Is it possible to compute optimal simple rules (OSR) using a planner discount factor different from 1? This is an option available for the Ramsey and the discretionary commands, but not for the OSR. I have been trying to compute OSR using a planner discount factor of 0.99, but when I introduce the option “osr(planner_discount = 0.99)” there is an error. I cannot find one example for OSR where the planner discount factor is different from 1.
Possibly this option is not available yet for OSR!? Any help?
I don’t understand your point. If you look at the manual at dynare.org/manual/index_29.html, osr minimizes the contemporaneous variance. There are not future periods involved and discounting takes place.
My problem is that I want to compare the welfare stabilization costs from optimal simple rules with the optimal solution (ramsey_policy). In the Ramsey policy case, with a loss function as the planner objective, the welfare that I obtain is a discounted sum of the future squared deviations from the mean (I’m using a planner discount factor equal to 0.99). In the optimal simple rule case, the objective function is the expectancy of the squared deviations. These are two different sums and I don’t see how to compare the two. Maybe I could consider the Ramsey planner objective and compute an equivalent welfare for the simple rule case using the impulse response results (from osr)!?
The two things use inherently different objects (maximizing utility vs. minimizing some variance). While there is some relation (agents are risk averse and care about low fluctuations) I am not aware of an easy metric to compare the two. Probably you could look at consumption equivalents relative to steady state you would need to make agents indifferent between the two regimes.