Font Size: a A A

Statistic Analysis Of APT Model And Research Of Application In China's Stock Market

Posted on:2003-02-04Degree:MasterType:Thesis
Country:ChinaCandidate:J R LiFull Text:PDF
GTID:2156360062980646Subject:Applied Mathematics
Abstract/Summary:PDF Full Text Request
Statistic Analysis of APT Model and Research of Application in China's Stock MarketLi Jiao- ruiAbstract In the course of finance mathematics, CAPM ,which bases on expectation-variance analysis, describes a market determined mechanism of the capital asset profits . when capital market reaches balance. It interprets that why different portfolios has different results.CAPM is a quite perfect model in theory In addition, it has the advantage of being operated simply and easily. However, it is not easy to test because it difficult to gain real market composition and more serious consequence is that it is conform to the experiment lesult .So , APT , another important pricing model , was established in the 1970s, it is superior to CAPM .Compared with CAPM , APT makes much less assumptions in which the market composition plays mo role. One of the basic assumptions of APT is that every one may take advantage of the chance when the investors have the chance to increase their returning rate without taking any risks梚ndividual is unsatisfacted. Another important assumption is that there are all kinds of independent stocks in the stock market, which makes APT easily be checked by the market .This paper deals with statistics analysis about factor choose, model improvement and other concrete questions in APT, and experimental test about Shenzhen's stock market. This essay can be divided into 3 chapters. In chapter one, we argue about how to choose factors with influence stock market results, build orthogonal factor model and give the factors meaning of matrix.In addition, we improve APT in three ways. Firstly quote suppositional Variable, which makes APT in consistent with the economic objectivity in reality. Secondly, revise factor coefficient with probability distribution , which given by experienced experts. Thirdly ,use Bayes statistic deducing method to bind together the income rate of prior distribution and sample in formation, which makes forecast stocks in Shenzhen stock market as samples . Work out the series of weakly income rate. Do experimental test on APT model and improved model.The result shows that Shenzhen stock market almost conform to APT model. In chapter three, we introduce the use of martingale in APT , also give some useful results.
Keywords/Search Tags:APT model, improvement of APT, experimental test, method of Martingale
PDF Full Text Request
Related items