Is Smart Beta A Solid Concept?
Smart beta and its kissing cousin, factor-based investing, aim to take advantage of market anomalies—cases in which the market appears to incorrectly price securities in a consistent way. These anomalies violate the deeply held investment wisdom that states that expected returns should vary in line with expected risks. Are there really free lunches, or is there actually considerable risk attached to these anomalies? Put another way, is there an a priori reason to expect smart beta and factor investing to be successful, or are these new strategies poster children for data mining?
Most writers who discuss market anomalies such as the small-cap effect, the value effect, momentum and low volatility start by discussing CAPM, the capital assets pricing model. CAPM produces a single factor—beta—which describes the risk of a security, held in a well-diversified portfolio, relative to the market as a whole. CAPM’s predictive powers can be improved by the addition of factors. CAPM’s detractors posit that the existence of these factors prove CAPM’s irrelevance. Others take a broader approach, and include market beta alongside factors such as security size or valuation ratios in a multifactor model. To the extent that commonly included factors don’t seem to correspond to additional risk, many practitioners accept them as anomalies.
The most popularly used factors adjust for a security’s relative market capitalization; the relationship of its price to fundamental valuation metrics such as earnings or book value; and the existence of patterns in its recent trading values. Much ink has been spilled about identifying the factors, but most academics and practitioners accept two or three: size; value; and, sometimes, momentum. Recently, many studies have explored a fourth, which is identified either as low volatility or low beta. It is worth looking at each in turn, to understand the history of their discovery, explore potential risks they describe and evaluate the post-discovery persistence of the anomaly. Persistence is key to determining whether the effect accounts for investment risks, which do not decrease upon discovery.
To what extent do these factors correspond to risks, which explain their return? If these factors have no relationship to any known risk, then they either describe an irrational phenomenon, or describe a risk yet to be discovered. Investor irrationality is a core topic for behavioral finance, which has documented many ways that we humans systematically defeat ourselves. However, for any irrationality to persist in the marketplace, it must be immune to arbitrage. Therefore, even irrationalities often have roots in the peculiarities of market structures. A bit of digging will begin to tell the story of these factors, and the forces behind them.
Size
Initially documented by Banz in 1981,4 and incorporated into a CAPM framework by Eugene Fama and Ken French in 19925 and 1993,6 the “size effect” states that market-capitalization levels are predictive of future returns, with small-cap securities outperforming large-cap securities over time, in most market environments.
Explanations for the size effect abound, and include liquidity risk, information asymmetry/lack of analyst coverage, and trading costs. Some researchers have found evidence that the size effect is subsumed by the January effect, while others have found that it applies almost exclusively to the smallest tranche of securities by market capitalization [Horowitz (2000);7 Michou (2010);8 Fama (2008)9].