Market efficiency, CAPM, and value-relevance of earnings and EVA: A reply to the comment by professor Paulo.

Dr. Stanley Paulo makes two points or critiques in his Comment on our paper, "Operating Income, Residual Income and EVA: Which Metric is More Value Relevant?" published in the spring 2001 issue of the Journal of Managerial Issues (Chen and Dodd, 2001). First, EVA is useless because it is conceptually faulty either in an efficient or a non-efficient market. Second, there is growing evidence that accounting information such as earnings and dividends is becoming progressively less useful for stock valuation. Throughout the comment, he cites several published studies, primarily from the behavioral finance literature, to argue that stock price is not driven by intrinsic value, but by factors such as volatility, momentum, and financial herding. While an explicit statement is not made in the Comment, the tenor of the Comment clearly suggests that the author questions the value of capital markets research in accounting.

The Comment raises some very important questions given that market-based accounting research has been a primary paradigm of financial accounting research over the past three decades. However, we believe that some of his concerns need to be clarified and others discussed in a more balanced fashion. Consequently, in this Reply, we will attempt to clarify the issues that are specific to our article. We will also discuss evidence contradictory to the Efficient Market Hypothesis (EMH) and the Capital Asset Pricing Model (CAPM), plus the resulting implications for market-based accounting research.

EVA, Market Efficiency, and CAPM

Dr. Paulo states that EVA is a financial fiction in an EMH world because there could be no consistently abnormal profits due to arbitrage and competitive action. We believe that this comment confuses the EMH and the concept of economic rents or profits. According to the EMH, a market is efficient with respect to a particular set of information if it is impossible to make profits by trading on this set of information. The mechanism to make markets efficient is arbitrage and competitive action by sophisticated investors. However, the EMH does not prevent firms from earning economic rents, defined as profits that cover the opportunity cost of capital, which is a concept identical to EVA.

When an industry settles into a long-term competitive equilibrium, all assets are expected to earn their opportunity cost of capital because any economic profits have already been driven away by competition in the industry in terms of expansion by existing firms or entrance by new firms. However, the Long-term equilibrium is not a static point, but rather it is a dynamic process in which the equilibrium is constantly changing. In this process, firms enjoy economic profits due to either industry characteristics, such as industry concentration and industry-level barrier to entry, or firm characteristics, such as monopoly power and competitive advantages. It is because of economic profits that we advocate the use of the discounted cash flow analysis in capital budgeting to look for projects with positive net present values (Brealey and Myers, 2000). Thus, the existence of economic profits does not contradict the EMH as long as stock price fully reflects fundamental information about a firm's earnings generating ability such that at a particular time an investor cannot earn an abnormal return from a stock investment (Kothari, 2001).

Dr. Paulo also criticizes EVA as a deficient construct in a non-EMH world because the CAPM is not a valid model for computing expected returns, and stock prices are driven by non-fundamental factors other than earnings or dividends. We would like to make four observations.

First, we do not believe that the CAPM is derived from the EMH and is dependent upon the existence and functioning of the EMH; our belief differs from a statement made in the Comment. The CAPM only assumes efficient portfolios; it is not the same concept as the EMH. Although market-based event studies in accounting and finance generally assume the descriptive validity of both the EMH and the CAPM, the CAPM does not have to be maintained in an association study, such as our EVA article, where raw stock return is used as the dependent variable. The CAPM plays a role in our article because of its use in the computation of EVA. Since the purpose of our study was to evaluate the popular claim by the consulting firm Stern Stewart regarding the superiority of EVA in moving stock prices, we had to use EVA figures produced by Stern Stewart. Conceptually, our study does not assume the validity of the CAPM. In contrast, one possible explanation for the weaker association of EVA with stock return than accounting earnings documented in our article may be due to the noise introduced as a result of estimating the cost of capital based on the CAPM.

Second, although it is important to point out evidence against the CAPM, it is even more important to provide a valid alternative for computing the opportunity cost of capital. Economic profits, of which EVA is one version, are a sound economic concept. The issue here is whether the CAPM provides a valid basis to calculate the cost of capital. According to the CAPM, beta is the only reason for expected returns to differ. We understand Dr. Paulo's concern about evidence contradictory, such as three-factor models by Fama and French (1995), to the prediction of the CAPM. However, we would like to point out that defenders of the CAPM are equally forceful as critics of the model. For example, it has been stated that the anomalous evidence is a result of "data mining" or "data snooping" (Brealey and Myers, 2000; Kothari, 2001). In addition, defenders argue that the CAPM models the relationship between risk and expected return, while we can only observe actual return (Brealey and Myers, 2000). Even with the existing evidence often cited by critics, one cannot rule out the CAPM as the asset pricing model that still enjoys the broadest support in both the academic and the practicing community. The bottom line is that there is no widely accepted alternative model for computing expected returns. This explains why the CAPM continues to dominate capital markets research in accounting and finance.

Third, even with the evidence of non-fundamental factors affecting stock prices, we do not think that one should discard fundamental information, such as accounting earnings and EVA, in the equity valuation process. Three decades of empirical research, which we will discuss in the next section, and a casual observation of what investors do attest to the role of fundamental information. Many valuation models developed and used by the investment community have earnings as a primary input. P/E ratio is still the most frequently cited financial indicator in the stock market. If earnings are as trivial as Dr. Paulo believes, how can one explain investors' dramatic reaction when a firm's earnings surpass or fall short of an expected target? Financial headlines are flooded with such stories of stock price responses to earnings surprises during an earnings announcement season. Certainly, recognizing the role of earnings in stock valuation does not mean that earnings are the sole determinant of stock prices. As cited by Dr. Paulo, evidence shows that non-fundamental factors also affect stock prices. However, such evidence does not really invalidate the findings of our study because we employ the same model to compare the usefulness of earnings and EVA. There is no basis to believe that omitted variables would have any biased impact on one but not the other metric.

Fourth and finally, regarding Dr. Paulo's discussion of the long-term detached relationship between prices and earnings, we would like to point out that there are alternative explanations. For example, according to Givoly and Hayn (2000), financial reporting by U.S. companies has become increasingly conservative over time, which leads to depressed values in both earnings and book value of equity and thus high earnings and market-to-book multiples. Under this scenario, the time-series comparisons of these valuation measures as shown in Figures 1 and 2 of the Comment article may be misleading. The higher P/E ratios in more recent years do not necessarily suggest that stock prices are more disengaged from earnings over time, rather increasingly conservative accounting practices may be one of the culprits. To discriminate between competing explanations, we need more, not less, capital market-based accounting research.

Accounting Information and Capital Markets Research in Accounting

Dr. Paulo's comment on the relevance of accounting information is relatively short, but related to the recurring theme of the Comment. He first cites Lev and Zarowin (1999) as evidence of accounting information becoming progressively less relevant, and then repeats his main concern that stock prices are affected more by non-fundamental variables. We respond by addressing two issues. First, how do we evaluate the role of accounting information in equity valuation from existing evidence? Second, how do we assess the implications of market inefficiency for capital markets research in accounting?