VaR Method Based On ARCH Models For Managing Interest Rate Risk In Commercial Banks

In1996, our country began the reform of interest rate liberalization for-mally, the reform process was promoted steadily. Many interest rates were liberalized, includes interbank offered rate, bond market interest rates, the issuing interest rate of interbank market bond and policy financial bonds, for-eign currency loan interest rates for domestic Chinese ect. After liberalization, interest rates are expected to have an increased susceptibility to financial en-vironment, with clearly increased violation frequency and range. Most of the assets of commercial banks were financial assets, whose value will change while the interest rate violate. So interest rate risk is having a greater influence on commercial bank. Howerer, the most significant risk management in our com-mercial banks was the credit risk management before, as a result, the interest rate risk management is inexperienced. Researching the interest rate risk man-agement is crucial to our commercial banks that give positive action to interest rate risk and establish complete and effective systems to handle risks.This paper has five chapters. The first chapter mainly introduces the re-search background and significance, the foreign and domestic literatures about interest rate risk management in commercial banks, the research idea and framework of the issue. The second chapter introduces causes of interest rate risk in commercial banks from both external macroscopic and internal micro-scopic aspects; introduces Interest rate risk management methods in commer-cial bank including sensitive gap analysis, duration gap analysis, convexity gap analysis, especially VaR analysis and its principle, computational method, mer-its and drawbacks. The third chapter begins with volatility of financial series, introduces ARCH model, GARCH model, GARCH-M model, EARCH mod-el and IGARCH model, analyses their model features, application condition, merits and drawbacks. Furthermore, because of limitations of normal distri-bution in fat-tailed features, this chapter introduces t and GED distributions. The fourth chapter chooses data of Shanghai Interbank Offered Rate(O N) as sample, analyzes the stationarity, normality and fat-tailed features, autocor-relation, ARCH-effect. The fifth chapter build AR(1)-GARCH(1,1), AR(1)-GARCH-M(1,1), AR(1)-EGARCH(1,1), AR(1)-IGARCH(1,1) models under normal, t, GED distributions; then chooses models under GED distribution as the best ones by maximum likelihood function value and AIC SC value; fi-nally calculates VaR at both95-percent and99-percent confidence levels with conclusions following.As conclusions show, models under GED distributions are best ones to fit data’s fat-tailed feature. Furthermore, model estimation result shows O N In-terbank offered rate has reverse leverage effect and long memory effect. At the95-percent confidence level, only VaR calculated under AR(1)-EGARCH (1,1)-GED passes the Kupiec test, which proves EGARCH model that indicates asymmetric volatility can be used in calculating VaR effectively. At the99-percent confidence level, except of AR(1)-IGARCH(1,1) model, VaR under other models all pass the Kupiec test, which proves good risk-management-effect.