以DID計量方法分析:央行貸款成數限制對房貸逾期之影響

Abstract

我國央行有鑑於國內房貸過度擴張，一旦發生系統性風險，將不利於我國金融穩定，故於103年6月實施擴大房貸貸款成數(LTV)管制措施。本研究取得國內房貸龍頭銀行101~105年樣本數達25,000筆，資料擴及全國各區域。為探討我國央行貸款成數限制，即對於借款人之還款約束力控制，是否可有效降低房貸逾期情形。本次採取差異中的差異(difference in difference)計量方法，以103年6月央行政策實施時點，並以管制貸款成數新增區域之桃園市的桃園區、蘆竹區、中壢區、龜山區為實驗組，全台其餘區域為對照組(不含新北)。從本文實證結果可知幾點:第一，全體樣本於政策實施後，對於逾期比率有顯著下降。第二，政策實施後對於前後一年及前後兩年之樣本，逾期比率亦皆有顯著下降，前後一年之效果大於前後兩年。第三，政策實施對於投資族群及自住族群皆有顯著下降效果，其中自住族群的逾期比率下降效果高於投資族群。

In view of the excessive expansion of domestic housing loans, once a systematic risk occurs, it will be detrimental to the financial stability of our country. Therefore, in June 2014, the Central Bank implemented the control measures to lower the threshold loan-to-value (LTV) ratio. This article has obtained 25,000 samples of leading domestic mortgage banks, and the data are completed and extended to all regions of the country, in order to verify whether the control of the LTV can effectively reduce the residential mortgage default. The difference in difference measurement method is adopted this time. This policy was implemented in June 2014, and the four restricted areas of Taoyuan, the newly-added of controlled LTV, were used as the experimental group. The others of Taiwan excluded the New Taipei City were used as the control group. First of all, after the implementation of the policy, the non-performing loans ratio of the entire sample has decreased significantly. Secondly, the non-performing loans ratio in a year or two, before and after the implementation of the policy has also decreased significantly, the effect in a year is greater than that in two years. Thirdly, the implementation of the policy has a significant reduction effect on both the investment group and the first purchase group. Between them, the reduction effect of the non-performing loans ratio of the first purchase group is higher than that of the investment group.