Clustering financial time series: an application to mutual funds style analysis

Classi#cation can be useful in giving a synthetic and informative description of contexts characterized by high degrees of complexity. Di3erent approaches could be adopted to tackle the classi#cation problem: statistical tools may contribute to increase the degree of con#dence in the classi#cation scheme. A classi#cation algorithm for mutual funds style analysis is proposed, which combines di3erent statistical techniques and exploits information readily available at low cost. Objective, representative, consistent and empirically testable classi#cation schemes are strongly sought for in this #eld in order to give reliable information to investors and fund managers who are interested in evaluating and comparing di3erent #nancial products. Institutional classi#cation schemes, when available, do not always provide consistent and representative peer groups of funds. A “return-based” classi#cation scheme is proposed, which aims at identifying mutual funds’ styles by analysing time series of past returns. The proposed classi#cation procedure consists of three basic steps: (a) a dimensionality reduction step based on principal component analysis, (b) a clustering step that exploits a robust evolutionary clustering methodology, and (c) a style identi#cation step via a constrained regression model #rst proposed by William Sharpe. The algorithm is tested on a sample of Italian mutual funds and achieves satisfactory results with respect to (i) the agreement with the existing institutional classi#cation and (ii) the explanatory power of out of sample variability in the cross-section of returns. c © 2003 Elsevier B.V. All rights reserved.