Font Size: a A A

Comparative Analysis Of China Stock Market Risk Models

Posted on:2012-07-27Degree:MasterType:Thesis
Country:ChinaCandidate:D Q ZhangFull Text:PDF
GTID:2219330338465399Subject:Investment economics
Abstract/Summary:PDF Full Text Request
All the investors are running after the maximization of the profit and the minimization of the risk. Most of the financing institutions use a number of resources to research and develop the technology of the financing risk management, especially the central part and the foundation of risk management, which is the technology of risk evaluation. At present, VaR is used broadly and considered to be the most efficient technology. However, VaR model has some kind of deficiency, so that there are some deviation from the reality. Lots of solutions are proposed by literatures and CAViaR is one of them. CAViaR models combined quantile regression theory with VaR model and are more close to the reality. AAVS—CAViaR model is derived from CAViaR models.We analyzed AAVS—CAViaR model and four CAViaR models with 4849 data chose from SSEC, using the Nelder-Mead Simplex Method and the Quasi-Newton Method recommended by Engle and Manganelli. We used the function fminsearch and fminunc included MATLAB7.0. The Nelder-Mead Simplex Method is the method used in fminsearch and the Quasi-Newton Method is the method used in fminunc.In order to find the most suitable model for the China stock market, we tested the stability of the five models with the method of Chow test, using the data of SSEC from the year 2004 because the data before are highly influenced by the government policy. After the comparative analysis of the test results, we selected the AAVS—CAViaR model and the I-GARCH model as the most stable models. Since the I-GARCH model is symmetric while the security market should be asymmetric, we chose the AAVS—CAViaR model as our final choice.At the last, we gave some suggests for the China stock market:First, we may use the time-vary parameter model to improve the AAVS-CAViaR model. Second, we should enhance our supervision and control of the market to increase its transparency. Third, we should broaden the investment channel. Finally, in order to make investors more sensible, we have to popularize the knowledge of investment and financing.
Keywords/Search Tags:CAViaR, AAVS-CAViaR, quantile regression
PDF Full Text Request
Related items