“Optimal Filtering with Potential Output Uncertainty” by James YETMAN
University of Hong Kong
Existing papers studying optimal monetary policy when the central bank makes errors in measuring potential output typically treat those errors as exogenous to the model of the economy. Here we show that this is not an innocuous assumption, and can have important implications for optimal monetary policy. If measurement errors result from optimal filtering by the central bank, then they are a function of the shocks that hit the economy. Using a simple New Keynesian model that allows for analytical solutions to the optimal filtering problem, we show that a speed limit policy will be favored if the central bank places too much weight on recent data, and a price level target will be favored if the central bank places too little weight on recent data when estimating potential output.