Study On The Default Risk Management Concerning Residential Mortgage Loan

From the perspective of developing process of China’s real estate market, the reform on urban residential housing system, which was implemented by the government in1998, has served as a watershed in the development of real estate market. Having taken the opportunity, the real estate market in China has stepped into a rapid developing track. During this process, real estate financing has been playing an important role, and the residential mortgage loan has be witnessed the fold increases in the recent dozen of years. Residential mortgage loan becomes one of the important credit species in commercial banks. But as a long term loan, it also carries the considerable credit risks. To strengthen the study on the residential mortgage default risks is of great importance for the commercial banks in regard to increase the risk recognition level, to strengthen the risk prevention and control capacity, to enhance the core competitiveness.From the domestic and oversea study, the foreign scholars have proposed and applied various economic theories, and set up the various econometric models, and made the multi-angle researches on the residential mortgage default risks with application of the multi-dimensional variables. The domestic experts have also made the exploration in the terms of qualitative analysis and mathematical model, while there are much fewer empirical studies using the econometric models to analyze the various kinds of default risks respectively.The essay takes the default risks concerning residential mortgage loan as the study object, analyzing and reviewing the current overseas and domestic theories and research achievements meanwhile investigating the practice of default risks concerning residential mortgage loan; thus proposing the concrete research prospective and the basic arguments. On the basis of reference on the related research at home and abroad as well as the management practice on default risks, combining the current situation in China, twenty variables and quantitative methods have been determined for the empirical study of the thesis, thus quantitative study model being formulated. The samples adopted in the empirical analyses have been sifted and chosen from16308residential mortgage loans, which were issued by the Nanjing branch of a state-owned bank. Among these samples, there are1340normal samples.4040prepayment samples,2676delinquency samples and27substantive default samples. In the whole context, the main research achievements of the thesis present on the following aspects: 1. To make innovation on the research methods, trying hard to use the actual loan data for multi-layer empirical research including factor analysis, discriminant analysis, cluster analysis and logistic model regression analysis, etc.2. The factor analysis and discriminant analysis serve the basis to analyze and judge the impact orientation and degree of the every variable on the default risks.(1) In the empirical analysis on prepayment risk, the factor analysis produces seven factor variables; according to discriminant analysis, six out of the seven hypothetical propositions have been verified. What influence the prepayment risks considerably are education/occupation factor, absolute financial situation factor and loan situation factor. The cluster analysis classifies the prepayment samples into three groups based on risks levels. The further logistic model regression analysis shows, under the95%confidence level, education/occupation factor, loan situation factor and gender factor would be of the greater importance to the group "high prepayment risk" meanwhile, education/occupation factor, financial burden factor, absolute financial situation factor and loan situation factor will exert more influence on the group "median payment risk", then to the group "low prepayment risk", the more influential factor variables are absolute financial situation factors and education/occupation factor.(2) In the empirical analysis on delinquency risk, the factor analysis produces eight factor variables; according to discriminant analysis, three out of the five hypothetical propositions have been verified. Concerning the delinquency risks, the more influential factor variables are loan situation factors, education/occupation factor and absolute financial situation factor. The cluster analysis classifies the delinquency samples into three groups based on risks levels too. The further logistic model regression analysis shows, under the95%confidence level, education/occupation factor and absolute financial situation factor would be of the greater importance to the group "high delinquency risk"; meanwhile, the more influential factor variables for the group "median delinquency risk" are loan situation factor, residential factor, absolute financial situation factor, education/occupation factor and gender factor. What influence more on the group "low delinquency risk" are residential factor and absolute financial situation factor.(3) In the empirical analysis on substantive default risk, the factor analysis produces seven factor variables; one out of the four hypothetical propositions has been verified completely, while other two have been verified partially. The factor variables which impact more on the substantive default risk are financial expectation factor, marriage/age factor and education/occupation factor. The logistic model regression analysis on all the substantive default samples shows that all of seven variables are indistinctive. So this paper does not continue to make cluster analysis and the further logistic model regression analysis.3. According to the analysis of discriminant coefficient and component matrix, combining the features of samples, this thesis, from the perspectives concerning the features of borrowers, the features of real estate, the features of loans and the features of locations, concludes the impact of every variable on prepayment risk and delinquency risk. For instance, the education level of the borrower will have a positive impact on the prepayment. The higher the borrower’s education level would be, the more possible the loan would be paid in advance. However, there is a negative correlation between education level and probability of default risk. The lower the borrower’s education level would be, the less his income and level of stability, thus the more possible the loan would become delinquent. This conclusion also indicates that the innovation of this thesis is theoretically acceptable, i.e. the delinquency risk and prepayment risk should be investigated separately.4. Concerning the countermeasure and proposal, based on the conclusions of normative research, practical research and empirical research, this paper aims to construct risk prevention and control system against residential mortgage loan defaults innovatively. This system that features as "four in one" combines government, commercial banks, borrowers and intermediaries.From the macro-regulation aspect, the government should adopt the following measures:(1) to handle properly the fluctuation and stability in macro-regulation of real estate;(2) to coordinate and settle the interest relations among the central government, local government, real estate developer and lending institutes;(3) to serve as a dominant guidance to the housing supply;(4) to enrich the regulation measures and methods;(5) to broaden the investment channels of the residents. On the aspect of constructing a sound credit and legal environment, first of all, the credit regulation and laws that apply to all of the society should be formulated as soon as possible. Secondly, the unified individual personal credit information system should be constantly improved and completed. Thirdly, the government should launch promptly the complementary laws and regulations to guarantee the safety of the individual residential loans issued by commercial banks.From the aspect of risk assessment, risk prevention, risk monitoring and risk management,(1) the commercial bank should complete management system and business process related to mortgage loan;(2) to strengthen the research and analysis, continually upgrade the methods of inner investigation and inspection;(3) strictly inspect the related scope and standard, strengthen the ablity of inspection and approval;(4) strengthen the active risk management on delinquency through contract stipulation;(5) to establish the risk monitoring and early warning system for mortgage loan;(6) attach great importance to management after loan and strengthen group construction of the professional talents. The borrower should provide the sound record on the individual credit information and credit grade, and be honest in the whole process of mortgage loan. The agent institutes should provide the professional and standardized service in the fields of laws, insurance, guaranty, assessment, risk diversification and business innovation.5. Take a review on the research of the essay, the paper still has the relative insufficiency in sample, variable, and accuracy of classification in discriminant analysis as well as conclusion test. In the future, the further research could be carried out on the residential mortgage default risk from the following aspects:(1) to use cross-regional sample;(2) to investigate the samples with larger time-span;(3) to make the empirical research combining the option theory, optimaization theory of consumer choice, etc. with econometric models;(4) to apply the larger quantities of substantive default samples in research;(5) to choose samples of prepayment, delinquency, and substantive default in approximately identical number, and to make overall analysis of the three risks and the econometric model analysis.