Essays on monetary economics

This dissertation consists of three essays on monetary economics. The first paper examines how optimal monetary policy responds to credit shocks. When the shock occurs monetary policy is eased dramatically, to the extent that the deflationary forces associated with a positive credit shock are reversed, so a mild inflation is induced. Subsequently, inflation goes below baseline. Output initially falls at the time of the shock, and then spends a considerable amount of time above baseline as the capital stock is replenished. It is far less volatile than under a Taylor Rule benchmark. The design of optimal policy critically depends on whether the economy begins at the first best level of output. The second paper examines the role of multiple aggregate shocks in monetary models with imperfect information. Because agents can draw mistaken inferences about which shock has occurred, the existence of multiple aggregate shocks profoundly influences macroeconomic dynamics. In particular, after a contractionary monetary shock these models can generate an initial increase in inflation (the "price puzzle") and a delayed disinflation (a "hump"). A conservative calibration exercise exhibits these patterns. In addition, the model shows that increased price flexibility is potentially destabilizing.

The third paper considers the claim made in Rudebusch (2002b) that the Federal Reserve does not smooth interest rates. Support for this proposition was offered in a New Keynesian model, where smoothing led to predictability in interest rate changes, predictability not seen in the data. This paper estimates some of the key parameters of Rudebusch's model using a minimum distance estimator. The data support an interest rate policy that is very inertial. Indeed, policy is so inertial that it may be better characterized as an equation in changes. With the estimated parameters the inertial policy does not result in predictable changes in the interest rate.