EViews.com

Dear Eviews-users,

I hope someone can help me out.

Attached is the data containing daily returns for the Deutsche Mark / US dollar exchange rate, over the period January 6, 1987 - June 30, 1999 (T=3036 observations).

I have estimated a GARCH(1,1) model for the daily exchange rate returns for the period January 6, 1987 - December 29, 1994 (apprx. 1868 observations).

C -0.016837 0.015712

Variance Equation

C 0.017960 0.005520 3.253806 0.0011
RESID(-1)^2 0.048925 0.008696 5.626234 0.0000
GARCH(-1) 0.916621 0.016644 55.07274 0.0000

And made a one-step ahead forecasts for the variance for the period January 3, 1995 - June 30, 1999 (using forecast --> forecast name rf / garch (optional) rf --> statistical --> sample 1868 - 3036).
The conditional variances can be used to make a volatility forecasts by regressing the squared daily return of the conditional variance forecast.
Estimate equation --> LS --> equation specification rsquared rf c and the complete sample. But the answer to the question isn't correct.

Values should be b0=-0.075(0.075) and b1=1.035(0.164) with an R2 = 0.036.

Can someone tell me what i am doing wrong?

Thanks in advance!

I am not sure whether it is a typo, but you should assign a different name to save the garch forecasts (i.e. not rf). As far as I understand, you are expected to use the conditional variance forecast as a regressor to volatility, which is defined as the squared daily returns. Make sure that you choose "static forecast" and set the sample to "1869 3036" in the garch estimation.