透過您的圖書館登入
IP:18.220.187.178
  • 學位論文

台灣金融機構的危機預警應用之研究

An Application of Distress Prediction Model for Taiwan Financial Institution

指導教授 : 張麗娟
若您是本文的作者,可授權文章由華藝線上圖書館中協助推廣。

摘要


建立存活率預警模型並與常用之Logit模型區別力作比較,期能發現在預警模型實務上應用性與優越性,並對金融檢查與監理提供建議。 本文以台灣10家財務危機銀行及其配對40家正常銀行為研究對象,利用存活分析方法中之加速失敗時間模型〈AFT--accelerated failure time model〉,建立風險預警模式:偉柏模式〈Weibull model〉,對數常態模式〈Log-normal model〉及對數邏輯斯模式〈Log-logistic model〉,再利用殘差分析法來判斷此三種風險預警模式之配適度程度。 實證結果顯示,對數常態模型〈log-normal model〉之配適最好,故建議使用此模型做為預測金融機構危機之工具。

並列摘要


This study examines 10 crisis banks in Taiwan and their matching 40 normal banks, utilizing one of the survival analysis methods—AFT (Accelerated failure time) model to set up three kinds of risk precaution models (Weibull model, Log-normal model and Log-logistic model), and uses residual analysis to compare the fitness of the three risk precaution models. Our empirical results show that the fitness of the lognormal model is the best. We therefore suggest that the model be used as the distress precaution model for financial institutions. This study also establishes a survival rate precaution model and compares with the popularly used logit model for Discriminant Analysis. We hope to find the practical applicability and superiority of the precaution model, and to provide advice for financial inspection and supervision.

參考文獻


2.林淑萍,「企業危機預警模式之研究- DEA-DA、邏輯斯迴歸與類神經網路之應用」,會計與公司治理,第4卷第1期,2007年,35-56頁。
17.鄭婷月,「汽車貸款客戶之風險研究」,成功大學統計所碩士論文,2003年。
1.Altman, E., “Financial Ratios, Discriminant Analysis and the Prediction of Corporation Bankruptcy,” The Journal of Finance, vol. 13, 1968 ,pp. 589-609.
2.Altman, E., Bharath, S. and Saunders A., “Credit Ratings and the BIS capital Adequency Reform Agenda,” Journal of Banking & Finance, vol.26, 2002, pp. 909-921
3.Altman, E. and Sabato, G.,“Modeling Credit Risk for SMEs: Evidence from the US Market, ” 2005a.

延伸閱讀