基于动态随机优化方法的商业银行利率风险最优控制

Optimal Control of Interest Rate Risk in Commercial Banks with Dynamic Stochastic Optimization Method

利用Zimmerman的随机优化方法,构造一个多目标优化模型,通过合理确定其中决策变量的值,使商业银行在决策期末满足最小化利率风险的条件下收益最大化。实例计算表明,相对于利率风险静态久期免疫模型而言,动态随机久期免疫模型对于给定的初始值和约束条件,可以较好地减少利率风险暴露头寸和提高收益;同时,利率风险较大时,后者比前者更好地降低利率风险暴露,鲁棒性更强。

Traditional static interest-rate-risk-management models are only suitable for immunization from a certain change of interest rate,otherwise,we can be confronted with Stochastic Process Risk;and these models require that all of interest rate term structure be completely interrelated.Under interest rate marketization,these defects are more obvious.This paper constructs a target mapping model with Zimmerman＇ stochastic optimization method,which makes commercial banks at the end term of decision satisfy interest rate risk minimized and yields maximized,after reasonably determining the value of decision variables.Calculation case shows that in the given initial value and restraints,dynamic stochastic duration immunization model can be better to lessen the exposure position of interest rate risk and increase yields than static interest-rate-risk-management model.Meanwhile,when facing with higher interest rate risk,the former can be better to mitigate risk exposure than the later,and then it is more robust.