Risk, the Pricing of Capital Assets, and the Evaluation of Investment Portfolios: Comment

Abstract

In a classic article appearing in this Journal, Jensen presented empirical results from which he concluded that mutual fund portfolios showed inferior performance over the 10-year period 1955-64 after the deduction of all operating expenses, management fees, and brokerage commissions generated in trading activity.' Jensen argued further that, as a group, the mutual funds performance was neutral when all operating expenses and brokerage commissions were added back to the fund returns; therefore, the resources spent by the funds in attempting to forecast security prices did not yield higher portfolio returns than those which could have been earned by randomly generated portfolios. Finally, Jensen suggested that his evidence supported the strong form of the efficient market hypothesis-that is, the current prices of securities completely reflect the effects of all information concerning the securities, and efforts to acquire and analyze this information cannot produce consistently superior results. This comment will demonstrate that Jensen's empirical analysis and conclusions were based on questionable methods of estimating the mutual fund rates of return and levels of systematic risk. More specifically, it will be shown that Jensen's methodology (1) understated the mutual fund rates of return (and therefore understated the measures of excess return), and (2) introduced unnecessary measurement error into the analysis by assuming that the measures of systematic risk for the mutual funds were stationary over time.