Abstract
In this study we consider two methods of returns based style analysis for classification of investment styles for a single asset class, US Diversified Equity Funds. We extend Sharpe’s (1992) style Returns Based Style Analysis (RBSA) by forming style groups using cluster analysis and RBSA factors. We also introduce a parsimonious Best Fit Index (BFI) of style classification which explicitly acknowledges the existence of market segmentation and practitioner benchmarking. The methods provide complementary information about mutual fund returns. Both methodologies explain a significant proportion of the cross section of out of sample returns, but the BFI method performs better out-of-sample is more transparent and more closely aligned to investment practice.