FilteredVariables vs. Actual data

Hello everyone,

I am working on a DSGE Model and want to test model fit by comparing oo_.FilteredVariables and actual data. The attached excel is the comparison. Could you please give me some advice about how to improve or what’s the problem in model may cause this kind of output? I’ve tried many but still no improvement. Appreciated in advance!

BTW, if the one-step forecast match actual data very well, can we say the model is reliable then it can be used to do forecasting? THANKS.

Helen
FiltererVariables_140218.xls (146 KB)