GARCH(1,1)模型的M估计

M-estimators of GARCH(1,1) model

常用的参数估计方法有最小二乘法,但最小二乘估计存在不稳健性.为了找出更稳健的估计方法,提出了M估计的定义及其算法,然后对QMLE和LAD估计做了模拟比较,并对道琼斯指数做了实证分析.

We find out a stable estimation method, propose the definition of M-estimator and its algorithm for LSE is not as stable as the most common method of the parameters＇ estimation. Then QMLE and LAD were compared by simulation, and LAD estimator was used to simulate the Dow lones index.