China's Stock Market Rate Of Return Calculation Method Yields The Distribution Of Empirical Research

After entering the WTO, China has its financial market gradually open up, and the stock market, as a cornerstone of the nation's economy, rises to a more prominent position. Thus, the stock return, widely known as a fundamental concept of financial theories and a direct index to appraise investment practices, has attracted more attention. Unlike Western countries, whose security markets have enjoyed flourish after a long time development and having already set up a mature theoretical model and research approach on the proposition of stock returns, Chinese stock market has just emerged and is far from perfection. Therefore, the way of duplicating the Western models onto the present Chinese stock market has become a question drawing public concern.This thesis, based on a profound analysis on the computation methods of stock returns, tries to investigate the distribution of stock returns from empirical perspectives. Firstly, by employing statistical methods, this paper, under the classical Normal distribution hypothesis, gives an inference on how different returns shift between each other and the deviation when using the distribution of log returns to substitute for the distribution of simple returns. Then, on the other hand, this thesis, countering the traditional normality hypothesis, attests that the distribution of stock returns in China exhibits fat tails and high peaks, and is in a symmetrical pattern. In order to testify the previous findings, some alternatives of Normal distribution, such as Scaled-t distribution, Logistic distribution, Exponential Power distribution as well as Mixtures of tow normal distribution, are fitted into the data drawn from Chinese stock market, and finally it finds out that the Scaled-t distribution fits best while the traditional Normal distribution exhibits the worst fit.This thesis takes great effort to conduct analysis mainly in terms of the following three aspects:1. It systematically induces the calculation methods of stock returns, and briefly compares the characteristics of simple returns and log returns;2. It gives a further analysis on the relations of statistical characteristics between the simple returns and the log returns under the classical normal distribution...