The Construction Of Index Fund And Evaluation Of Performance

China has joined the World Trade Organization at the beginning of this century. Sincethen, the Chinese economy has boomed for almost a decade. During this decade, manycompanies went public because of the growing of the earning power. Meanwhile, some newstock indexes were introduced into Chinese stock market, which marks the very beginning ofthe index funding time. Portfolio theory is the main strategy of modern security analysis.Traditionally, mutual fund managers tried to analyze stocks and timing to make the bestchoice. This is called active investment. Although the stock market has been through ups anddowns many times, these active mutual funds always failed to beat the market. The Paretophenomenon shows us when the market index is going up, the prices of some stocks can stillgo down. That is why active investment cannot have a better rate of return than the market.The sub-prime crisis hit the U.S. economy badly in2008, and led to a world-widefinancial crisis. In the same year, the Chinese stock market went down over50%, and sincethen, it has been a bear market till now. When the index went down, the net value of mutualfunds went down even more. Under these circumstances, the passive investment became moreand more popular in China. The number of index funds is growing rapidly in the past fewyears. Investors and investment agencies who believe in index trading have to build an indexportfolio regularly, so the model of index tracking is the foundation of index portfoliomanagement. A good tracking model can minimize the tracking error between the portfolioand the index.This paper introduces the history and the process of index tracking in the beginning, andthen choose Hushen300index as the aiming index. By using optimized replication method,the process of index tracking can be realized.Firstly, in the stock picking process,3important factors are considered. They are stockbeta, average trading amount and market capitalization.Secondly, calculate different tracking error under different circumstances. Set the numberof stocks at30or60in the portfolio, and then use genetic algorithm to calculate the trackingerror as the cost of trading goes up. The results show that1) when the number of the stocks inthe portfolio is the same, the tracking error goes down as the cost goes up.2) when the cost isfixed, the portfolio with more stocks has a higher tracking error.In the end, evaluate the performance of the portfolio by using excess rate of return, Jensen index, Sharpe index, and information rate.This research shows that the performance of index tracking portfolio does not change alot as the cost of trading changes. Theoretically, full replication method is the perfect one inindex tracking. However, in the real world, it is totally unprocurable, especially in China. TheChinese stock market is still not mature, that is why optimized replication method is the bestchoice for index fund managers in China.