An Empirical Research On Tracking Error Of Exchange Traded Fund In China

After 2006 and 2007 the two years of rapid development, the fund industry is becoming dazzling in the stock market. Either, after the drastic adjustment of stock market in 2008 and the strong rebound in 2009, the exchange traded fund becomes the fast developed fund product in the fund market for its lower investment costs, higher efficiency of resource allocation and transparent and simple operation, meanwhile, which is favored by most of the investors. By the product design, the ETF is one kind of index fund which tracks and clings the underlying index, if the trend of ETF coincides with the underlying index, it's a successful operation for the ETF. However, the trend of ETF in reality is different from the underlying index, it means the yield of ETF is not always equal to the yield of underlying index, and the tracking error is an important indicator to reflect risk and investment value of ETF. In order to understand the tracking error of ETFs, five ETFs are chosen as research objects whose operation periods are more than three years: Shanghai Stock 50 Exchange Traded Fund (shorted for SSE-50 ETF), Shanghai Stock 180 Exchange Traded Fund (shorted for SSE-180 ETF), E Fund Shenzhen Stock 100 Exchange Traded Fund (shorted for SE-100 ETF), Small and Medium Enterprise Board Exchange Traded Fund (shorted for SME Board ETF), Shanghai Stock Dividend Exchange Traded Fund (shorted for SSE Dividend ETF), and three different methods are adopted to analyze the tracking error of ETFs in China, on the basis of the three methods, the thesis shows the comprehensive evaluation of ETFs by the set pair analysis (SPA). Next, the thesis also analyzes how the purchase and redemption of ETF, dividends and transaction costs influence the tracking error. Based on the analysis above, this thesis not only provides the empirical foundation for the ETF investors to choose the appropriate fund products, but also helps the fund manager recognize the problems exist in the operation and management process of ETF, meanwhile, it also helps the investors invest rationally and the fund operate efficiently and improve the investing efficiency of ETF.With Eviews software, the thesis chooses five ETFs from the stock market in China and then analyzes empirically tracking error of five ETFs, the results are:1. Through three measurement methods of tracking difference volatility, absolute mean deviation and regression analysis, it shows that the tracking effect of SSE-50 and SME Board ETF are not good enough which don't achieve the tracking aims. We advise investors to reduce the investment of SSE-50 and SME Board ETF and also suggest the fund manager to strengthen the management and operation of the two ETFs.2. The tracking error is the result of calculation, so the sampling frequency and the autocorrelation of return sequences could influence the result in a certain extent during the process of calculation. After the empirical analysis, the tracking error of ETFs will increase with the decrease of sampling frequency of day, week and month. In order to make sure the calculation results are comparable, the results should be annualized based on different frequency.3. On the basis of tracking difference volatility, absolute mean deviation and regression analysis,the thesis shows the comprehensive evaluation of ETFs by the set pair analysis (SPA) which is the tracking error of SE-100 ETF and SSE Dividend ETF have the identical degree with the real tracking error respectively.4. The tracking error of ETFs are mainly affected by three factors: the purchase and redemption of ETF, dividends, fund management fee and custodian fee and some other fees. In fact, the tracking error of ETFs are affected by many factors which are not independent, meanwhile, they have a complicated influential relations. But in a certain extent, some factors are controllable for the fund manager, so the fund manager should strengthen the management of controllable factors and reduce the tracking error between ETFs and the underlying index.This thesis is mainly consisted of three contents:1. According to the development, characteristics and study history of ETFs, this thesis shows the theory and calculating model to measure the tracking error, and then design the empirical research plan so that the tracking error of ETFs in China could be calculated.2. The adjusted net asset value of these five ETFs from July 1, 2007 to April 30, 2010 are adopted to calculate the daily and annual tracking error of each ETF by tracking difference volatility, absolute mean deviation and regression analysis, and compare the difference among these three methods. Based on that, the thesis shows the comprehensive evaluation of ETFs by the set pair analysis (SPA) which makes the result reliable.3. The mainly factors which effect the tracking error of ETFs in China are analyzed empirically and the relations between these factors and tracking error are represented pictorially.