We use robust control to study how a central bank in an economy with imperfect interest rate pass-through conducts monetary policy if it fears that its model could be misspecified. We find that, first, whether robust optimal monetary policy under commitment responds more cautiously or more aggressively depends crucially on the source of shock. Imperfect pass-through amplifies the robust policy. Second, if the central bank is concerned about uncertainty, it dampens volatility in the inflation rate preemptively but accepts higher volatility in the output gap and loan rate. However, for highly sticky loan rates, insurance against model misspecification becomes particularly pricy. Third, if the central bank fears uncertainty only in the IS equation or the loan rate equation, the robust policy shifts its concern for stabilization away from inflation.
Learning and monetary policy shifts
