Stress Testing On Credit Risk Of Real Estate Loan In Commercial Bank Based On CPV Model

Chinese banks are closely related to real estate market. Correspondingly, the realestate market in turn affects the bank asset quality and credit risk provisions. In recentyears, with the continuous reinforce in real estate market regulation, the housingbubble could burst at any time. Conventional risk management tools have been unableto meet the risk management needs of financial institutions. As a tool of measuringcredit risk, stress testing has been regarded as important as Var. The World Bank andIMF jointly launched FSAP program in1999, and took stress testing as an importanttool to measure the impact from “extreme but credible” tail events. China officiallylaunched the FSAP program in2009and has already organized several overall-testsand sub-tests among our bank system. In this background, this paper attempts to usestress testing tool to analyze the credit risk of segments among commercial bankloans-real estate loan.This paper adopts theoretical research and empirical analysis at the same time.Firstly we systematically elaborate the relevant literature on tress test and thenanalyze the credit risk of commercial bank’s real estate loans from different angles.On this basis, we select Credit Portfolio View model and a dozen banks’non-performing loans ratio as a sample data source to launch the credit risk empiricalanalysis. Furthermore, by selecting related macroeconomic variables and conductingco-integration analysis, we find that only GDP growth rates、housing price growthrates and long-term loan interest rates indicate significantly. In addition, GDP ratesand GFJ are negatively correlated with NPL ration while DKL is positively correlated.Then according to the fitting result, we design pressure scenarios to forecastnon-performing ratio and credit losses. Finally, relevant policy recommendations areproposed according to the stress test results.