China's Financial Crisis Warning: Based On The Empirical Analysis Of Logit Model

Since 70's last century,the international financial system suffered impact frequently,and countries all over the world sufferred financial crisis. Especially in 2006, the subprime crisis break out in the United States evolved into a financial crisis and engulfed the globe, bringing a great economic loss to our society. The accession into the WTO speeds up China's globalization; it makes China's dependency on foreign trade increase obviously. As the recession sweep all over the world, China can not avoid and the foreign exchange risk is increasing. Therefore, this article tries to study the early warning of China's financial crisis with the help of binary Logit model.The first section of this thesis is the introduction where the background and purport of this article are introduced. Literature related to the research on early warning system from home and broad will also be reviewed in this part. Then the main contents and methods are briefly described in this passage. Three generations of financial crisis model are introduced in the second part firstly, on which the research of early warning on financial crisis is theoretically based. Then three major models of early warning will be elaborated: Subjective probability method, STV cross-sectional regression method and KLR signal analysis method; and their defects are also simply illustrated. In view of the defects of three methods above, binary Logit model is first introduced in the third part of this paper. Then on the basis of foreign definition for crisis, the paper presents foreign exchange market pressure index (EMP) which is appropriate for Chinese actual situation and establishes Chinese financial crisis in early warning indicators according to EMP. Finally, the main factors which have impact on Chinese foreign exchange market are analyzed and 6 indexes for the study on financial crisis are primarily selected: the growth rate of GDP, the growth rate of domestic credit, the growth rate of export etc. In the fourth section, the warning indicators primarily selected are examined respectively with EMP by Granger Causality Test. The result of the test shows that several indexes have obvious impact on economy: the growth rate of GDP, the growth rate of nominal exchange rate between China and USA, the growth rate of export rate and the growth rate of M2/foreign exchange reserves. While these warning indexes are Logit estimated with the warning index of financial crisis, the growth rate of GDP, nominal exchange rate between China and USA, M2/foreign exchange reserves rate are found fitted well. Subsequently, early warning indexes are shortly predicted respectively with ARIMA and the probability of crisis take place from June 2010 to June 2011 is obtained by the Logit model established. According to forecasting results, The relevant policy recommendations are put forward.