Welfare and volatilty

Dear Users,
I have a question related to Welfare evaluation. I am computing welfare as

W = U + beta*W(+1)

and I am solving the model with second order approximation. What I have noticed is that the theoretical mean of Welfare (stored in oo_.) is increasing in the variance of the shocks in the model. Is this due to precausionary savings or is there something else which is going on? (agents are risk-averse of course)

Thank you

Depends on your model. But precautionary savings may be one reason as a higher capital stock may lead to higher average capital.

Thanks for your reply.
I am studying Iacoviello and Neri (2010) “Housing Market Spillovers: Evidence from an Estimated DSGE Model”.
This issue is a big deal to study optimal policy though…