[We evaluate the performance of models for the covariance structure of stock returns, focusing on their use for optimal portfolio selection. We compare the models' forecasts of future covariances and the optimized portfolios' out-of-sample performance. A few factors capture the general covariance structure. Portfolio optimization helps for risk control, and a three-factor model is adequate for selecting the minimum-variance portfolio. Under a tracking error volatility criterion, which is widely used in practice, larger differences emerge across the models. In general more factors are necessary when the objective is to minimize tracking error volatility.]

We evaluate the performance of models for the covariance structure of stock returns, focusing on their use for optimal portfolio selection. We compare the models' forecasts of future covariances and the optimized portfolios' out-of-sample performance. A few factors capture the general covariance structure. Portfolio optimization helps for risk control, and a three-factor model is adequate for selecting the minimum-variance portfolio. Under a tracking error volatility criterion, which is widely used in practice, larger differences emerge across the models. In general more factors are necessary when the objective is to minimize tracking error volatility.

[Academic and practitioner research evaluates portfolio performance using size and value/growth attributes or factors. We assess the merits of popular evaluation procedures based on matched-characteristic benchmark portfolios or time-series return regressions by applying them to a sample of active money managers and passive indexes. Estimated abnormal returns display large variation across approaches. The benchmarks typically used in academic research—attribute-matched portfolios from independent sorts, the three-factor time-series model, and cross-sectional regressions of returns on stock characteristics—track returns poorly. Some simple alterations improve the performance of these methods.]

Most mutual funds adopt investment styles that cluster around a broad market benchmark. Few funds take extreme positions away from the index, but those who do are more likely to favor growth stocks and past winners. The bias toward glamour and the tendency of poorly performing value funds to shift styles may reflect agency and behavioral considerations. After adjusting for style, there is evidence that growth managers on average outperform value managers. Though a fund's factor loadings and its portfolio characteristics generally yield similar conclusions about its style, an approach using portfolio characteristics predicts fund returns better.

Academic and practitioner research evaluates portfolio performance using size and value/growth attributes or factors. We assess the merits of popular evaluation procedures based on matched-characteristic benchmark portfolios or time-series return regressions by applying them to a sample of active money managers and passive indexes. Estimated abnormal returns display large variation across approaches. The benchmarks typically used in academic research--attribute-matched portfolios from independent sorts, the three-factor time-series model, and cross-sectional regressions of returns on stock characteristics--track returns poorly. Some simple alterations improve the performance of these methods. The Author 2009. Published by Oxford University Press on behalf of The Society for Financial Studies. All rights reserved. For Permissions, please e-mail: [email protected]., Oxford University Press.

Most mutual funds adopt investment styles that cluster around a broad market benchmark. Few funds take extreme positions away from the index, but those who do are more likely to favor growth stocks and past winners. The bias toward glamour and the tendency of poorly performing value funds to shift styles may reflect agency and behavioral considerations. After adjusting for style, there is evidence that growth managers on average outperform value managers. Though a fund’s factor loadings and its portfolio characteristics generally yield similar conclusions about its style, an approach using portfolio characteristics predicts fund returns better.