Learning, Hedging, and the Natural Rate Hypothesis∗

We assume that firms actively manage risk by learning and hedging in two macroeconomic models, a simple cobweb model and a model of monopolistic competition that allows for the analysis of monetary policy and welfare. In both models, firms that learn (by paying a cost to observe the model’s stochastic shocks) face less uncertainty and produce more output than firms that hedge. Parameter or policy changes that increase the attractiveness of learning therefore induce a higher steady state level of output. When we introduce monetary policy into the latter model, we obtain three results that are profoundly different from all or most of the related literature. First, by affecting the volatility of the system, monetary policy affects the supply side decisions of producers and therefore affects the steady state level of output and not just its second moment. Second, if the fraction of firms that learn is exogenous, then the optimal policy maximizes steady state output and welfare by minimizing the volatility of the aggregate price level. Third, if the fraction of firms that learn is endogenous, then the optimal policy maximizes steady state output by ensuring an intermediate level of price stability, close to the minimum stability needed to induce the maximum amount of learning.