Recall that the CAPM is the foundation of asset pricing in modern finance. Perhaps the most important lesson we can learn from CAPM is that the single biggest factor in determining the premium that investors can receive from any investment is the risk premium. With very few esoteric exemptions, the risk premium is the only premium. If we follow the CAPM and view all risk as market risk as defined by beta, then alpha is a mere illusion that active managers use to justify charging investors exorbitant fees. From this perspective, alpha cannot exist. Thus an investment manager cannot deliver it.
Using this basic framework, we can conclude that factors are just tilts that cut out cross-sections of beta. Investors tend to wrongly conclude that beta is a unidimensional construct, and make mistakes based on the fallacy. I argue that beta is a multidimensional construct and that factors represent the major underlying constructs. While I find this concept intriguing and want to see the factor structure tested against a single factor model using Structural Equation Modeling (SEM), such musings go beyond the scope of this text (if you are in a Ph.D. program, note the free dissertation idea). My point here is that paying a manager substantial fees to deliver alpha is a complete waste of money if all we need are cross sections of beta. We can do that ourselves once we identify the factor structure (which has already been done by the likes of Markowitz, Fama, and French), and use the savings in fees to bolster our own returns.
I say that CAPM is the foundation of asset pricing because it made the critical error of treating beta as a unidimensional construct. Taken as such, all of the risks and returns are homogenous, and thus you cannot beat beta because beta bakes in every possible tilt. Considered as a unidimensional construct, beta also bakes in mediocrity. Mediocrity may seem like a pejorative term, but, in reality, the mean return is good, and it is easy to achieve at a very low cost. Most investors perform below the mean because they try to time markets, try to find alpha, and other investing mistakes. Further, if we start dissecting beta and identifying factors that increase returns, we are by all accounts increasing risk.
Because beta is summative point estimate, it necessarily involves the systemic risk that cannot be diversified away (at least from within the market under consideration). To dampen the effects of equity market beta risk (according to the CAPM logic), we need to diversify using different asset classes. If we partition beta into various factors and examine the risks and returns of those factors individually and in concert, we make a surprising and very useful observation: The risk and returns of individual factors are not perfectly correlated.
In fact, some of them have nearly independent variances. With this fact in mind, we can view market risk as a sort of smoothed average of individual factor risks. We can extend this logic to form the idea that the factor investors can diversify away risks that “buy the market” investors cannot. This, in turn, suggests that we can design portfolios that can beat the return of the S&P 500 and not take on anymore (portfolio) risk. It also suggests that conservative investors can achieve market returns with lower risks. The implications are profound.
Last Updated: 6/25/2018