標題: | 中小企業違約信用風險評估流程 Constructing a Procedure to Evaluate Credit Risks Due to Defaults for Small and Medium-sized Enterprises |
作者: | 吳莉安 Li-An Wu 唐麗英 張永佳 Lee-Ing Tong Yung-Chia Chang 工業工程與管理學系 |
關鍵字: | 中小企業;償還率;羅吉斯迴歸;新巴塞爾資本協定;small and medium-sized enterprise;recovery rate;Logistic regression;Basel II |
公開日期: | 2005 |
摘要: | 近年來經濟不景氣與金融變遷快速,使得國內金融機構之逾期放款比率逐年高昇及其借貸客戶之償還比率逐年降低,造成金融機構嚴重的損失。此外為落實國內金融業風險管理理念,政府要求國內銀行在2006年底實施符合國際清算銀行標準之新版巴塞爾資本協定(the New Basel Capital Accord, 簡稱Basel II),以增加銀行業者的風險敏感性。因而銀行及金融機構必須針對借款之企業制定一套有效的違約信用風險評估流程,以做為銀行制訂授信放款策略及客戶違約催討策略之依據,以降低信用風險並減少損失。
目前中、外文獻相關之研究多以上市上櫃公司為研究對象,這與國內金融機構以中小企業為主要放款對象有所出入,造成其所建議之模型無法直接應用於國內的金融機構。因此,本研究針對國內金融機構之中小企業借款者,發展出一套違約信用風險評估流程。此流程主要衡量企業之違約信用風險等級及建構償還率預測模型,主要分為四階段:(1) 選擇變數與蒐集整理資料,本研究共蒐集了企業內部財務、客戶信用評估項目、客戶評等評分及總體經濟指標等四大構面變數進行分析;(2) 對自變數進行主成份分析(Principle Component Analysis),以縮減變數個數並消除變數間之共線性; (3)利用羅吉斯迴歸(Logistic regression)構建違約信用風險等級;(4)利用羅吉斯迴歸構建償還率預測模型以有效幫助金融機構決定最適宜之催討決策。本研究蒐集了國內某金融機構所提供之中小企業客戶歷史資料進行模型建構,驗證本研究模型之可行性。 Economic recession and unstable financial markets have resulted in some companies not being able to make on-time payment on their mortgage loans. This in turn has resulted in high overdue loan and low recovery rates to banks and financial institutions where consequently result in huge financial losses to these loan agencies. To meet the requirements from the New Basel Capital Accord (Basel II) and to minimize the credit risks due to incorrect loan decisions, it is very important for banks and financial institutions to construct a reliable procedure to evaluate credit risks resulting from defaults. Most of the studies on loss due to defaults (or recovery rate) are based on the data drawn from publicly traded companies. However, there is some difficulty to apply these results directly to those loan agencies whose customers are mainly small and medium-sized enterprises (SMEs). This research, focused on small and medium enterprises specifically, aims to construct an effective and reasonable procedure to assist loan agencies to evaluate credit risks due to default. The procedure consists of four stages: (1) selecting relevant variables and collecting data, (2) using Principle Component Analysis to reduce the number of variables, (3) applying Logistic regression to construct credit risk models, (4) constructing recovery rate prediction models to assist financial institutions making loan recovery policy. Finally, a case study is provided to demonstrate the effectiveness of the proposed procedure. |
URI: | http://140.113.39.130/cdrfb3/record/nctu/#GT009333504 http://hdl.handle.net/11536/79464 |
Appears in Collections: | Thesis |