Although buzzwords such as "investment factors", "factor funds", "smart/strategic beta" have recently gained popularity, Factor investing was pioneered back in the 1960s.
This era gave birth to many modern finance theories including the seminal Capital Asset Pricing Model (CAPM) which was developed by Jack Treynor, William Sharpe, John Lintner and Jan Mossin (Perold, 2004). CAPM's central thesis is that the return generated by a security is a function of its volatility compared to the market as a whole, describing this parameter as “Beta”. From an evolutionary point of view, this was the first time that a mathematical explanation of a security's return was attempted and CAPM was, for all practical purposes, the first single-factor investment model.
However, CAPM only captures the security’s exposure to systematic risk i.e. the component that is inherent in the market or the overall economy. Security returns also depend on various firm-specific/idiosyncratic surprises viz. CEO changes, product launches, and better-than-forecasted earnings announcements etc. that are neglected by CAPM. In addition, CAPM’s assumption of frictionless markets with no taxes, no transaction costs and no impediments to short selling mean that, CAPM assumes that normally securities are priced solely based on their beta. Furthermore, CAPM suggests a linear relationship between a security’s expected return and its beta, a relationship difficult to prove statistically.
Although a seminal development, CAPM has low explanatory power, largely due to its far-fetched assumptions that do not reflect the functioning of markets in the real-world. The relative outperformance by low-beta (low volatility) stocks compared to high-beta stocks is itself a testament of non-linearity, reinforcing CAPM’s failure in the real world. Such anomalies motivated the academic community to devise more sophisticated asset pricing models that take into account multiple factors, thus setting the stage for modern-day factor investing.
One of the most notable efforts in this domain is the work of Nobel Laureate Eugene Fama and Kenneth French in 1992. Fama and French created the Fama-French 3-Factor Model (FF3F) that explains security returns using three risk-factors; beta, size and value (Fama & French, 1992). Considered the cornerstone of factor investing, the FF3F plays a pivotal role as it successfully includes the impact of firm-specific characteristics proving mathematically that Size (relative outperformance by smaller firms) and Value (relative outperformance by undervalued firms) play a dominant role in explaining security returns across markets, geographies, sectors, and time. Due to its significantly higher explanatory power, the FF3F is regarded as the first truly multi-factor investment model and paved the way for future developments.
The success of the FF3F enticed academics to develop even more sophisticated asset pricing models which take into account several other factors. One such multi-factor model is the Carhart 4-Factor Model that adds a momentum factor to FF3F, proposing that a security’s returns are also influenced by the tendency of securities that have experienced a recent trend to continue to move in the direction of the trend in the near future (Carhart, 1997). Similarly, the Pastor-Stambaugh Model (Pastor & Stambaugh, 2003, #) adds a liquidity factor to the FF3F, stating that security returns are also impacted by a liquidity premium.
Thanks to the advancements in asset pricing theories and factor investing, we have various multi-factor models that seek to explain security risks and returns using a variety of investment factors. These models can largely be classified as macroeconomic, fundamental, or statistical models. Macroeconomic models explain actual returns using expected returns (calculated using asset pricing models such as CAPM) along with macroeconomic surprises (actual vs expected GDP growth, actual vs expected inflation etc.), fundamental models explain expected returns using a security’s fundamentals (earnings, growth, intrinsic value etc.), while statistical models explain a security’s returns using return statistics such as variance or covariance.
The Evolution of Factor Investing
Source: Fidelity Investments (Overview of Factor Investing | Investment Portfolio Building & Strategies | Fidelity)
NOTE: The proportions of the different factors (company-specific as well as systematic) displayed in the pie charts above are purely for illustrative purposes and these proportions do not accurately reflect the extent to which they impact security returns.