A Study On Passive Indexing Strategy Based On Tracking Error Minimization

Indexing strategy is a way of passive investing strategy, aiming to track target indexand minimize the tracking error between index return and indexing portfolio return. Since2002, the first batch index funds launched in mainland China market, indexing strategy inthe mainland has been more than ten years.In recent years, China’s macro economy continues to improve, with rapid developmentof the securities market as a result, indexing products get broader space for developmentunder the background of economic prosperity. As of the end of August2013, China’smainland indexing products soared to248from22of2008, the size of the assets soared to343.8billion yuan from107.8billion yuan of2008. At the same time, active investingstrategies are experiencing poor performance, indexing products are attracting more andmore investors’ attention, especially for institutional investors. With great development ofindexing strategy, exploring how to build an investing portfolio which aims to minimizetracking error is of great theoretical and practical significance.Based on theoretical analysis of passive indexing strategy and tracking error, thispaper proposes tracking error minimization capital allocation model. In the part ofempirical research,in order to improve the method of stock selection based on correlationcoefficient with the order from highness to lowness and the method of optimization basedon nonlinear programming, use the method of tracking CSI300Index, based on K-MeansClustering in stock selection, while Genetic Algorithm is applied to optimize the allocationof funds under certain constraints, and achieve the goal of constructing optimal indexportfolio with minimized tracking error. The empirical results show that the combination ofK-Means Clustering and Genetic Algorithm in constructing an indexing portfolio can getsmaller tracking error and achieve better simulating effect. This paper explores how to setup passive indexing strategy based on tracking error minimization from aspects oftheoretical analysis and empirical research.