Longlasting IRFs

Stochastic simulations
1

Dear all,
I have a question regarding impulse response functions. They do converge but after 50+ periods. What period can be considered normal? 10?

I investigated the earlier posts and I found that problem may be related to eigenvalues close to 1. I used the check comment and one eigenvalue is indeed 0.9894.

My question is why this problem arises? Is the model set up wrong? Or is it the parameters? The problem started after introducing the nontradable goods sector to a perfectly working model. Can this be the reason? If so why?

Thanks very much…

2

That is hard to tell. Most of the times, we actually want out model to feature persistence, i.e. endogenous propagation. In a quarterly model, 50 periods is a bit more then 10 years. That can still be considered a business cycle frequency.