Oil, Inflation Expectations, and Household Characteristics: A Nonlinear Heterogeneous Agent VAR Approach

Abstract:

In this paper, we develop a scalable micro-macro modelling framework that integrates linear multivariate time series models with nonlinear panel models to investigate to what extent oil supply shocks affect household inflation expectations in the Euro area in a heterogeneous and nonlinear way. We rely on a unique and very rich multi-country micro dataset of quantitative inflation expectations from the European Commission’s monthly business and consumer survey that allows us to construct pseudo individuals based on demographic and socioeconomic characteristics. To capture nonlinearities in group-level dynamics, we explicitly model the responses of our pseudo individuals as nonlinear functions of area-wide macroeconomic aggregates using Bayesian Additive Regression Trees (BART). We find pronounced asymmetries in the response of aggregate inflation expectations to large oil supply shocks and a considerable degree of heterogeneity across countries, gender, income, and age groups. We explore several economic mechanisms to explain cross-country and group-level differences in the adjustment of inflation expectations to oil supply shocks of different magnitude and sign.

Gary Koop, University of Strathclyde