financial accounting characteristics, i.e., the book values of total assets and financial leverage. By using the book value of total assets, one avoids using a potentially biased measure for comparison. Specifically, if stock markets are inefficient, then stock prices would be biased. Since stock price is used in determining both total market value (firm size = # of common shares publicly held times market price) and P/E ratio, abnormal returns obtained might be due to mispricing. The total value of assets reported on the firm's balance sheet does not suffer from this bias.
It is demonstrated that the book value of total assets is not a factor common to both the P/E effect and firm size effect. Market value of common equity is highly correlated with total asset value while P/E ratio is not. The evidence presented here suggests that the two anomalies are independent of each other rather than proxies for each other, as has been asserted by previous researchers. Furthermore, the book value of financial leverage as a contributor to explaining the P/E effect produces mixed results.
Using data from a combined COMPUSTAT Industrial and Industrial Research file, Basu  grouped New York Stock Exchange (NYSE) issues by P/E ratio and segmented them into quintiles (five groups, each representing one-fith of the total sample). He found the lowest P/E ratio firms provided abnormal, positive, risk-adjusted returns over the 1956 to 1969 time period. Basu concluded that perhaps the Capital Asset Pricing Model (CAPM) is misspecified.
In 1981 Banz  reported results similar to those of Basu, except Banz ranked firms by market value of common equity instead of P/E ratio. Using the Center for Research in Security Prices (CRSP) monthly returns file for the time period 1926 to 1975, Banz observed that "the common stock of small firms had, on average, higher risk-adjusted returns than the common stock of large firms" [pp. 3-4]. The determination of whether a company is small or large, however, is based on market value of common equity only, i.e., market price times number of common shares publicly held.
Reinganum  investigated both the small firm and P/E effects in 1981 using NYSE and American Stock Exchange issues from the CRSP daily returns file. He calculated the market value of common equity and the E/P ratio (the inverse of the P/E ratio) for each firm used. Even though Reinganum found evidence of positive abnormal risk-adjusted returns associated with both high P/E (low E/P) stocks and low market value stocks, he concluded: