Tuesday, May 28, 2013

The Capital Asset Pricing Model – “Understand markets on their own terms and not on our own”?


 
"Grau, theurer Freund, ist alle Theorie."
(Johann Wolfgang von Goethe, Faust)

The Abacus Journal has recently published a series of articles about the use of the Capital Asset Pricing Model (CAPM). Before summarizing the academic discussion, let's first understand what the CAPM is about:

With the CAPM you can calculate the price of a financial assets. If you buy a stock, you should get the rate that everyone gets who invests in stock markets (This is the risk-free rate.) plus a surplus which is a function of the risk involved with the specific company you invest in (This is the equity risk premium.).

This leads us right away to the CAPM formula:

k = rf + Beta * (km – rf)

Definitions:

k = the required rate of return on the financial asset = cost of equity
rf = the risk-free rate
Beta = the beta of the specific stock in question
km = the required rate of return for the market = the expected return of the market portfolio
Beta * (km – rf) = the equity risk premium


The CAPM is built on two main assumptions:

  • Investors act rationally.
  • Investors dispose of all important information about the financial assets in question.


The Abacus supplement discusses an article from Mike Dempsey who says:

In adhering to the CAPM we are choosing to encounter the market on our own terms of rationality, rather than the market’s.”

Dempsey's basic criticism is that the CAPM is built on rationality of markets although empirical evidence contradicts this assumption. In his view, markets can self-correct irrational reactions only in the long run. Dempsey is not convinced of post CAPM models, which simply add additional factors to the CAPM equation:

The identification of the correlation of a variable with asset returns is then presented as either an ‘anomaly’ or as the demonstration that the variable is ‘priced’ by the markets. This is what Black meant by saying that the exercise amounts to data mining.”

The reactions of the other authors to Dempsey's statements are mitigated:

  • D. J. Johnstone focuses on the debate whether an investment should be appreciated on the basis of mean-variance (MV) of its returns or on its subjective expected utility (SEU) for investors. The author criticizes that this debate is no more present in today's finance which is built on mean-variance.
  • Karen Benson and Robert Faff criticize Dempsey and call his work “greatly overstated” and “unduly ‘nihilistic’”. They argue that asset pricing is a long-term task and, therefore, the CAPM still useful. To counter contradictory empirical evidence, the authors argue that the CAPM cannot be tested at all. The reason is that one fundamental parameter of the CAPM is the market portfolio, which is, however, unobservable in practice. Any index which replaces the true market portfolio is, therefore, doomed to distort the results of the CAPM. The authors accept these shortcomings of the CAPM but argue that it is still preferable over no pricing model at all.
  • Henk Berkman says that the lack of empirical support for the CAPM does not make the model void. Rather, it “should be viewed as a necessary step in the normal scientific process resulting in a better understanding of asset pricing”.
  • Philip Brown and Terry Walter respond to the lack of empirical proof of the CAPM by stating that the CAPM is an ex ante concept and, thus, cannot be tested ex post. In addition, as the true market portfolio is not and may never be known, any empirical test of the model is, in the authors' view, necessarily flawed.
  • Charlier X. Cai, Iain Clacher, and Kevin Keasey agree with Mike Dempsey and call for understanding markets better: “We need to understand how markets actually behave rather than assuming they behave in a given way.” Their article includes pretty strong language such as “Through its assumptions of rationality and efficiency, academic finance has legitimated an emphasis on financial markets, unfettered profit and the ability of markets to self-regulate, and this has filtered through into government policy”.
  • Imad A. Moosa fully agrees with Mike Dempsey and says that “the model is theoretically bankrupt, empirically unsupported, and practically useless at best and misleading at worst”. “Excessive mathematization of economics and finance has resulted in work that has no relevance to reality.”
  • Graham Partington takes a pragmatic view. He emphasizes that the critics must consider what the CAPM is used for. Graham argues that discount rates are not critical to budgeting decisions. As a consequence, the job of the CAPM is of much less importance than usually thought which compensates for its insufficiencies. In addition, he points out that there is simply no alternative to the CAPM today.
  • Tom Smith and Kathleen Walsh say that “the CAPM is half-right and everything else is wrong”. As a matter of fact, since it is still very difficult to make abnormal returns from publicly available information, financial markets should still be considered efficient. Besides, the authors argue that the CAPM is actually not based on the market efficiency hypothesis. Instead, the model has its roots in the concepts of time value of money (A USD today is worth more than a USD tomorrow.), diversification (Diversification reduces risk.), and arbitrage (Same cash flow – same price).
  • As does Graham Partington, Avanidhar Subrahmanyam takes a pragmatic viewpoint. “He asks “What is the cost to practitioners to using the wrong model?” and “What is the alternative model?”


In short, the authors seem more or less to agree that the CAPM has serious shortcomings. But as we have nothing else, it will probably be still around for quite a while. As said Johann Wolfgang von Goethe in his Faust: “Es irrt der Mensch solange er strebt.”


You can download the Abacus supplement on the CAPM here.