Dynamic Return-Based Classification of European Mutual Funds
46 Pages Posted: 15 Aug 2015
Date Written: April 30, 2015
Abstract
In this study, we analyze the monthly fund migrations in a return-based classification scheme for European mutual funds. Accounting for the time-varying commonalities in returns, we derive a classification scheme consisting of four layers with two, six, 15, and 20 groups respectively. In the two-group solution fund migrations are low while in the classification with six groups migrations are low during normal times but high during the financial crisis for stock funds and during the debt crisis for bond funds. Fund migrations in the classification with 15 and 20 groups are high overall. In line with prior research, we confirm that return-based classifications are better able to explain the cross-sectional differences in mutual fund returns than existing industry classifications based on portfolio holdings and self-declared investment objectives.
Keywords: Fund Migrations, Hierarchical K-Means, Mutual Funds, Return-Based Classification, Transition Probabilities
JEL Classification: C63, G23
Suggested Citation: Suggested Citation