Research On Early Warning System Of International Financial Crisis

Since the 20th century, the international financial crisis has taken place for many times such as the Latin American financial crisis in 1995, the Asian financial crisis in 1997, the modern international financial crisis caused by American sub-prime crisis in 2008, etc. Each time the international financial crisis not only took a heavy toll on the economic development of the countries involved but massively impacted the global economic situation and financial order. With rapid financial globalization, the potential risks in financial markets increase and the factors that may give rise to the financial crisis and the international financial crisis have become more complex. Prepared for danger in times of peace, for many years scholars have never stopped their research on the early warning of financial crisis. Actually, scholars from various countries have raised different early-warning models of financial crisis, among which are some major ones:the subjective probability model, STV model, FR model, KLR models, DCSD model, ANN model, Simple Logit model and so on. Since the backgrounds where these models arise vary and differences exist in the research methods as well as the indicators of early warning, it is therefore difficult to determine which model best meets the demands from the modern society. Besides, their researches tend to serve a particular country and those on the early-warning system of the international financial crisis are relatively rare.The modern international financial crisis in 2008 sounded the alarm to the world’s financial markets once again. Because it varies from other crises in history in both inducing factors and transmission routes, it is urgent to reexamine the early warning of the financial crisis and the international financial crisis. As is argued in the current study, the financial crisis should be warned earlier in a single nation to prevent the potential international financial crisis. Although the available early-warning models failed to effectively predict the international financial crisis in 2008, some ideas and research methods of them are meaningful and adoptable. In the present research, several major early-warning models are discussed in detail to find out their advantages and disadvantages. Based on what they have achieved, the author constructs an early-warning system of the international financial crisis from the view of economic correlation, which proves effective when empirical tests are carried out with abundant sample data from many countries.The dissertation takes as the core a new early-warning system of international financial crisis, which is probed into with qualitative and quantitative approaches from four aspects:its construction, its verification, its application and its improvement. In particular, five chapters are included:Chapter One is the introduction. The background and significance of the present study are illustrated. Literature review is made on relative theories of financial crisis at home and abroad. The research methods and the main structure of the dissertation are described. At last, the innovation and possible improvement of this study are summarized.Chapter Two supplies a theoretical analysis framework for the early warning of international financial crisis. It works on the financial crisis’definitions and classifications as well as the contents and characteristics of the international financial crisis. Theories of the financial crisis in the past years are reviewed in detail. Besides, the warning indicators and warning methods are analyzed for some classic warning models of financial crisis to lay a theoretical basis for a new early-warning system of international financial crisis,In Chapter Three a new early-warning system of the international financial crisis is built based on economic theories and mathematical methods. The system is composed of five parts, i.e. research objects, warning indicators, optimal thresholds, crisis prediction and efficiency examination. The research objects are decided according to the research objects’economic correlation, which is measured by the weight of import and export trade among them; as to the choice of warning indicators, it is taken under consideration not only the available research and economic theories but also aspects such as similarity, availability, data frequency, etc.; the selection of the optimal threshold is made according to the noise-to-signal ratio and the method of VaR. Three bivariate Probit regressions are run on the pooled panel, where each indicator is inspected and the weights of those with more apparent regression results are set to calculate the probability weight of each nation in facing a financial crisis. At last, to see whether the system is efficient, the principle of the noise-to-signal ratio is again applied to choose the optimal threshold of warning probability. Then the achieved probability weight of financial crisis is compared with the occurred international financial crisis to test the validity of the new system in early warning the financial crisis and the international financial crisis.In Chapter Four the new early-warning system of international financial crisis is verified. The eight target countries for empirical research of international financial crisis are chosen by the ranking in the accruing amount of importing from and exporting to the United States. Results show that the early warning system is effective in predicting the Asian financial crisis in 1997 and the contemporary international financial crisis in 2008.Chapter Five provides the evaluation of the system and its practical significance in China. It first assesses the advantages and disadvantages of the new early-warning system, and then points out its possible improvement of in the future research. After that, an empirical analysis is made to discuss how the new early-warning system can be applied in China and what can be further done in Chinese early-warning research of financial crisis.