Dynamic Excess Autocorrelation and Mutual Fund Performance

AFA 2014 Philadelphia Meetings Paper

86 Pages Posted: 13 Mar 2013 Last revised: 31 Mar 2020

See all articles by Xi Dong

Xi Dong

CUNY Baruch College

Massimo Massa

INSEAD - Finance

Date Written: January 1, 2014

Abstract

Positive return correlation signals slowly-diffusing information. Short sell-constrained institutions are mainly informed in their buy trades. Building on these facts, we identify informed investors ex ante by focusing on mutual funds. We propose a measure of the dynamic excess autocorrelation (DEA) of fund portfolios. High-DEA funds persistently outperform by 3.5% annually. Our framework shows that the DEA identifies informed managers who dynamically rebalance to buy portfolios with slowly-diffusing positive information, which are thus positively autocorrelated and subsequently outperform. We support such intuition with high- and low-frequency evidence. The DEA performance predictability stems from a novel portfolio-level, cross-stock linkage.

Keywords: Mutual Fund Performance, Serial Correlation, Informed Traders

JEL Classification: G11, G14, G23

Suggested Citation

Dong, Xi and Massa, Massimo, Dynamic Excess Autocorrelation and Mutual Fund Performance (January 1, 2014). AFA 2014 Philadelphia Meetings Paper, Available at SSRN: https://ssrn.com/abstract=2232397 or http://dx.doi.org/10.2139/ssrn.2232397

Xi Dong (Contact Author)

CUNY Baruch College ( email )

Massimo Massa

INSEAD - Finance ( email )

Boulevard de Constance
F-77305 Fontainebleau Cedex
France
+33 1 6072 4481 (Phone)
+33 1 6072 4045 (Fax)

Do you have a job opening that you would like to promote on SSRN?

Paper statistics

Downloads
154
Abstract Views
2,163
Rank
485,678
PlumX Metrics