Font Size: a A A

Some Statistical Analysis Of CAPM In Chinese Security: An Regression View

Posted on:2005-11-26Degree:MasterType:Thesis
Country:ChinaCandidate:J YangFull Text:PDF
GTID:2156360125961947Subject:Basic mathematics
Abstract/Summary:PDF Full Text Request
In this article, we arc mainly concerned that how to choose the rm, the "return of the Market" in the CAPM model, in the Chinese security market. The candidates are Shanghai composite index, 180 index and A-index. Our main tools are statistic regression models. Furthermore, viewing that the CAPM is just a special regression model between the return r of a single stock (or a portfolio ) and the market return rm, we study the following problem: Is there any regression model better than CAPM? How to find the best one? For this purpose, we use the famous Box-Cox transformation to construct a class of models which contains CAPM, and then use the well-established criteria, such as F statistic to form an optimization procedure. The results we get arc: For the problem of choosing of the " return of the market", the composite index is the best one in most cases but not in all 5 cases; For the problem of choosing the best regression model, CAPM is not the best regression models in many cases, and we give the better model in those cases. We believe that more importantly, these results show that when using CAPM, one must be very careful about the choosing of "the return of the market"', and that to find a reasonable model for analyzing, it is necessary and helpful to conduct a. regression analysis, for each single stock or portfolio.
Keywords/Search Tags:Portfolio theory, capital asset pricing model (CAPM), regressive analysis, Box-Cox transformation, optimal regressive model.
PDF Full Text Request
Related items