The Performance of Mutual Funds in the Period 1945–1964
Source: Jensen, M. C. (1968). Journal of Finance 23(2), 389–416.
TL;DR
Derives Jensen's alpha — the intercept from regressing a fund's excess return on the market's excess return under the CAPM (Sharpe–Lintner–Treynor) — and applies it to 115 open-end mutual funds over 1945–1964. On average the funds could not outperform a buy-the-market-and-hold policy, and there is very little evidence any individual fund beat the market by more than random chance. The verdict holds even gross of management expenses: average net alpha ≈ −1.1% per year, average gross alpha ≈ −0.4% per year. The first rigorous, risk-adjusted assessment of active management.
The idea
Portfolio "performance" has two dimensions — forecasting ability (picking mispriced securities) and efficient diversification. Jensen isolates forecasting ability with a single number: alpha is the average return a manager earns beyond what the fund's market exposure (beta) would predict. Estimate Rₚ − R_f = α + β(R_m − R_f) + ε; a passive buy-and-hold position has α = 0, so a reliably positive α signals genuine selection skill and a negative α signals underperformance. Returns are annual, continuously compounded.
Evidence
Sample: 115 open-end funds; annual data 1945–64 (complete 10-year data 1955–64 for all funds; 56 funds had 10+ earlier years).
Net of all expenses: the average estimated intercept (alpha) was −0.011 (−1.1%/yr), ranging from a minimum of −0.078 to a maximum of +0.058. Only 3 of 115 funds had significantly positive alpha at the 5% level — fewer than the ~5–6 expected by pure chance; 14 funds were significantly negative.
Gross of management expenses (1955–64): average alpha was −0.004 (−0.4%/yr), with 67 funds showing negative alpha. So even ignoring management/research costs (but still bearing brokerage commissions), funds on average failed to recoup their trading costs.
The preponderance of negative alphas implies funds were not able to forecast security prices well enough to cover research, management, and commission expenses.
Why it matters
Defines alpha, the central measure of active performance, used everywhere from fund evaluation to factor research (where alpha means the return left unexplained by the asset-pricing model).
Early, influential evidence supporting market efficiency and indexing, and against the average active manager — a cornerstone of the active-vs-passive debate.
Caveats
Single-factor (CAPM) alpha conflates true skill with exposure to omitted factors (size, value, momentum), later addressed by multi-factor alphas (Carhart 1997).
The gross-return adjustment is approximate (expense data available only for 1955–64; brokerage commissions could not be added back), and differing per-fund degrees of freedom complicate the significance tabulation.
Survivorship and sample-selection issues in fund data can bias measured performance.
Key references
Jensen, M. (1968) — The Performance of Mutual Funds in the Period 1945-1964 — Journal of Finance
Sharpe, W. (1964); Lintner, J. (1965); Treynor, J. (undated) — Capital Asset Pricing Model foundations
Carhart, M. (1997) — On Persistence in Mutual Fund Performance — Journal of Finance
Fama, E. & French, K. (2010) — Luck versus Skill in the Cross-Section of Mutual Fund Returns — Journal of Finance
Provenance: verified/generated from the paper's full text.
