Exploring Monetary Policy Shocks with Large-Scale Bayesian VARs

Dimitris Korobilis · 2025

Read Exploring Monetary Policy Shocks with Large-Scale Bayesian VARs by Dimitris Korobilis (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2505-06649.txt.

Introduction

Dimitris Korobilis studies how central bank rate shocks spread through the United States economy, in a paper from two thousand twenty-five. He builds a large Bayesian model that tracks many series at once and treats shocks as hidden factors. The design keeps the math fast even as the data set grows, and it lets policy effects shift over time.

Shocks are picked out with high-speed market surprises plus sign rules on the results. The method held up across many model choices. A key result is that the Federal Reserve sway over single consumer prices shifted a lot in the high price-rise years from two thousand twenty-two to two thousand twenty-four. Readers get new detail on how policy reaches prices.

Catalog record

CREATORDimitris Korobilis
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE69 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2505-06649

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Exploring Monetary Policy Shocks with Large-Scale Bayesian VARs 0 0 footnotetext: I would like to thank Martin Bruns, Luca Gambetti, Domenico Giannone, Michele Lenza, Nicolò Maffei-Faccioli, Mirela Miescu, Ivan Petrella, Giorgio Primiceri, Barbara Rossi and Lorenza Rossi for their valuable comments and suggestions. I also thank participants at the University of Lancaster’s Workshop on Empirical and Theoretical Macroeconomics, the University of East Anglia “2nd Time Series Workshop”, the Collegio Carlo Alberto conference on “The Economics of Risk: Econometric Tools and Policy Implications”, and seminar participants at Universities of Manchester and Paris Dauphine for their insightful feedback. Any remaining errors are solely my responsibility. Correspondence: Professor of Econometrics, Adam Smith Business School, University of Glasgow, 2 Discovery Place, Glasgow, G11 6EY, United Kingdom; email: [email protected] . Dimitris Korobilis Abstract I introduce a high-dimensional Bayesian vector autoregressive (BVAR) framework designed to estimate the effects of conventional monetary policy shocks. The model captures structural shocks as latent factors, enabling computationally efficient estimation in high-dimensional settings through a straightforward Gibbs sampler. By incorporating time variation in the effects of monetary policy while maintaining tractability, the methodology offers a flexible and scalable approach to empirical macroeconomic analysis using BVARs, well-suited to handle data irregularities observed in recent times. Applied to the U.S. economy, I identify monetary shocks using a combination of high-frequency surprises and sign restrictions, yielding results that are robust across a wide range of specification choices. The findings indicate that the Federal Reserve’s influence on disaggregated consumer prices fluctuated significantly during the 2022–24 high-inflation period, shedding new light on the evolving dynamics of monetary policy transmission. Keywords: Disaggregated consumer prices; Latent factors; High-dimensional Bayesian VAR; Time-varying parameters; Sign restrictions; High frequency data JEL Classification: C11, C32, C55, E31, E52, E58, E66 1 Introduction This paper quantifies the effectiveness of conventional monetary policy shocks during the post-pandemic inflationary period using large-scale vector autoregressions (VARs). While previous studies ( Bańbura et al., 2010 ; Bernanke and Boivin, 2003 ; Giannone et al., 2015 , see, for example,) have established the benefits of identifying monetary policy in data-rich environments, this research develops a flexible yet numerically robust VAR framework to address emerging estimation and identification challenges. First, after a prolonged period of near-zero interest rates, historical data may be less informative for evaluating recent monetary policy effectiveness. 1 1 1 Studies examining conventional monetary policy (MP) shocks, such as Arias et al. (2019) ; Read (2024) ; Schlaak et al. (2023) , rely on data ending in 2007, that is, just before the period when nominal interest rates effectively reached their zero lower bound. Aruoba and Drechsel (2024) perform their empirical analysis to 2008, arguing that their findings remain relevant for the 2022–2023 inflationary period, provided that the transmission mechanism of MP shocks has not changed. However, they caution that factors such as state dependency, non-linearities, and structural shifts may alter response dynamics and invalidate projections to the recent period. Second, the COVID-19 pandemic introduced substantial outliers, disrupting macroeconomic correlations and complicating structural identification. 2 2 2 Lenza and Primiceri (2022) and Carriero et al. (2024a) develop VAR methods to accommodate COVID‑19 outliers and stochastic volatility; they do not, however, address structural shock identification or inference in high-dimensional systems. Third, in an increasingly globalized and financially volatile environment, enhancing the VAR with relevant predictors is crucial for statistical fit and for ensuring that structural shocks are well identified. Moving forward, addressing these data challenges requires novel statistical methods and estimation algorithms, capable of handling high-dimensional data, nonlinearities, and other complexities. A major contribution of this study is the development of such methods, allowing the integration of high-dimensional predictors to the application of robust identification techniques for structural shocks. To tackle these challenges, I develop a high-dimensional and data-analytic approach that integrates economic identification schemes with a Bayesian shrinkage estimation methodology. Building on Korobilis (2022) , I extend the VAR to integrate high frequency instruments, fat-tailed errors, time-varying parameters, and stochastic volatility. The approach uses “target” and “path” factors of high frequency surprise series from Gürkaynak et al. (2005) to look at how the markets react to Federal Reserve Bank’s policy announcements. The target factor describes the effects of surprise changes in the current federal funds rate target, that is, conventional shocks that are of interest in this paper. The path factor captures the future policy path which is used to identify a residual monetary policy shock. 3 3 3 There are varying views about the number and effects of future policy path and unconventional shocks, in general. Nakamura and Steinsson (2018) argue that the Fed has private information (what they call the “Fed information effect”), while Bauer and Swanson (2023) argue that both the Fed and professional forecasters respond to incoming economic news, which naturally generates a positive correlation between monetary policy surprises and forecast revisions, without requiring the Fed to have private information about the economy. Swanson (2021) extends Gürkaynak et al. (2005) to include a third, quantitative easing factor, and Jarociński (2024) identifies up to three unconventional monetary policy shocks. I first show that in a small-scale VAR the use of high-frequency surprises as internal instruments for identifying monetary policy is not always robust and in line with theory (see also Acosta, 2023 ; Lakdawala, 2019 , for similar arguments) . 4 4 4 Several other studies have raised concerns about the use of high-frequency instruments (HFIs) in identifying monetary policy shocks, particularly regarding issues of exogeneity, temporal aggregation bias, and event window limitations; see Bauer and Swanson (2023) , Paccagnini and Parla (2021) and Casini and McCloskey (2024) , among others. Next, I show that a plausible solution to this lack of identification is a hybrid method that merges proxies with zero and sign restrictions to the response of key macroeconomic aggregates -- namely measures of output, inflation, short/medium/long interest rates, and money supply. While sign restrictions to these variables align with conventional theoretical predictions of a monetary policy shock, they alone are also insufficient for unique identification and thus need to be combined with high-frequency instruments. 5 5 5 Braun and Brüggemann (2023) present similar arguments in the context of VARs that combine external instruments with sign restrictions. While their algorithm is very useful, it is not computationally feasible for large VARs, particularly when allowing for time-varying shock impacts. An emerging line of research, such as Acosta (2023) , Braun and Brüggemann (2023) , Carriero et al. (2024b) , Read (2024) and Schlaak et al. (2023) , demonstrates the advantages of combining established identification schemes, such as sign, zero and narrative restrictions, proxies, and identification via heteroskedasticity and non-Gaussianity. Prüser (2024) and Chan et al. (2022) similarly extend the VAR methodology of Korobilis (2022) by adapting it to identification by non-Gaussianity and heteroskedasticity, respectively. However, these identification approaches are statistical in nature, and may not always ensure adequate identification of conventional monetary policy shocks. In contrast, I provide comprehensive evidence that combining high-frequency instruments with sign restrictions remains economically robust across various data and modeling assumptions, including lag length, stationarity transformations of variables, assumptions about residual shocks, the measurement of short-term interest rates, and different assumptions regarding the exogeneity of instruments. After establishing that the hybrid identification scheme performs effectively in a

Read the complete text