Jensen's Alpha
Jensen's Alpha is a risk-adjusted performance measure developed by Michael Jensen (1968) that quantifies a portfolio or investment manager's excess return above the theoretical expected return predicted by the Capital Asset Pricing Model (CAPM) given the portfolio's systematic risk (beta), calculated as the actual portfolio return minus the CAPM-predicted return: α = R_p - [R_f + β(R_m - R_f)]. A positive alpha indicates the manager generated returns above what beta alone would predict, suggesting genuine stock selection or market timing skill.
Key takeaways
- Jensen's Alpha = R_portfolio - [R_f + β × (R_market - R_f)]; positive alpha indicates outperformance after adjusting for systematic market risk.
- Unlike Sharpe and Treynor ratios, which are useful for comparing managers with different risk levels, alpha measures absolute excess return in percentage points, making it directly interpretable as the economic value added by the manager.
- Jensen's original 1968 paper found that the average U.S. mutual fund earned a negative alpha of -1.1% annually net of costs, providing early empirical support for the efficient market hypothesis.
- Alpha is sensitive to the benchmark and factor model used: a multi-factor alpha (controlling for size, value, momentum, and quality factors) provides a more demanding and rigorous measure of manager skill than single-factor CAPM alpha.
- Statistical significance of alpha requires long track records: even a true alpha of 2% annually requires approximately 8-10 years of monthly data to distinguish statistically from luck at the 95% confidence level.
Explanation
Jensen's Alpha emerged from the theoretical framework of CAPM, which predicts that the expected excess return of any asset or portfolio equals its beta multiplied by the market excess return: E[R_p] - R_f = β × (E[R_m] - R_f). Under the strict CAPM assumption that markets are fully efficient and all systematic risk is captured by beta, no portfolio should earn a positive alpha in expectation. Jensen's 1968 paper was significant not just for providing the mathematical formulation of alpha but for its empirical finding: applying the measure to 115 U.S. mutual funds over 1945-1964, Jensen found that the average fund earned a CAPM alpha of -1.1% annually—net of costs, active management appeared to destroy value rather than create it.
The intuition behind Jensen's Alpha is straightforward: any investor can passively earn the CAPM-predicted return for a given beta by holding the appropriate combination of the market portfolio and the risk-free asset. If a fund earns more than this CAPM-predicted return, the excess (alpha) must reflect either genuine skill (superior stock selection or market timing) or exposure to risk factors not captured by market beta. In Jensen's original single-factor framework, alpha represents all unexplained excess return—a mixture of genuine skill and uncompensated factor exposures. Multi-factor extensions of alpha address this limitation.
The evolution from Jensen's single-factor CAPM alpha to multi-factor alpha represents a significant refinement in performance measurement. The Fama-French three-factor model (1992) extended CAPM to include size (SMB: small minus big) and value (HML: high minus low book-to-market) factors. A four-factor model adds the Carhart momentum factor (MOM). Contemporary factor models include profitability (RMW), investment (CMA), and quality factors as well. Computing alpha against these multi-factor models requires regressing portfolio excess returns against all factor returns and interpreting the regression intercept as the alpha. A fund generating 2% CAPM alpha but zero multi-factor alpha is simply loading on known factor premia (small-cap, value, momentum) that are available at low cost, not generating genuine skill-based alpha.
For hedge fund investors (LPs), Jensen's Alpha is used in combination with other performance metrics to evaluate manager skill. The practical challenges are substantial: alpha estimates from historical returns are noisy due to the short time series available for most funds (the typical hedge fund has a 3-5 year track record), the true factor model capturing all risk exposures is unknown, and fund returns may reflect survivorship bias (failed funds exit databases, leaving a database of stronger performers). Statistical significance tests—requiring alpha t-statistics above 1.96 for 95% confidence—impose demanding data requirements. A manager with a true alpha of 3% per year would need approximately 7 years of monthly data to distinguish this from zero at the 95% level, assuming 15% annual portfolio volatility.
The relationship between Jensen's Alpha and portfolio construction is also significant. The Treynor-Black model (1973) shows that the optimal portfolio allocation to an active manager depends on the ratio of their alpha to their residual risk (non-systematic risk): a manager with high alpha per unit of residual risk (high appraisal ratio) deserves a large allocation, while a manager with lower alpha-to-risk earns a smaller weight. This provides a rigorous framework for combining multiple active managers and passive index exposure to maximize the expected Sharpe ratio of the aggregate portfolio, with each manager's allocation determined by their contribution to portfolio-level alpha net of the residual risk they introduce.
Formula
Jensen's Alpha (α) = R_p - [R_f + β_p × (R_m - R_f)]
Example
A long/short equity hedge fund earns a net return of 14.5% over a year in which the S&P 500 returns 20% and the risk-free rate is 5.0%. The fund's beta, estimated from a regression of monthly returns, is 0.60. CAPM-predicted return = 5.0% + 0.60 × (20% - 5%) = 5.0% + 9.0% = 14.0%. Jensen's Alpha = 14.5% - 14.0% = 0.5%. The fund barely outperformed its CAPM-adjusted benchmark, earning only 50 basis points of positive alpha despite delivering an absolute return of 14.5%. A competing fund with a 10.0% return, a beta of 0.30, and the same risk-free rate earns CAPM alpha = 10% - [5% + 0.30 × 15%] = 10% - 9.5% = 0.5%—the same alpha as the first fund, despite the dramatically different absolute return, revealing that the two funds delivered equivalent risk-adjusted performance once CAPM beta is accounted for.
Related terms
Alpha Basis Beta Cap Capital Asset Pricing Model Efficient Market Hypothesis Equity Equity Risk Premium Factor Model Fama French Three Factor Model Hedge Fund Idiosyncratic Risk Premium