本文藉由貸款成數與借款人違約機率之間是否存在顯著正向關係之檢定,探討台灣不動產抵押貸款市場借貸雙方的資訊不對稱問題。實證結果發現,不論是應用Chiappori and Salanié(2000)模型或Dionne et al. (2001)模型,不論是控制利率變數或未控制利率變數,貸款成數與違約機率之間均呈現負向相關,顯示台灣不動產抵押貸款市場似乎不存在資訊不對稱問題,但由借貸人的違約機率愈高其貸款成數愈低的現象觀察,台灣不動產抵押貸市場可能存在著「另一種資訊不對稱現象」,即高風險型態的借款人較傾向於選擇低貸款成數契約的「反逆選擇」行為,或是選擇高貸款成數契約的借款人其違約風險反而降低的「反道德風險」現象。此外,實證結果亦顯示,台灣銀行不動產抵押貸市場的「另一種資訊不對稱現象」主要發生在銀行與年輕族群借款人之間;以及當利率愈高時,台灣不動產抵押貸市場的「另一種資訊不對稱現象」愈明顯,顯示利率愈高使得借款人負擔愈重時,他們愈會採取愈安全的投資或消費行為,以降低違約發生的機率。
By testing for the relationship between “loan-to-value ratio” and “borrower’s default probabilities”, this paper explores the problems of asymmetric information between banks and borrowers in Taiwan’s real estate mortgage markets. Using the extended alternative paired probit models of Chiappori and Salanié and the econometric models of Dionne et al., we first investigate whether asymmetric information exists in the markets. We further examine whether banks can use relative variables (e.g. interest rates and borrowers’ ages etc.) to control the asymmetric information problems. Chiappori and Salanié(2000) found no empirical support for the existence of asymmetric information in the liability insurance market of France. Dionne et al. (2001) showed that that there is no residual adverse selection on risk types in the automobile insurance portfolio studied. On the contrary, we find that opposite empirical evidence to support the existence of asymmetric information in Taiwan’s real estate mortgage markets. It seemly exists another pattern asymmetric information phenomenon. That is, the higher borrowers tend to select lower loan-to-value ratio contracts to reduce their default probabilities. Moreover, we also find that another pattern asymmetric information phenomenon mainly occur between banks and younger mortgagers subgroup, and banks could somehow control the asymmetric information problems by their pricing (interest rates) systems.