开发者:上海品职教育科技有限公司 隐私政策详情

应用版本:4.2.11(IOS)|3.2.5(安卓)APP下载

🍑🍑🍑🍑🍑🍑🍑 · 2019年11月09日

问一道题:NO.PZ2016062402000045

问题如下:

Suppose σt2\sigma_t^2 is the estimated variance at time t and μt\mu_t is the realized return at t. Which of the following GARCH(1,1) models will take the longest time to revert to its mean?

选项:

A.

σt2=0.04+0.02μt12+0.92σt12\sigma_t^2=0.04+0.02\mu_{t-1}^2+0.92\sigma_{t-1}^2

B.

σt2=0.02+0.04μt12+0.94σt12\sigma_t^2=0.02+0.04\mu_{t-1}^2+0.94\sigma_{t-1}^2

C.

σt2=0.03+0.02μt12+0.95σt12\sigma_t^2=0.03+0.02\mu_{t-1}^2+0.95\sigma_{t-1}^2

D.

σt2=0.03+0.03μt12+0.93σt12\sigma_t^2=0.03+0.03\mu_{t-1}^2+0.93\sigma_{t-1}^2

解释:

The persistence (α1+β)(\alpha_1+\beta)span> is,respectively0.94, 0.98, 0.97, and 0.96. Hence the model with the highest persistence will take the longest time to revert to the mean.

the he persistence ()span> is, 这什么意思?

1 个答案

品职答疑小助手雍 · 2019年11月09日

同学你好,garch模型里后两个项也就是α和β是代表和当前相关的系数,和当前相关的系数越高,回归到均值的速度就越慢,所以就找这俩的合最大的模型就好了,0.94+0.04那个最大。