Date of Award

11-2-2006

Degree Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Finance

First Advisor

Jason T. Greene - Co-Chair

Second Advisor

Stephen D. Smith - Co-Chair

Third Advisor

Richard D. Phillips

Fourth Advisor

David C. Nachman

Abstract

By construction, actively managed portfolios must differ from passively managed ones. Consequently, the manager’s problem can be viewed as selecting how to deviate from a passive portfolio composition. The purpose of this study is to see if we can infer the presence of superior skills through the analysis of the portfolio deviations from a benchmark. Based on the Black-Litterman approach, we hypothesize that positive signals should lead to an increase in weight, from which should follow that the largest deviations from a benchmark weight reveal the presence of superior skills. More precisely, this study looks at the subsequent performance of the securities corresponding to the largest deviations from different external benchmarks. We use a sample of 8385 US funds from the CRSP Survivorship bias free database from June 2003 to June 2004 to test our predictions. We use two external benchmarks to calculate the deviations: the CRSP value weighted index (consistent with the Black-Litterman model) and the investment objective of each fund. Our main result shows that a portfolio of the securities with the most important positive deviations with respect to a passive benchmark (either CRSP-VW or investment objective), would have earned a subsequent positive abnormal return (on a risk-adjusted basis) for one month after the portfolio date. The magnitude of this return is around 0.6% for all the funds, and can be as high as 2.77% for small caps value funds. This result is robust to all the performance measures used in this study.

Share

COinS