Indexing Investment Models And Performance Evaluation Based On Tracking Error

Portfolio investment is the main modern security investment strategy. Traditional investment strategy take a positive approach mainly including stock analyzing and time choosing. Lots of research and data show that the traditional positive investment strategy can not beat the market, while the indexing investment strategy, represented by index funds which began to pop up in the 1970s, have gained the same or even higher level of performance than the market average returns. Therefore indexing investment and indexing investment models received more and more attentions.As theoretical circles and investment communities have put more attentions on indexing investment and indexing investment models, the need of professional fund managers to evaluate overall performance has been induced and the comparison between portfolio's return and the announced benchmark's return has to be built. Against such a backdrop, the study of tracking error, indexing investment models and performance evaluation which are the core aspects in indexing investment and management of index funds is of tremendous theoretical and practical significance.This paper's main line falls on tracking error which is the core concept of indexing investment and index tracking technology. On the grounds of a comprehensive literature summary in domestic and abroad researches of index tracking, this paper make a in-depth research, from the perspectives of theory and application, on tracking error and tracking error risk, indexing investment models based on tracking error, the comparisons between indexing investment models which are derived from different covariance matrix and the performance evaluation system of index fund based on the tracking error. We aim to provide useful guidance to index funds and the innovation of index derivative product which are still in the developing stage in China securities market.The full text is divided into six chapters, the specific structure and contents are as follows:The first chapter is the foreword. The background, research significance, content, structure and innovations of this paper have been outlined.Chapter II provides a study review of index tracking error. This chapter describes the background and theoretical basis of index tracking error, summarizes the research of index tracking error, domestic related researches and status of index investment.Chapter III introduces tracking error and tracking error risk. This chapter presents fundamental issues of which call for clear descriptions: the theory and definition of tracking error and tracking error risk; a focus investigation of the tracking error and tracking error risk from angels of basic concepts, measure indicators, characteristics and influencing factors; not least the introduction of tracking error risk indicator RVaR and the comparison between TEV and RVaR.Chapter IV emphasizes on indexing investment models based on tracking error and empirical study of the models.The systematic study of the tracking error-based indexing investment models in the framework of traditional mean-variance model is conducted in this chapter. First of all, it briefly introduces the traditional mean-variance model and efficient frontier. Then it focuses on two tracking error-based indexing investment models and the empirical comparative study of the two models.Chapter V works over comparison between indexing investment models which spring from different covariance matrix. This chapter thoroughly discusses the impact of different covariance estimation methods on indexing investment models. In the first place, this chapter put emphasis on significance of covariance matrix estimation to portfolio models and different covariance estimation methods. Then empirical study and comparative study are employed to indexing investment models within the estimated samples according to different covariance estimation methods. In the last, we perform a fitting test within fitting samples intervals.Chapter VI sums up the tracking error-based performance evaluation of index fund. In view of the importance of tracking error in index funds and index tracking management, this chapter injects the notation of tracking error risk into performance evaluation which consists of fund performance evaluation theory, fund performance evaluation based on the tracking error, empirical evaluation based on the track error. Some meaningful conclusions have been emerged from this paper:1. Through literature study, we comprehensively review the background and theoretical basis of the index tracking error; demonstrate the feasibility and effectiveness of indexing investment, the rationality of index tracking management, and the importance of index tracking error research; clear the main line of our research, which is also the core concept of indexing investment and index tracking management, that is, tracking error.2. Tracking error and the tracking error risk are the basic issues that should be specific at first among the emphases in this research. This paper did an emphasis research on tracking error and tracking error risk from the aspects of basic concepts, measure index, characteristic and influencing factors. To measure the tracking error risk, a tracking error risk measure indicator RVaR is proposed, this is defined by VaR downside risk measure. The validity of this indicator for measuring tracking error risk reflected in the validity of improved information ratio for evaluating the performance of index funds. The empirical result shows that RVaR and information ratio are both efficient.3. This paper makes a scenario that tracking error is the risk control constraint conditions of index investment model and makes a deep empirical comparison research on index investment TEV model and C-TEV model in the aspects of construction of portfolio position, portfolio performance and index tracking effect. The empirical result shows that, both of the TEV model and the C-TEV model can enhance the income of index investment combination comparing with the benchmark income. But the C-TEV model perform better in the balance of risk control and increasing income, especially when the efficiency of benchmark is comparatively low, it is reasonable to add risk control constraint conditions in the model.4. Different covariance matrix estimation measures can affect the index tracking and tracking error evidently. This paper adopt 5 different covariance matrix estimation measures, which are sample covariance matrix V, single model matrix F, constant matrix C, scalar matrix K and two-parameter model matrix P, to research on TEV model and C-TEV model by constructing portfolio within samples and fitting test outside samples. The result which make the risk as the measure index shows that, when constructs portfolio within samples, the result from two-parameter model matrix P is close to benchmark index mostly. When fitting test outside samples, the TEV model can adopt constant matrix C to get a risk curve which is very close to the benchmark index, and adopt two-parameter model matrix P to get a same curve with the benchmark index. The empirical result from adoption of scalar matrix K underestimated the risk heavily in each model; it increased the tracking error in index investment and impacted the effect of tracking error.5. Performance evaluation of index fund based on the tracking error serves as an applied research on tracking error which is the backbone of this paper. This paper attempts to build performance evaluation system which is composed of investment style of index funds, tracking capability of index funds, historical performance and ability of selecting stocks and occasions and conduct an empirical study of performance evaluation of index funds available in the market. The results reveal that: There exists a strong positive correlation between daily yield of index funds sample and daily yield of tracking index in the sphere of investment style. Regression model has strong capacity of interpreting fluctuations of China′s index funds. Most of the non-systemic risk of funds samples is eliminated through the investment portfolio which follows that the strategy of investment portfolio lends a hand to reducing the non-systemic risk. In the aspect of the evaluation of index fund tracking capability, index funds which fail to strictly meet the investment objectives vary in the levels of controlling tracking error. In the process of evaluating history performance of index funds, taking advantage of classical methods (Treynor's ratio, Sharpe's ratio, Jensen's measure) and improved risk-adjusted methods (information ratio, improved Sharpe Ratio, improved information ratio), moment method based on investors preferences, this paper verifies that different methods have certain diversions and information ratio which is directly related to tracking error and tracking error risk, the improved information ratio and moment method based on investors preferences possess some reference value to the performance valuation of index fund. On occasion of the ability evaluation of selecting timing and stock, timing factors have more positive values than negative values with just a few to pass the significance test. Empirical results show that it boasts a weak (but not significantly) stock-selecting capacity, a weak (but not significantly) time-selecting ability.