Jensen's Alpha: Risk-Adjusted Performance Beyond CAPM

For educational purposes only; not investment advice.

Direct answer

Jensen’s alpha measures how much a portfolio’s return exceeds or falls short of the return predicted by CAPM for its market beta.

alpha = portfolio return - [risk-free rate + beta × (market return - risk-free rate)]

It asks whether performance remains unusually high or low after adjusting for exposure to broad market risk. It is most often used to evaluate active funds, managers, or strategies.

How it works

CAPM says expected return should rise with market beta. A fund with beta 1.5 is expected to move more than the market when the market risk premium is positive. Therefore, simply beating a market index does not prove skill if the fund took more market risk.

Jensen’s alpha can be calculated for one period or estimated through a regression:

R_p - R_f = alpha + beta × (R_m - R_f) + error

The regression intercept is the estimated alpha. If monthly data are used, the alpha is monthly unless it is explicitly annualized. A serious estimate also reports standard errors, t-statistics, confidence intervals, sample period, fees, and benchmark choice.

Example

Suppose a fund earns 14% in a year. The risk-free rate is 4%, the market return is 10%, and the fund beta is 1.2.

CAPM expected return is:

4% + 1.2 × (10% - 4%) = 11.2%

Jensen’s alpha is:

14% - 11.2% = 2.8%

In that period, the fund earned 2.8 percentage points above the CAPM return implied by its beta.

Now suppose the market return is -10%, the risk-free rate is 3%, beta is 1.2, and the fund return is -8%.

CAPM expected return is:

3% + 1.2 × (-10% - 3%) = -12.6%

Alpha is:

-8% - (-12.6%) = 4.6%

The fund lost money, but it still had positive alpha relative to its high-beta expected loss.

Risks

Benchmark risk: the wrong market proxy can create false alpha.
Beta risk: beta estimates change with time period, frequency, and market regime.
Omitted-factor risk: value, size, quality, momentum, or sector exposure can explain what looks like alpha.
Sample risk: short histories can make random luck look persistent.
Fee risk: gross and net returns can produce different alpha.
Purpose risk: hedging or tax-aware strategies may intentionally sacrifice CAPM alpha for another objective.

Common misconceptions

Jensen’s alpha is not raw outperformance.

Positive alpha does not automatically prove manager skill. It may come from an omitted risk factor, benchmark mismatch, leverage, timing, or luck.

Negative alpha does not automatically mean a strategy is useless. The strategy may serve a hedging, diversification, liquidity, or tax purpose.

One alpha estimate without statistical uncertainty is weak evidence.

Sources

Jensen, Sharpe, and Fama-French academic papers: alpha, CAPM, and factor-adjusted return context.
FINRA: investment performance reporting and interpretation context.