Statistical Analysis Of Overreaction Of Stocks Based On Intraday Price

In the modern financial theory, as the representative of the financial anomalies op-posite to the Efficient Market Theory (EMH) may be the overreaction phenomenon. The first research of overreaction phenomenon was Debondt and Thaler's article,"Dose the stock market overreact?(1985)", which has showed,there was long-term price reversal phe-nomenon in U.S. stock markets. But the traditional finance theory just viewed the over-reaction phenomenon as the " anomaly".Fama and French(1998) thought that abnormal returns comes from risk compensation, and tried to introduce a new factor to characterize the risk. Interpretation of the traditional financial theory is not convincing. At the same time, the explanation given by behavioral finance scholars and the related models seem to be more persuasive. For example:Barberis, Shleiferv and Vishny with BSV model, Daniel,Hirsheifer and Subramanyam with DHS model,and Hong and Stein(1999)with HS model. Many scholars has conducted a large number of empirical tests for the stock market overreaction, but the conclusions differ. One of the most important reason is the difference of the definition of overreaction standard and testing methods. In this article, we discuss the improvement about the definition and testing methods.The first chapter presents the research background,existing problems of the overre-action. The second chapter, from a quantitative approach, under the assumption with the stock intraday price follows the geometric Brownian motion, we construct a new mea-sure of volatility. Based on that, we derive a test statistic which detect the inspection days price overreaction from the point of view of the Hausman test and its theoretical distribution. The third chapter presents the robustness of the test statistic under the volatility changes(GARCH), the drift rate changes(AR), and the volatility and drift rate changes(AR-GARCH)with the corresponding correction test statistics. The last chapter, we give the empirical test of the global market with the overreaction statistic test, and discuss the robustness and applicability of the correction of the test statistic based on the actual stocks data. Finally, we're try to explore the Trading Strategy based on the overreaction of the test statistic F.