| Summary: | 碩士 === 國立高雄應用科技大學 === 金融資訊研究所 === 96 === In this paper we combine traditional default forecasting model based on financial accounting analysis and Merton model into a Hybrid model of credit risk measurement. This paper investigates whether the Hybrid model can boost predictability of traditional default forecasting model. At first we explore the predictability of traditional default forecasting model based on financial accounting information. Then, we refer to the previous research paper to consider some variables about corporate governance in the traditional forecasting model. Finally, plugging Merton model into this traditional default forecasting model, we explore the implement and application of this Hybrid model in all listed companies in Taiwan. Our research can provide domestic banks with some useful suggestion in development of credit model.
Our observations include 85 listed companies with financial distress versus 85 non-defaulted ones during the period of 2000~2005 (we exclude companies with financial distress in financial services industry). We employ the Logistic regression and Kolmogorov-Smirnov test to investigate the relationship of financial accounting variables, corporate governance variables, default probability in Merton model and the probability of business financial distress. Empirical results show that quick ratio, operating margin, accounting receivable turnover rate, inventory turnover rate and ownership of directors and supervisors have negative significantly link with possibility of business financial distress; debt ratio and pledged ratio of directors and supervisor have positive significantly link with possibility of business financial distress. We also find information is much more close to the time when financial distress happened, the accuracy of traditional default forecasting model is enhanced. In addition, the Hybrid model which combine Merton model is superior to traditional default forecasting model in accuracy of predictability. This framework can set a good benchmark for financial institution in Taiwan to set up a forecasting model of credit risk. As a result, we suggest they can combine the Merton model with financial accounting information into a complete forecasting warning model of credit risk. Thanks to the creation of immediate information in Merton model, it can help enterprises, who only integrate financial accounting information, to manage credit risk.
|
|---|
