| With the globalization and internationalization of the world finance, risk management has become one of the most popular topics in the world of global finance in recent years. The process of China's interest rates liberalization started from last century and so far has achieved important initial results. We can foresee that in the near future, China's interest rates would be determined by the supply and demand of currency. Up till that time, commercial banks would suffer larger interest rate risks than it does now. Therefore, it is of great significance to do studies on the measurement of interest rate risk and to enhance risk management of commercial banks'interest rates through the introduction of advanced technology and tools. The significance not only lies in the current and future commercial banks'interest rate risk management, but also could be reflected by meeting the inevitable requirements of China's economic and financial development. The global financial tsunami triggered by the U.S. sub-prime crisis, is a great catastrophe for the global banking sector. The financial crisis caused a massive outbreak of credit risk, which has a lot to do with interest rate risk. Therefore, the financial crisis requires China's banking to attach great importance to risk management, in particular, real-time measurement, monitoring and management of interest rate risk.Studies on the interest rate risk of commercial banks at home and abroad mainly focused on the term structure of interest rates, duration and convexity gap, and the interest rate risk with embedded options, most of which are theoretical studies or model derivation. Few studies have touched upon both quantitative research on the interest rate risk of commercial banks and empirical research on the measurement of a single interest rate risk. Starting from the inter-bank bond repo market, using GARCH models and Risk Metrics-based mixed-normal distribution fitting method, this thesis has made empirical analysis of the Repo Rate and China's interbank treasury bonds repurchasing market, aiming to find a model suitable for the risk measurement of China's interbank treasury bonds repurchasing market and to help the sub-market of our commercial banks measure and manage the interest rate risk.The author gets a conclusion: to calculate the repo rate risk VaR with the Risk Metrics-based mixed normal distribution simulation method not only takes account of the ARCH of financial time series, but also well depicts the fat tails characteristics of time series. In addition, the mixed-normal distribution fitting method can adjust the distribution of forms based on the sample data selected at different periods, and makes sure that the distributions become even more timely and accurate. At the end of this thesis, the author proposes Chinese commercial bank risk management, and the future research directions of Risk Metrics-based distribution fitting method. One is the Distribution Fitting and risk measurement based on the high-frequency financial time series, the other is the full range of risk management models, including both theoretical and empirical research on the system. |