我也正被此问题所困扰啊

ln(p)= ln(p(-1))+residum

then check residum if r=f(r(t-1)), it is a MA PROCESS

If r^r= F(r(t-1)^r(t-1)),then check if there is ARCH effect , or only Hetroscedacity, when ARCH(1) exists, then test ARCH(2), when there are higher ARCH, you should use GARCH(1,1) or more higher order

when no ARCH, just only white hetroscedacity, do error correction in estimation equation(OLS), select (option) , it is near specification, select white hetroscedasity, the cofficient will be pratically not changed, just the t -value will be reduced, because the standard variation will be enlarged.

I suggest you would better read some user guide, and follow the operation steps, and do many comparison, then you can understand better

感谢两位的指点! 觉得自己真的欠看书欠太多了.

我找不到make Variance选项，哪位知道怎么得出波动率？

着急，谢！