The Multi-factor Capital Assets Pricing Model Based On Downside Risk And Its Empirical Analysis

The pricing of capital assets has been the focus of the financial field study andbeen the basic research problems. Traditional Capital Asset Pricing Model is used tomeasure the mean variance risk. But the risk is measured by variance has itsdrawbacks. But there are drawbacks to use mean variance to measure the risk.Because it contrary to investors real psychological feelings to the risk. Investorconcern is the possibility of earnings below expectations and the excess is notconsidered. So the methods of mean variance exaggerate the extent of the risk. Thefull text of a total divided into five chapters, as follows:ChapterⅠintroduces the topic of background and significance, the purpose ofthe study and research at home and abroad, and problems in the current situation.Finally notes the direction, ways of thinking and methods of the subject.ChapterⅡ, first reviewed the theory of financial economics, and introduces thetheoretical stage of infancy and prosperity. Then focus on the theory of capital assetpricing, from the establishment of CAPM to the development and improvement andto the existing problems of the model.ChapterⅢestablished a risk-based downside the Asset Pricing Model(D-CAPM). D-CAPM and the traditional CAPM are identical in form, only replaceβ_i in CAPM withβ_i~D which has the similar meaning withβ_i. Then on theβ_i~D -valued made a brief introduction, but also discussed theβ_i~D estimationmethod. The paper then randomly selected Shanghai Stock Exchange 50 A shares asa research object and introduce the methods of data processing. Finally we test thetraditional and improved single-factor model, from the test results can be seen thatthe improved pricing model is superior to the traditional model.ChapterⅣwe raise the three-factor Capital Assets Pricing Model Based ondownside risk. In practice, in addition to B some of the factors related with thefeatures of the company also has obvious explain ability of the stock yield. Thesefactors include the company size(ME),price-earnings ratio countdown(E/P) and book-to-market value(BE/ME) etc. Therefore, we propose to the third chapter of thesingle-factor model to improve again, and we add in the factors of ME and BE/MEto build a new model. Then we test the traditional and improved three-factor model,the result shows once again that both in the ability to explain or adjustedgoodness-of-fit, improved model of the three-factor should shown the superiority inthe theory.ChapterⅤ's main work is a summary.