Investment performance above a stated benchmark or model-implied return, whose meaning depends on the calculation used.
Alpha is return above a stated comparison return, but the comparison must be identified. In some reports, alpha means simple portfolio return minus benchmark return. In risk-adjusted analysis, it usually means return above the amount predicted by a model such as the Capital Asset Pricing Model (CAPM).
A simple benchmark-relative calculation is:
Where (R_p) is portfolio return and (R_b) is benchmark index return for the same period.
This result shows how far the portfolio finished above or below the benchmark. It is useful, but it is not automatically risk-adjusted. A portfolio can outperform simply because it carried more market, sector, duration, credit, currency, or leverage exposure.
The single-period CAPM form, commonly called Jensen’s Alpha, is:
Here, (R_f) is the matching risk-free rate, (R_m) is market-benchmark return, and (\beta_p) is portfolio beta.
| Label | Calculation | Risk adjustment | Best interpretation |
|---|---|---|---|
| Active return | Portfolio return minus benchmark return | None by itself | Value added relative to a stated benchmark |
| Jensen’s alpha | Portfolio return minus CAPM-implied return | Adjusts for estimated market beta | Return unexplained by the CAPM specification |
| Multifactor alpha | Regression intercept after specified factor exposures | Adjusts only for included factors | Return unexplained by that factor model |
Assume a portfolio returned 9.4%, its broad market benchmark returned 8.0%, the risk-free rate was 3.0%, and portfolio beta was 1.10.
Simple active return is:
The CAPM-implied return is:
Jensen’s alpha is therefore:
The 1.4% active return and 0.9% Jensen’s alpha answer different questions. The first is direct outperformance; the second removes the return associated with the portfolio’s estimated market beta under CAPM.
A positive result means the portfolio exceeded the stated comparison return. A negative result means it fell short. Neither sign proves or disproves manager skill from one period.
Before comparing alpha figures, verify:
Alpha or active return measures the amount of estimated value added. The Information Ratio compares average active return with tracking error. A strategy can have high alpha but a modest information ratio if its benchmark-relative outcomes are inconsistent.
Changing the benchmark, beta estimate, or factor model can change alpha materially. An omitted exposure may be mislabeled as skill.
Gross alpha belongs to the strategy before specified costs. Investors experience returns after fund expenses and other applicable costs, so net alpha may be lower or negative.
Short samples, multiple strategy tests, survivorship bias, and favorable start dates can produce positive estimates by chance. Statistical significance does not make future alpha certain.
Alpha omits absolute volatility, drawdown, liquidity, capacity, leverage, taxes, and how the portfolio combines with other holdings. It is one performance estimate, not a suitability conclusion.
This page is for financial education and does not recommend a manager, fund, benchmark, or strategy.