The Study of Company’s Credit Risk Measurement-Hybrid Model’s Application

• Main Authors: Jing-Lin Fang, 方敬霖
• Other Authors: Mei-Ying Liu
• Format: Others
• Language: zh-TW
• Published: 2008
• Online Access: http://ndltd.ncl.edu.tw/handle/mwkr9n

碩士 === 東吳大學 === 企業管理學系 === 96 === It is a quite important issue in the study of credit risk management to accurately predict whether the company will default or not, but now we already have too much models to predict the financial crisis, so it become another big issue to know which model is suitable for Taiwan’s financial environment. Maybe someone will ask the question-Why can’t we just use the outcome of foreign country’s study? But according to the past studies to compare the performance between Merton model and others financial cirsis predite model in Taiwan, we found that the ability of Merton model to predict financial crisis was not as good as foregine study’s outcome. So our study will focus on the assessment of four model’s performance –Logistic, Merton, Hybrid, and Cascaded Logistic models, and we also want to know what kinds of effort to the performance of model’s predicition will macroeconomic and industry make. Our study collect all the public companies’ data from 1998-2007, and use the concept of Ohlson’s (1980) research to fit every model’s predicting time point in every 5/1, and finally we use the Kolmogorov-Smironov (K-S) test to assess the performance of each model. According to the result of our research, the Hybrid model in K-S value is the highest, then the second one is Logistic model, the third is Cascaded Logistic model, primitive Merton model is the last, and the result of consider macroeconomic and industry’s effect show the model with macroeconomic and industry’s effect will have better ability to predict the company’s default.