Beta and alpha

From Chapter 6 of The Analyst's Handbook, by Matt Lang and Bree Lang

Suppose that two student funds both returned 12.0% in a year when the market returned 10.0%. Each fund beat the market by 2.0 points, and that difference is called the active return. Take a moment and ask yourself whether the two funds showed the same amount of skill.

We cannot say yet, because every return is two things mixed together, the part that the market handed over and the part that the fund added. In order to separate them we first need to know how much market each fund was carrying, and the number that tells us is called beta:

β = Cov(r_p, r_b) / Var(r_b)

Here r_p is the fund's series of daily returns and r_b is the benchmark's, and the calculation wants at least thirty days of them. A beta of 1 means that the fund rises and falls with the market one for one, which is what an index fund does on purpose. Suppose that the first fund ran the year at a beta of 0.6, so it felt six-tenths of every market move, and that the second fund ran at a beta of 1.6, so every move reached it amplified by sixty percent in both directions.

Now we can ask what each fund should have earned for the market exposure it took, with no skill at all. Alpha is whatever the fund earned beyond that amount:

α = R_p − [Rf + β × (R_b − Rf)]

Here R_p is the return of the fund, R_b is the return of the benchmark and Rf is the risk-free rate. With a risk-free rate of 5.3%, the first fund was expected to earn 5.3% + 0.6 × (10.0% − 5.3%) = 8.12%, so its alpha is 12.0% − 8.12% = +3.88 points. The second fund was expected to earn 5.3% + 1.6 × (10.0% − 5.3%) = 12.82%, so its alpha is 12.0% − 12.82% = −0.82 points.

Take a moment with that last number, because the second fund beat the market and still earned a negative alpha. With that much market exposure in a year that good, a fund with no skill was supposed to win by more than this one did. The first fund, meanwhile, beat its smaller handicap by almost four points. This means that two funds with the same return and the same active return can deserve very different grades, and that active return taken by itself cannot tell us which is which.

Chapter 6 of the textbook computes a beta by hand from ten days of returns and then builds the alpha from it.

Read Chapter 6: Beta, alpha, and active return

Related topics: Tracking error and the information ratio · Performance attribution · All topics

The Analyst's Handbook is Sifter's free textbook for student investors, written by Matt Lang and Bree Lang. Sifter is free for students. This page as markdown.