金融機構在進行房屋抵押貸款授信時,需對貸款人進行提前清償行為的風險加以評估,以避免產生資金流量與利息收益不確定之風險。國內有關提前清償之相關研究,在實證方法上大都利用羅吉斯迴歸來分析,以估計貸款人提前清償之機率。故本文以羅吉斯模型(Logistic regression model) 來探討貸款人提前清償之行為。 在實證時係利用羅吉斯模型分析影響貸款人較易提前清償之因素,進而估計出提前清償之機率。研究結果發現在個人借款條件中,借款金額、償還期間、償還方式、攤還金額、利率型態與提前清償呈正相關,其中以償還方式最顯著,進一步實證得知採用先付息後本息平均攤還時,最容易提前清償。
The early payoff behavior for loan borrowers has become an important studying topic for financing institutes in the present days. Such issue must be carefully considered during in the evaluation of accepting new loan and mortgage requests to avoid risks brought by the uncertainties of cash flow and interest earnings. Most of the previous studies concerning this topic have adopted Logistic Regression Model to predict the rate of early payoff for loan borrowers. This paper has also applied the same model for the early payoff behavior research. This research has analyzed the main factor concerning early payoff behavior and evaluated the rate of early payoff for loan borrowers to reach the following conclusion. The empirical study has revealed a positive correlation between early payoff and total borrowed amount, payback period, payback method, payback amount and interest rate. Payback method has showed the most significant influence to the early payoff decision. This paper has further found the pay-interest-first method has the highest early payoff rate.