應用邏輯斯迴歸與存活分析建構台灣中小企業貸款提前清償預測模型

Abstract

台灣中小企業在發展時通常需要大量的資金，其中大部分的資金是向銀行或金融機構以授信方式取得。近年來向銀行或金融機構貸款之中小企業提前清償的情形日漸增多，使得銀行或金融機構在財務管理上碰到許多問題。過去中、外文獻所提出之提前清償預測模型多是用在租賃業，尚未見利用於企業貸款方面。因此，本研究之主要目的是針對以中小企業為主要放款客戶之金融機構或銀行，建構一套有效之提前清償預測模型，以準確地判斷企業是否會提前清償貸款以及發生提前清償行為的時間點，使銀行或金融機構之財務規劃決策者能事先規劃投資策略，以減少財務投資之損失。本研究首先利用邏輯斯迴歸建立一個中小企業貸款者之提前清償模型，再利用存活分析法建立一個存活機率表，以預測企業可能發生提前清償行為之時間點。最後，本研利用台灣某金融機構之中小企業貸款案例，說明本研究方法確實有效可行。

The development of small and medium enterprises (SMEs) in Taiwan is crucial without sufficient financial support. Most of funds for development are borrowed from banks or financial institutions. Recently, banks or financial institutions encounter many problems about SMEs’ prepayment. According to previous studies, the risk assessment model of prepayment is mainly utilized in mortgage loans. It is rarely seen that the prepayment model is employed in enterprises’ loans. In this study, the logistic regression is utilized to construct a prepayment model of SMEs’ for banks or financial institutions. Then, the survival analysis is employed to establish the lifetime table for predicting the time point that the prepayment may occur for each SME. Using the proposed procedure, preventive actions can be taken by banks or financial institutions’ policymakers to reduce the financial losses due to prepayment of SMEs. The effectiveness of the proposed method has been examined with a real case of SME loaners in Taiwan.