Birkbeck A&F Research Seminars
When:
—
Venue:
Online
A&F research seminar; 5 February 2026 (12- 1 pm)
Presentation Title: Monetary Policy, Large VARs and the Information Paradox
Invited Speaker: Professor Dimitris Korobilis, University of Glasgow, UK
Abstract: This paper documents an information paradox in monetary policy identification. Adding theoretically motivated variables to vector autoregressions can break rather than strengthen identification of structural shocks. In a core specification with output, inflation, interest rates, and financial conditions, three prominent high-frequency instrument sets, each constructed to address different concerns about information effects, news contamination, and weak identification, produce nearly identical monetary policy shocks. However, when the VAR expands to include additional macroeconomic variables such as unemployment or consumption, instruments alone yield economically implausible responses unless combined with sign restrictions. This finding challenges the conventional wisdom that larger information sets automatically improve structural inference. The paper develops a computationally tractable Bayesian framework that combines high-frequency instruments with sign restrictions and allows for time-varying transmission to disaggregated prices. Applied to U.S. data from 1995 to 2023, the results show substantial heterogeneity in how monetary policy affects different consumer price categories. Goods prices respond quickly while services prices, particularly housing, adjust sluggishly. The intensity of transmission has strengthened in the post-COVID period, suggesting that conventional monetary policy remains effective even in a high-inflation environment.
Bio:
Dimitris Korobilis is Professor of Econometrics at the Adam Smith Business School, University of Glasgow, and an adjunct researcher at the Center for Applied Macroeconomics and Commodity Prices (CAMP) at BI Norwegian Business School. He is also affiliated with the School of Computing Science at the University of Glasgow, a senior fellow of the Rimini Center for Economic Analysis (RCEA), and serves on the board of the European Seminar on Bayesian Econometrics (ESOBE). He is an associate editor for the Journal of Business and Economic Statistics and Studies in Nonlinear Dynamics and Econometrics.
His research focuses on time series analysis and forecasting of macroeconomic and financial data, with particular expertise in high-dimensional inference and computation. His work combines Bayesian statistics with machine learning methods to develop flexible forecasting and structural econometric models for central banks, governments, and policy institutions.
Contact name: Ellen Pei-yi Yu
Speakers-
Professor Dimitris Korobilis
—
Bio:
Professor Dimitris Korobilis is Professor of Econometrics at the Adam Smith Business School, University of Glasgow, and an adjunct researcher at the Center for Applied Macroeconomics and Commodity Prices (CAMP) at BI Norwegian Business School. He is also affiliated with the School of Computing Science at the University of Glasgow, a senior fellow of the Rimini Center for Economic Analysis (RCEA), and serves on the board of the European Seminar on Bayesian Econometrics (ESOBE). He is an associate editor for the Journal of Business and Economic Statistics and Studies in Nonlinear Dynamics and Econometrics.
His research focuses on time series analysis and forecasting of macroeconomic and financial data, with particular expertise in high-dimensional inference and computation. His work combines Bayesian statistics with machine learning methods to develop flexible forecasting and structural econometric models for central banks, governments, and policy institutions.
Tags: