Active Share and the Predictability of Portfolio Performance: Out-of-Sample Evidence from Separate Accounts
35 Pages Posted: 20 Apr 2020
Date Written: April 2020
We use a novel sample of separate accounts to perform an out-of-sample test of the predictive power of active share (Cremers and Petajisto, 2009). While active share has limited predictive power unconditionally, it has significant power conditional on past performance. We find strong positive performance persistence only for high active share separate accounts: the portfolio of separate accounts with high active share and strong past performance subsequently outperforms by 1.48% per year. That result is even stronger when return dispersion is high and among separate accounts with a small cap style, a fundamental investment approach, or low cash holdings.
Keywords: Active share, separate accounts, mutual funds, alpha, out-of-sample, predictability
Suggested Citation: Suggested Citation