Portfolio performance measures using holdings data are panel regressions. The returns of a fund’s stocks are regressed on its lagged portfolio weights. Stock fixed effects isolate average performance from time-series predictive ability. Control variables condition for fund performance on the characteristics of the stocks held. The long-term performance of average holdings drives some of the classical measures, while predictive ability drives others. A “buy-and-hold drift,” where portfolio weights increase over time in the higher alpha stocks, affects performance measures. Investor flows respond to average performance net of the buy-and-hold drift. (JEL G11, G14, G23, G29).

Mimicking portfolios have long been useful in asset pricing research. In most empirical applications, the portfolio weights are assumed to be fixed over time, while in theory they may be functions of the economic state. This paper derives and characterizes mimicking portfolios in the presence of predetermined state variables, or conditioning information. The results generalize and integrate multifactor minimum variance efficiency (Fama (1996)) with conditional and unconditional mean-variance efficiency (Hansen and Richard (1987), Ferson and Siegel (2001)). Empirical examples illustrate the potential importance of time-varying mimicking portfolio weights and highlight challenges in their application.

ABSTRACT The methods of Gibbons and Ferson (1985) are extended, relaxing the assumption that expected returns are linear functions of predetermined instruments. A model of conditional mean‐variance spanning generalizes Huberman and Kandel (1987). The empirical results indicate that more than a single risk premium is needed to model expected stock and bond returns, but the number of common factors in the expected returns is small. However, when size‐based common stock portfolios proxy for the risk factors, we reject the hypothesis that four of them describe the conditional expected returns of the other assets.

Previous studies identify predetermined variables that predict stock and bond returns through time. This paper shows that loadings on the same variables provide significant cross‐sectional explanatory power for stock portfolio returns. The loadings are significant given the three factors advocated by Fama and French (1993) and the four factors of Elton, Gruber, and Blake (1995). The explanatory power of the loadings on lagged variables is robust to various portfolio grouping procedures and other considerations. The results carry implications for risk analysis, performance measurement, cost‐of‐capital calculations, and other applications.

We develop methods for testing factor models when the weights in portfolios of factors and test assets can vary with lagged information. We derive and evaluate consistent standard errors and finite sample bias adjustments for unconditional maximum squared Sharpe ratios and their differences. Bias adjustment using a second-order approximation performs well. We derive optimal zero-beta rates for models with dynamically trading portfolios. Factor models’ Sharpe ratios are larger but standard test asset portfolios’ maximum Sharpe ratios are larger still when there is dynamic trading. As a result, most of the popular factor models are rejected.

ABSTRACT Even though stock returns are not highly autocorrelated, there is a spurious regression bias in predictive regressions for stock returns related to the classic studies of Yule (1926) and Granger and Newbold (1974) . Data mining for predictor variables interacts with spurious regression bias. The two effects reinforce each other, because more highly persistent series are more likely to be found significant in the search for predictor variables. Our simulations suggest that many of the regressions in the literature, based on individual predictor variables, may be spurious.