| dc.contributor | 金融系 | |
| dc.creator (作者) | 張春雄 | zh_TW |
| dc.creator (作者) | Chang, Chun-Shyong | |
| dc.date (日期) | 1990-06 | |
| dc.date.accessioned | 12-十月-2016 11:27:21 (UTC+8) | - |
| dc.date.available | 12-十月-2016 11:27:21 (UTC+8) | - |
| dc.date.issued (上傳時間) | 12-十月-2016 11:27:21 (UTC+8) | - |
| dc.identifier.uri (URI) | http://nccur.lib.nccu.edu.tw/handle/140.119/102700 | - |
| dc.description.abstract (摘要) | 臺灣廠商發表的財務報表，其可信度多年來一直受到社會的質疑，此研究旨在對臺灣廠商財務報表的可信度做實證分析，搜集銀行貸款廠商的財務資料，包括97家發生呆帳的廠商和200家正常的廠商，運用此等資料進行貸款風險之測定，進而分析財務報表的可信度。此研究包括兩部分，第一部分採用差異分析法，運用廠商財務資料，測定各項財務比率的差異分類能力，以導出風險預測方程式；惟前項方法不能檢定方程式中各項財務比率之統計特性。因此 ，第二部分乃採用probit分析法，用以檢定各項財務比率之統計特性。此研究導出之差異方程式和probit方程式，經反覆測定，顯示臺灣廠商財務資料可信度不高，財務報表之品質有待提昇，不過差異方程式可供做貸款風險分析之一項輔助工具。The reliability of financial statement disclosures in Taiwan is popularly believed to be low. This research represents a first step in empirically examining the reliability of Taiwanese financial statement disclosures. Financial statement disclosures are examined in connection with loan default.The existence of information content in financial statements is examined using two approaches. First, loan default prediction models are developed using accounting data and multivariate discriminant analysis. A drawback of this approach is that variables entering the discriminant equation are chosen solely on discriminatory power without regard to statistical significance. Therefore, a second approach using conditional probability models is employed. With this approach, variables are tested for statistical significance.A by-product of this research is the development of loan default models for predictive purposes. This could have significant social value in improving the efficiency of lending decisions in the Taianese economy. | |
| dc.format.extent | 4682749 bytes | - |
| dc.format.mimetype | application/pdf | - |
| dc.relation (關聯) | 國立政治大學學報, 61, 563-628 | |
| dc.title (題名) | An Empirical Test on the Use of Financial Statements for Loan Default Risk Analysis | |
| dc.type (資料類型) | article | |
