I have a two sector model with patient households and impatient entrepreneurs. My model works just fine when I run it with first order approximation. However, when I add the value functions of the households and entrepreneurs to evaluate the monetary policies and I take the second approximation to those functions and their autocorrelation, variance, correlations are NaN, and it takes too long to get the results about 1-hour whereas it takes 10seconds with first order approximation.
Do you have any idea about what might be causing this problem and how it can fixed?
Thanks for your help in advance.