Working Paper: CEPR ID: DP4135
Authors: Martin Ellison
Abstract: In this Paper, we suggest a new motivation for why central banks appear averse to reversing recent changes in their interest rate. We show, in a standard monetary model with forward-looking expectations, data uncertainty and parameter uncertainty, that there is a learning cost associated with interest rate reversals. A policy that frequently reverses the interest rate makes it more difficult for the central bank and private agents to learn the key parameters of the model. Optimal monetary policy internalizes this learning cost and therefore has a lower number of interest rate reversals. The incentive to reduce the number of interest rate reversals is in additionto the optimal policy inertia created by the presence of forward-looking expectations and uncertainty in the model.
Keywords: interest rate smoothing; learning; monetary policy
JEL Codes: D83; E52; E58
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| frequent interest rate reversals (E43) | hinder ability to learn monetary transmission mechanism (E49) |
| optimal monetary policy (accounting for learning costs) (E63) | lower frequency of interest rate reversals (E43) |
| active learning policy (J68) | less interest rate reversals compared to passive learning policy (E43) |
| active learning policy (J68) | increased persistence of short-term nominal interest rate (E43) |
| active learning policy (J68) | reduced volatility of short-term nominal interest rate (E43) |
| optimal monetary policy (E63) | promotes learning and improves welfare (I25) |