Applying Portfolio Change and Conditional Performance Measures: The Case of Industry Rotation via the Dynamic Investment Model

This paper applies portfolio change and conditional performance measures to assess the performance of the dynamic investment model in various industry-rotation settings spanning the 1934–1995 period. The dynamic investment model employs the empirical probability assessment approach in raw form. In addition, it incorporates three adjustments for estimation error: James–Stein, Bayes–Stein, and CAPM-based corrections. The tests are unanimous in their conclusion that the excess returns attained by the (unadjusted) historic, the Bayes–Stein, and the James–Stein estimators are (sometimes highly) statistically significant over the 1966–1995 and 1966–1981 sub-periods. This lends support to the idea that the joint empirical probability assessment approach based on the recent past, with and without Stein-based corrections for estimation error, contains information that can be profitably exploited. The relationship of these findings to the extant literature on momentum and contrarian strategies is addressed.