Credit Default Prediction for Small-Medium Enterprise

• Main Authors: Chi-Jui Chung, 鐘啟瑞
• Other Authors: Rwei-Ju Chuang
• Format: Others
• Language: zh-TW
• Published: 2015
• Online Access: http://ndltd.ncl.edu.tw/handle/6224vq

碩士 === 輔仁大學 === 統計資訊學系應用統計碩士在職專班 === 103 === Taiwan enterprises, mostly SMEs, so that when banks doing business loan, mostly belonging to the small and medium enterprises, but the problem often facing SME financial statements the Bank, therefore how to company data and the financial statements of the financial data contained in the messages revealed to find the clue of possible default, credit quality is relatively important. This study in order to understand the General SME when the breach occurred, its financial statements, financial ratios for the possibility of default factors, through banks of enterprise credit rating listed in the tables of the financial variables and non-financial variables, while also increasing selection its not listed in the other financial variables in order to collect data, using independent samples t Test, χ² test, decision tree analysis and logistic regression to analyze its probability of default, and to establish a default model. Decision tree analysis, it was found that "industry boom in the coming year", "the financial cost ratio" and "bank credit" for whether it is easy for "default" risk factors. Elected in logistic regression, "current ratio", "the financial cost ratio" and "net interest rate before tax", "head of General credit rating", "head of experience" and "bank credit", "guaranteed capacity" and "industry boom in the coming year" for the model variables.