OSR: indeterminacy

Hi All
Is it possible that the optimal simple rules computed by dynare violate the Blanchard-Kahn determinacy conditions?

No, that should never happen. If you encounter this in the most recent Dynare version, please send me the files to check for the presence of a bug.

Thanks Johannes. It is my colleague who is facing the problem using dynare 4.3.0. The problem occurs when when the osr command is used within a loop over lambda_y and lambda_r, the weights in the loss function over the variance of output and interest rate, in the monetary authority’s loss function. For each value of lambda_y lambda_r], the osr is initiated with the parameters set at the same standard values. At some point in the grid, once the optimal rule is found, dynare reports Blanchard-Kahn indeterminacy and the loop does not progress. Let me see if I can get the code.

Before doing so, please tell him to use the most recent unstable version. There were quite a few bugs affecting osr that have been fixed.