Credit score of Taiwans＇ElectronicIndustry Base on DEA approach

• Main Authors: Wang ,Wei-Chyan, 王威乾
• Other Authors: Du, Jian-Heng
• Format: Others
• Language: zh-TW
• Published: 2014
• Online Access: http://ndltd.ncl.edu.tw/handle/z9vejh

碩士 === 國立高雄應用科技大學 === 金融資訊研究所 === 102 === Traditional crisis prediction is mostly multiple discriminant analysis, Logistic regression analysis and neural networks, etc., to construct the prediction model, but most of these predictive models needed to derive the message afterwards, it is difficult to achieve in advance forecast demand. In this study, the DEA-BCC model by early enterprise financial ratios for credit score to reach crisis construct predictive models. Select the study sample to 101 years in Taiwan listed on the OTC and emerging companies in the electronics industry, and the variable selected 60 financial ratios TEJ information as the basis of the announcement, and in accordance with the Companies Act, securities exchanges, Taiwan economic Journal (TEJ) and domestic scholars argument, the 923 electronics companies divided into 69 crisis companies and 854 normal company, through factor analysis extracted, CAP model potency testing, data envelopment analysis (DEA) of the BCC model research methods, linear regression analysis, discriminant analysis and classification to make the credit score of the enterprise, the empirical results show that dependence on borrowings elected under this Institute, debt ratio, cash flow per share, the net per share (TSE), total asset growth rate and operating profit / paid-up capital ratio by six financial ratios as credit score, for the year when determining the correct classification rate achieve 90.9%, then another 102 years for the six financial classify discrimination, the correct rate also up to 85.9%, thus proving financial ratios elected under this Institute through DEA-BCC model is constructed of credit score, should be reached prior demand forecasting model prediction the crisis.