Individual Housing Mortgage Default Risk Empirical Research

Since the Construct Bank of China lent the first mortgage loan in 1992, it has become the major financing way by which people realize their dream of having a house. The quantities of mortgage loan increase at a speed of 40-50% per year nowadays. Although mortgage loans seem to be less risky comparing with other commercial loans in the balance sheet of banks, but they usually have long tem maturities and expose their risky features gradually during their running time. The research in default risk has been very sound in the developed countries as they have a long history in the carrying on mortgage business. Default risk is being concerned more and more in both financial industry and the academic field, and so is the necessity and urgency of reinforcing the management of default risk. However, there still lacks of empirical studies in default risk of residential mortgages. Because of the background mentioned above, this dissertation focuses on the empirical analysis about the major determinants of default risk. This study will help banks build theoretic and technological foundation to control default risk at the entering of residential mortgages.This empirical study uses actual mortgage loan data of a commercial bank in Sichuan province which include 65536 samples. After the abandoning those incomplete records, 35031 effective samples left, in which 29514 samples is of normal status and 5517 is of default status. Based on the literature review this study classifies the variables in four dimensions: borrower characteristics dimension, loan characteristics dimension, property characteristics dimension and regional characteristics dimension. 19 variables such as borrower's age, loan-to-value etc enter these four dimensions. Using the statistics software SPSS10.0, this study conducts descriptive statistics, factor analysis, discriminant analysis and logistic regression. The empirical results indicate that loan-to-value and...