Changing the Definition of "Manager Alpha"


The Fama and French Three Factor (Five Factors with bonds) Model changes the definition of alpha, as seen in the figure below. According to the one-factor CAPM, alpha is the amount by which an active money manager outperforms a broad market index. The Fama and French Three Factor Model defines alpha for equities more precisely as the return an active manager achieves above the sum of the portfolio’s expected return due to all three equity risk factors. Alpha measures a manager’s skill in earning a return that couldn't have been achieved by indexing the same exact risk exposure as the portfolio run by the manager. In short, did the money manager earn anything above the indexed return?

A portfolio can be plotted anywhere on a graph that represents its exposure to the risk factors of size and value, and it’s easy to calculate its expected return. For example, a small-cap manager may overweight value stocks relative to a benchmark, such as the Russell 2000 Small-Cap Index. As a result, the manager outperforms it. But if the extra return was simply compensation for taking additional non-diversifiable risk, why should the manager get credit? The job of an active manager is to consistently outsmart the millions of other traders who get the same news at the same second, and through this process provide additional returns that can’t be achieved through indexing. This is exactly what the alpha is in the Three-Factor Model. Investors should insist that a manager outperforms a three factor risk adjusted benchmark before crediting him with an alpha return. After all, active manager fees are supposed to pay for predicting the future of stock prices, not for taking additional risk compared to low cost index funds.

So, what “positive alpha” managers have been doing with the one-factor CAPM measurement model is just systematically subjecting their clients to two additional risk factors - size (small company) and high BtM value (distress). Thus, what’s showing up as alpha (skill) is nothing more than a measurement error. If the performances of active managers are compared between CAPM and the Fama and French model, there are radical changes in the outcomes. Any evidence of manager skill just vanishes under the Fama and French model. The formula for this type of analysis is summarized in the figure above.

Even though active managers focus on alpha, the amount of return due to alpha from stock picking or market timing is random, and on average is expected to be negative. It turns out that alpha is nothing more than a myth perpetuated by the improper measurement of a manager’s performance.