Analysis On Stock Prices Impacted By Unexpected Negative Events

With the development of social economy and the individual earning increase, the investment concept goes deep in people's mind, and this makes capital market became more and more active, including stocks market, bonds market, future market etc.. The capital market has developed quickly during the past 20 years, and the investment behaviors is turning to be more rational, so researchers want to find out some models that can describe the economic time series better. Based on this circumstance, I did some research on the stock prices which are impacted by unexpected negative events using the time series data of those listed companies.According to precious study, I made a definition on unexpected negative events under the research condition of this subject. In the article, all unexpected negative events are divided into two groups:one is those have direct impact on listed companies and the other is those have indirect impact. And there are three levels of each impact:particularly serious, serious, light. So an unexpected negative event is belonged to one of the six groups:direct and particularly serious impact on listed companies, direct and serious impact on listed companies, direct but light impact on listed companies, indirect but particularly serious impact on listed companies, indirect but serious impact on listed companies, indirect and light impact on listed companies.For the research purpose, I collected those unexpected negative events that took place from January 2005 to March 2011 and then fond out related companies on list and divided those impacts into six groups. I got 200 samples for study,150 of which took place between January 2008 and March 2011.Using the ADF method to test whether there are unit roots in the time series, then I can make a conclusion that whether the time series is stationary or not. If the time series is non-stationary, using difference method to make it stationary. Choose one of the combined modesl among ARMA-ARCH, ARMA-GARCH, ARIMA-ARCH, ARIMA-GARCH models. Normally, we can get an equation that can describe the time series, then forecast next value with this equation, and calculate the volatility of stock prices. This volatility is the benchmark to evaluate the impact.After numerous calculations and tests, I got two groups of unexpected negative events that are effective. One is the group which has direct and particularly serious impact on listed companies; the other is the one which has direct and serious impact on listed companies. The samples are 31 and 60. Test for equality between the two groups, I got the conclusion that those impact on listed companies are apparently different. Unexpected negative events in the group "direct and particularly serious impact on listed companies" decreased the listed companies'stock prices by 8.48% on average, and those unexpected negative events in the group "direct and serious impact on listed companies" decreased the listed companies' stock prices by 5.53%...