Abstract
Why are markups cyclical, and why does their cyclicality vary across firms, sectors, and business cycle episodes? This paper develops a theory of pricing under uncertainty in which risk-averse firms set prices before costs and demand are realized. In this environment, pricing serves a precautionary role: optimal markups incorporate a risk premium that reflects the value of limiting exposure to uninsurable cost and demand risk. A central implication is that markup dynamics depend on the composition of risk firms face. Cost risk and demand risk affect pricing incentives in systematically different ways, so whether markups are procyclical or countercyclical depends on which source of risk dominates and how it co-moves with economic activity. We test these predictions using industry-level data for U.S. manufacturing over 1976–2018. Consistent with the theory, a one-standard-deviation increase in cost uncertainty raises prices by 5.8%, while a comparable increase in demand uncertainty lowers prices by 6.6%. The structural calibration of the dynamic GE model and the corresponding quantitative exercises—impulse responses to uncertainty shocks, welfare costs, the historical 1976–2018 path, and untargeted cross-industry validation—are work in progress and will appear in a subsequent draft.
