After you read this chapter, you should be able to answer the following questions:
• What are the deficiencies of the capital asset pricing model (CAPM) as an explanation of the relationship between
risk and expected asset returns?
• What is the arbitrage pricing theory (APT) and how is it similar and different from the CAPM?
• What are the strengths and weaknesses of the APT as a theory of how risk and expected return are related?
• How can the APT be used in the security valuation process?
• How do you test the APT by examining anomalies found with the CAPM and why do some authors contend that
the APT model is untestable?
• What are multifactor models and how are they related to the APT?
• What are the steps necessary in developing a usable multifactor model?
• What are the two primary approaches employed in defining common risk factors?
• What are the main macroeconomic variables used in practice as risk factors?
• What are the main security characteristic-oriented variables used in practice as risk factors?
• How can multifactor models be used to identify the investment "bets" that an active portfolio manager is making
relative to a benchmark?
• How are multifactor models used to estimate the expected risk premium of a security or portfolio?
Chapter 7 and Chapter 8 introduced Markowitz portfolio theory and the capital asset pricing model (CAPM), which represent the foundation for understanding the connection between risk and expected return. This chapter considers several extensions of this framework. Specifically, whereas the CAPM designated a single risk factor to account for the volatility of an individual security or portfolio, in this chapter we develop the intuition and application of multifactor explanations of risk and return. We begin with an explanation of the leading alternative to the CAPM—the arbitrage pricing theory (APT), which was developed by Stephen Ross. The chief difference between the CAPM and the APT is that the latter specifies several risk factors, thereby expanding the definition of systematic investment risk compared to that implied by the CAPM's single market portfolio.
After developing the conceptual basis for the APT in the next section and contrasting its major assumptions with those of the CAPM, we also examine the empirical evidence supporting the theory. One of the practical challenges that an investor faces when attempting to implement the APT is that the risk factors in the model are not defined in terms of their quantity (i.e., how many there are) or their identity (i.e., what they are). We conclude the chapter by discussing how investors use multifactor models, which can be viewed as attempts to convert the APT into a working tool for security analysis. A wide variety of factor models are currently in use. These models can be grouped broadly into those that use macroeconomic factor definitions and those that specify microeconomic factors. Examples are shown to illustrate each of the different approaches.
9.1 A RBITRAGE P RICING THEORY
The last chapter highlighted many of the ways in which the CAPM has contributed to the investment management field. In many respects, the CAPM has been one of the most useful—and frequently used—financial economic theories ever developed. However, some of the empirical studies cited point out deficiencies in the model as an explanation of the link between risk and return. For example, tests of the CAPM indicated that the beta coefficients for individual securities were not stable but that portfolio betas generally were stable. There was mixed support for a positive linear relationship between rates of return and systematic risk for portfolios of stock, with some recent evidence indicating the need to consider additional risk variables or a need for different risk proxies. In addition, other papers criticized the tests of the model and its usefulness in portfolio evaluation because of a dependence on a market portfolio that is not currently available.
