Michael Ehrmann's headshot

The Overdelivery Premium: When Monetary Policy Decisions Exceed Market Expectations

Michael Ehrmann is Professor of Economics at the Frankfurt School of Finance and Management. His research focuses on monetary policy, central bank communication, financial markets, and household expectations, with numerous publications in leading academic journals. He previously held senior positions at the Bank of Canada and the European Central Bank, where he contributed to the analysis of monetary transmission and communication strategies. Combining academic expertise with policy experience, he is widely recognized for bridging research and central banking practice.

Abstract:

A large literature studies the effects of monetary policy based on the identification of monetary policy surprises. This paper tests whether monetary policy surprises exert different effects depending on whether the central bank decision exceeds market expectations (i.e., the central bank “overdelivers”) or falls short of them (the central bank “underdelivers”). Based on a panel dataset covering monetary policy decisions and market expectations for 14 advanced economies over nearly thirty years, we find strong and robust evidence for what we call an “overdelivery premium”, whereby overdelivery exerts substantially larger effects on financial markets. This is especially the case for short-maturity interest rates, where the effect can be up to 9 times as large. The paper then analyses potential channels that can generate this overdelivery premium. Overdelivery does not lead to a revision of the perceived central bank reaction function, but it generates a different effect on macroeconomic expectations of private agents: in response to an overdelivery tightening, economic forecasters revise down their GDP forecasts and revise up their inflation and unemployment forecasts, whereas the revisions have the opposite sign for an underdelivery tightening. The paper also shows that the results are unique to monetary policy surprises. They do not carry over to other macroeconomic surprises, suggesting that the results do not stem from psychological biases such as reference-dependent utility.

Accessibility
If you require accommodation, please contact the event host as soon as possible.
Date and time
Feb 6, 2026
2:30 p.m. to 4 p.m.
Format and location
In person
Vanier Hall (VNR)
VNR 3075
Language
English
Audience
Professors, Graduate students
Organized by
Department of Economics