Study Of Treasury Rate Risk Based On VaR Model

Government bond’s yields reflect the risk-free interest rate. In the micro field, it is the foundation of asset pricing, financial products design, hedging and risk management, arbitraging and investment. In the macro field, the issuance of government bonds is an important means of macro-control and manipulating the level of interest rates. Term structure of interest rate, which is also called the yield curve, plots a set of yield to maturity of the zero-coupon bonds with different maturities. In Chinese in the background of gradually accelerating interest rate liberalization, study of fixed rate government bonds yield curve has practical significance.Interest rate risk is refers to the uncertainty of fund changes due to fluctuations in interest rates of the financial institutions. Interest rate risk is mainly divided into the repricing risk, basis risk, yield curve risk and embedded option risk. The measurement of interest rate risk is divided into the traditional gap analysis, duration gap method and simulation method. How to understand, cognition and control the interest rate risk is a key content of fixed income securities. Since2004our country interest rate fluctuated frequently, the traditional sensitive gap analysis can only deal with the situation of small interest rate fluctuation amplitude and frequency, the requirement of the development of interest rate risk management methods objectively. Since JP Morgan in1995put forward the concept of value at risk, value at risk has become an important tool for all types of financial institutions risk management. This paper attempts to introduce the VaR method to measure the interest rate risk, empirical data selecting the fixed interest rate government bond’s maturing rate of inter-bank bond market. Because of the investment portfolio of fixed income securities market variables for the term structure of interest rates, before risk value analysis we analyze the term structure of interest rates. The term structure of interest rate cointegration test found3of the existence of cointegration relationship between term structure of interest rates. The cointegration relationship exists between the sample sequence, so the application of principal component analysis to analyze the relationship between the principal components. The empirical results show that the principal component analysis of the cumulative variance of principal component can explain88%of the first three, visible in the first three principal components could explain the vast majority of term structure of interest rate change. Usually we call the first three principal components as the level factor, slope factor and curvature factor. Leve factor explains the term structure of interest rate change reached70%, slope factor and curvature factor explanation ability respectively is13.24%,6.33%.This paper respectively using historical simulation, variance covariance approach, Monte Carlo simulation and scene simulation method to estimate the remaining term of1year fixed rate bonds VaR value, the application of Kupiec inspection found99%confidence level only Monte Cark) simulation method can through inspection.So we choose use the monte carlo simulation to estimate VaR value of the95%confidence level of other key term. The empirical results show that the residual period for3to5years of fixed rate bonds interest rate risk is the largest, the main reason is that short-term bonds. Medium-term bond investors most for securities companies, fund companies, etc., focus on the purpose of speculation. Remaining term is greater than5years of fixed rate bonds interest rate risk increases gradually, the maturity is more than20years of long-term Treasury bonds for the duration of VaR value and surplus of3to5years of fixed rate bonds VaR value was flat, this is because long-term debt holders mainly for the commercial Banks, the asset allocation is the purpose of holding long-term debt, focus on long term operation, more attention to the security of the bonds, rather than a liquid. Banks will usually long-term Treasury bonds to maturity, not circulated to make these bonds are no longer to enter the market circulation of national debt, lack of market-oriented pricing long-term debt risk level is low.In this paper, the innovation mainly in two aspects:first is the study of the VaR model of national debt interest rate risk management. VaR in our country commercial bank interest rate risk management in the still at the fledgling stage. Adopted the historical simulation method. the variance-covariance method. the monte carlo simulation method simulation method, and scenario simulation method. Unlike previous sample test, the results of VaR taken outside the sample detection. Second is introduced in detail the scenario simulation method proposed by Jamshidian and Zhu (1997) to calculate the risk value, the method is first used principal component analysis method to reduce the number of risk factors, the market on the basis of a polynomial distribution factor to obey the multivariate normal distribution of discretization, generate a finite number of scenarios with different probabilities, thereby greatly simplifies the calculation.