question about the optimal simple rule
Posted:
Tue Aug 24, 2010 2:04 am
by marvinsyrene
I was wondering what is the optimization method used to calculate the optimal simple rule?
thanks
Re: question about the optimal simple rule
Posted:
Mon Aug 30, 2010 8:59 am
by SébastienVillemot
Hi,
The “osr” command uses a rather straightforward algorithm: it runs a numerical optimizer. The optimizer explores the space of values for instruments declared in “osr_params”. The objective of the optimizer is the weighted variance/co-variance objective declared by the user. For a given value of instruments, this objective is computed by solving the model (as with “stoch_simul”) and by computing the corresponding theoretical or empirical moments.
Best,