EViews.com

morning,

Does anyone know how to resolve this kind of problem:

Missing values in @logl series at current coefficients at observation 1/02/2002 in "do_garch.ml(showopts,m=250,c=1e-6)"

But,there is no missing value at the very date.

thanks!

There probably are missing values in the log-likelihood series due to bad starting values or a bad likelihood specification. A very simple example of this would be taking the log of a series as part of the log-likelihood evaluation. If the series has negative values, the log-likelihood will have missing values, even though the raw data does not.