依貸款者觀點探討高屏地區還款選擇之影響因素－壽險公司之實證研究

Abstract

近年來，壽險業在中長期放款市場上扮演著日益重要的資金中介角色，其中不動產擔保放款與壽險貸款為重要業務來源。由於台灣在經濟成長的帶動之下，使得房地產貸款案件大量增加，導致放款機構的核貸標準變得較為寬鬆。有鑒於美國次級房貸風暴所帶來的衝擊，放款者應重視不動產放款的違約風險。特別是餘裕資金相對充裕的高雄市（包含原高雄縣），貸款者的提前清償行為對放款機構資金之流動性亦甚為重要。因此，本文以國內某大型壽險公司高屏地區之放款資料為研究樣本，目的為分析影響貸款違約與提前清償之決定因素，並利用巢狀羅吉特模型進行實證分析。檢定結果顯示消費行為觀點所建構的貸款者模型，能同時符合替選方案間獨立無關之假設，以及效用極大化型態的巢狀羅吉特模型之理論假設。然而，放款者觀點所建構之模型則無法滿足理論假設之限制。此外，個人特徵、貸款條件以及住宅特徵所建立的各項假說皆受到部分支持。詳言之，非已婚者、貸款成數愈高、年繳保費愈少、屋齡愈長以及抵押品座落區域為屏東縣之貸款者，違約發生的機率愈高。此外，已婚者、年齡愈低、貸款年限愈短、貸款成數愈低以及抵押品座落區域為高雄市之貸款者，傾向採取提前清償貸款。

Recently, life insurance plays an increasingly important role in funding intermediaries according to medium- and long- term loans. In particular, the real estate mortgage loan and loan to policyholders greatly reflects the proportion of total mortgage loan in life insurance industry. The cases of the real estate mortgage loan substantially increase with the strong economic growth, which cause the loan approval standard to be relaxed according to lending institutions. Whereas the subprime mortgage crisis poses a growing default risk to the future loans of financial institutions. In addition, the prepayment behavior of borrowers is a very important issue to the liquidity of lending institutions. Especially for borrowers who live in Kaohsiung city (and Kaohsiung county) because they have sufficient surplus funds. Therefore, the major goal of this study is to analyze determinants of default and prepayment which affect profitability of loan activities. We use the borrowers' data of the life insurance company in Kaohsiung city, county and Pingtung county, applying the nested logit model to perform empirical analyses. Simultaneously, the test results show that the borrower's modeling according to consumer behavior is in accordance with independence from irrelevant alternatives, and is in line with theoretical hypotheses of the utility maximization nested logit model. However, the restrictions concerning above theoretical hypotheses do not apply in the lender's modeling. Furthermore, the hypotheses are partially supported which are separately constructed in borrow characteristics, terms of loan and housing characteristics. Specifically, non-married borrowers, greater loan-to-value ratio, lower annual premium, greater house age and the collateral located in Pingtung county are associated with higher probability of default. The other finding suggests that married and younger borrowers, shorter loan period, lower loan-to-value ratio and their collaterals located in Kaohsiung city are associated with higher probability of prepayment.