EViews.com

Hell all,

I was trying to test an ARCH effect for an ARIMA model using Correlogram Squared Residuals, but I found that there was no p-value in front of 1-4 lags,why? I am very worried about it.

The DoF for the p-values are corrected to include the ARMA terms in the original estimation. This means that the p-values for the first 4 lags (if you have 4 ARMA terms) cannot be calculated. It is nothing to worry about.