Advanced Portfolio Management

Chapter 4An Introduction to Multi-Factor Models

• What will you learn here: What are multi-factor risk models; how many different types of models are there; and how are they estimated? Most importantly, how do they allow you to separate two worlds: the world of factors from that of company-specific returns?
• Why does this matter: Because it is the skeleton that supports the whole thing. Even if you do not see it, it is there and it is essential.
• When will you need this: Always and never. Always, because you will consume its byproduct daily; (almost) never, because you won't have to think about “factor portfolios”, or cross-sectional regressions unless you really miss them.

4.1 From One Factor to Many

We have seen that alphas do not come from straightforward regressions. What about betas? Is the market beta the beta to end all betas? It turns out that this is just the beginning of the story. The single-beta factor model was proposed independently by Lintner and Mossin, and Sharpe, who gave it the now-standard name Capital Asset Pricing Model (CAPM), in the mid-1960s. It was immediately put to the test. Initial empirical studies confirmed the model [Black et al., 1972; Fama and MacBeth, 1973]. In the mid-1970s, three separate contributions by young researchers set the stage for a second revolution. First, Stephen Ross, then an assistant professor at Yale, extended the CAPM. His starting point is to assume that there is a small number of factors, compared to the number of ...