Limits of central bank forward guidance under learning

Central bank forward guidance has emerged as a pertinent tool for monetary policymakers since the Great Recession. The conventional monetary policy tool of lowering overnight interest rates was exhausted when the zero lower bound on U.S. short-term nominal rates was effectively reached during the 2007-2009 global financial crisis. The Federal Reserve and other central banks around the world responded by pursuing the unconventional monetary policy tool of forward guidance. This new monetary policy tool is defined as communication of information about the future path of the policy rate to the public by the central bank. Nevertheless, the effects of forward guidance remain unclear. This paper investigates the effectiveness of forward guidance while relaxing two standard macroeconomic assumptions: rational expectations and frictionless financial markets. Agents forecast future macroeconomic variables via either the rational expectations hypothesis or a more plausible theory of expectations formation called adaptive learning. A standard Dynamic Stochastic General Equilibrium (DSGE) model is extended to include the financial accelerator mechanism. The results show that the addition of financial frictions amplifies the differences between rational expectations and adaptive learning to forward guidance. The macroeconomic variables are overall more responsive to forward guidance under rational expectations than under adaptive learning. During a period of economic crisis (e.g. a recession), output under rational expectations displays more favorable responses to forward guidance than under adaptive learning. These differences are exacerbated when compared to a similar analysis without financial frictions. Thus, monetary policymakers should consider the way in which expectations and credit frictions are modeled when examining the effects of forward guidance.