The Optimal Foracst Period Of The Credit Risk Of Listed Compinies In China

Since the last century, several major financial crises bring turmoils and huge losses to the global economy. Thus the global financial industry begin to focus on the research and promotion of credit risk management. From the beginning of the1950s, the banking had already developed the classic credit analysis system. As research continues developing, quantitative measurement of credit risk models are constantly achieving progress and improvement. Structural model of credit risk is one important kind of quantitative credit risk model, providing important and useful tools and reference of credit risk management to financial institutions. With the development of the market economy in China and the deepening of economic globalization, financial institutions in China are required to improve their credit risk management capabilities more and more urgently. In this article there is a retrospective of the development history of the structural model, and use the Leland-Toft model for theoretical and practical research from the perspective of the credit risk management of the commercial banks credit risk management, in the hope of providing useful references for commercial banks’credit risk measurement and management.This article focuses on sorting out the development and principles of structural models, and selects the LT-model for making an empirical analysis based on the data of A share market in China. The main contents include:(1) Elaborating the development process of the Credit risk analysis system, recalling the development of major credit risk analysis methods, collating the current research situation of credit risk at home and abroad;(2) Stating the birth, development and characteristics of structural models, highlighting the Merton model and the initial representative LS model and LT model,introducing the construction framework, basic principles and characteristics of each model, and giving the calculation methods and formulas of the expected default frequency;(3) Applying the endogenous default boundary model-LT model to the A share market listed companies in China, in order to explore whether in the medium to long term, the LT model could identify the different categories of listed companies, and finally get the result that in the medium to long term,the ability of identification of the LT model can be effective within a certain time frame. Then this paper attempts to identify the best time for predicting credit risk through the using of misjudgment ratio and drawing ROC curve.In this study we draw:(1) in the case of the same forecast period, generally ST class company with respect to the non-ST companies have a higher probability of default, in line with our economic intuition;(2) the LT model can effectively distinguish between different types of companies based on relevant accounting data in3years or less, and when the forecast period is3.5to5years, the discriminant result is not satisfactory;(3)when the commercial banks choose to use LT model to estimate the expected default rates of the target company at a future point, the predict period should be controlled in less than3years (including3years), in order to ensure the model’s working;(4)When using the LT model to predict a company’s future credit risk, the optimal forecast period is1.5fiscal year ahead.