 |
 |
How to store the conditonal forecasted variance for the FULL ESTIMATION period (training and testing)? I need to extract the full conditional variance to use as a forecasted sereis then to calculate the error measures and to easily caclulate the MZ regression in another software. Once I estimate and test the model i get missing data for the less forecast(out-of-sample) period. This is normal but is there a method to extract the coditional variance from the drop down G@RCH menu tools? the full sample period is from 01/01/1998 to 12/31/2013: out-of-sample:03/16/2009 to 12/31/2013. One step a head forecast.