The Early Warning Study Of Systemic Financial Crisis:Based On Panel Data Logit Model

From1825economic crisis broken out in Great Britain to2010debt crisis in European, the financial crisis is always go hand in hand in the global process of economic development. With the global integration process, the relations between financial institutions are increasingly close, and potential systemic risks in the financial system continue to accumulate, compared with the previous, the financial crisis express as systemic crisis. According to the International Monetary Fund(IMF) report, there are114systemic financial crisis of93countries in1980-2002(Caprio、Klingebiel and Leaven,2003). The subprime mortgage crisis erupted in2007eventually evolved into the global financial crisis, the European debt crisis erupted in2010resulted many countries into the debt scrape these phenomena raised concerns about systemic financial crisis once again. As the rapid development in the financial industry, the systemic financial crisis becomes a new threat in the development of the global financial system. The method to guard against systemic risk, to avoid a systemic financial crisis, will cause a widespread concern. At the same time, the instability factors in the international financial environment increase gradually, we have to think about that the method to prevent crisis in China under the growing finance.This paper tries to build the early warning model of a systemic financial crisis on the basis of the early warning research. However each crisis occurred at different time, different countries, how to integrate these crises? Building the model requires the dependent variable, how to measure the variable of crisis? Which method is appropriate to use in the early warning model of systemic financial crisis? For the previous question, this article uses panel data, the panel data not only take into account the time factors, but also consider the cross-section differences; for the latter two issues, this article uses a Logit model, the model is more mainstream early warning method of crisis. The crisis in building process is defined as0-1variables, and becomes into a probability. Comprehensive available, this paper takes the systemic financial crisis as the starting point, and uses the panel data Logit model to build the early warning model of systemic financial crisis. The selection of early warning indicators is the premise of the model, this paper build the required14early warning indicators based on the integrating indicator from the typical systemic crisis and indicators from the existing research. The study found that some early warning indicators have mutual high correlations, and this phenomenon will affect the accuracy of the model. This paper uses the global principal component analysis to extract the common factors, and uses the common factors to build the Logit model. The effect of the model needs the samples inside and outside tests, the better model will use in our country. The chapters are as follows:The first chapter is the introduction:this section presents the research background, the research value, the research ideas, the research content and the research methods.The second chapter is the review:first sector explains the definition of financial security, financial crisis and systemic financial crisis, and their links. Second sector reviews the four generations of crisis theory. Third sector take the review of existing early warning research.The third chapter is a typical analysis:common and characterization. Through the analysis, the paper wants to find the common feature of crises.The fourth chapter is modeling:first part describes the choice of sample sources, the index selection and the selection of warning method. Second part constructs Logit models both based on the original indicators and the common factor.The fifth chapter is the tests and application:the tests contain inside sample test and outside sample test. The way to inside sample test is putting back the sample into the model; the way to outside sample test is taking the sub-prime crisis to measure the model.The sixth chapter is conclusions and outlook:in the summary on the basis of the full text, the paper points out the inadequacies and the further research.Compares with previous research, this paper’s new ideas reflect in the following areas:first, this paper uses systemic financial crisis to build the early warning model; second, this paper uses panel data to establish the early warning model; third, this paper introduces the global principal component analysis, and using common factors to build the early warning model.According to academic level, limited documents and data sources, this paper has many inadequacies. First, the selection of early warning indicators may owe to complete; second, the model needs to be further optimized to improve the accuracy; third, the model needs to exclude confounding factors.