文章基本信息

标题：Cluster analysis for regime identification and forecasting with application to the enhanced index tracking problem
本地全文：下载
作者：Wanderlei Lima de Paulo ; Oswaldo Luiz do Valle Costa
期刊名称：Investment Management & Financial Innovations
印刷版ISSN：1810-4967
电子版ISSN：1812-9358
出版年度：2014
卷号：11
期号：3
页码：44-52
语种：English
出版社：LLC “Consulting Publishing Company “Business Perspectives”
摘要：This paper deals with the asset allocation problem in the presence of regime switching in asset returns.Considering a financial market subject to changes in regime,it is assumed that the expected value and covariance matrix of the returns of the assets can change according to a Markov chain taking values in a finite set.Generally,to apply a portfolio selection model in a switching regime approach it is necessary to estimate the market parameters and determine the number of regimes.In this paper it is proposed a non-parametric procedure to determine the number of regimes and define in which regime the market belongs to along the time,based on analyzing the historical stock return patterns using cluster analysis tools.The proposed methodology is applied to a portfolio optimization problem with enhanced index tracking and switching regime.The results show a satisfactory performance of the model with regime switches when compared to the case without regime switches.
关键词：switching regime;cluster analysis;enhanced index tracking;Markov process.