Premium
The Style Consistency of Hedge Funds
Author(s) -
Gibson Rajna,
Gyger Sébastien
Publication year - 2007
Publication title -
european financial management
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.311
H-Index - 64
eISSN - 1468-036X
pISSN - 1354-7798
DOI - 10.1111/j.1468-036x.2006.00355.x
Subject(s) - hedge fund , returns based style analysis , consistency (knowledge bases) , econometrics , cluster analysis , business , fuzzy logic , computer science , actuarial science , open end fund , economics , finance , institutional investor , artificial intelligence , corporate governance
This study examines the style classification and the style consistency of hedge funds using a new proprietary database over the period May 1989 to April 1999. First, a hard clustering procedure is applied to classify hedge funds into homogeneous groups. It is shown that the methodology is robust and can be used to build stable hedge funds indexes. The method performs equally well as the principal component analysis in explaining in‐ and out‐of‐sample cross‐sectional hedge funds' returns. Second, we extend hard to fuzzy cluster memberships, relaxing the full assignment of the funds to individual clusters. We apply the fuzzy clustering methodology to estimate hedge funds' probabilistic exposure to various styles. We introduce three consistency indicators to quantify the hedge fund managers' style opportunism levels. We finally document that there is no evidence that style consistency leads to superior hedge funds' performance.