Should a monetary policy maker following a Taylor‐type rule set a higher policy rate than the level suggested by the rule because of a possibility of an asset price bust in the near future? Our answer to this question for monetary policy makers who have two scenarios of ‘boom–bust cycle’ and ‘stable growth’ is yes if the following two conditions are satisfied. First, early warning indicators based on credit and residential investment data show a high probability of a boom–bust cycle occurring. Second, the policy rate path that minimizes the boom–bust probability‐based expected value of a social loss associated with inflation and the output gap over the two scenarios is higher than the rate path by the Taylor‐type rule. Our counterfactual analysis shows that the Fed should have raised the federal funds rate by a small amount over and above the level suggested by a Taylor‐type rule in the early 2000s.