Francesca Monti
Université catholique de Louvain, Belgium

Abstract:

This paper explores the dynamics of U.S. sectoral producer prices in a large Bayesian Vector Auto Regression (BVAR) model where the Input-Output (IO) matrix is used to structure their long-run relationships, allowing to capture flexibly the propagation of micro shocks to aggregate variables. This model’s forecasts of headline inflation have accuracy comparable to the Survey of Professional Forecasters’ and greater than those generated by a standard BVAR with Minnesota priors, confirming that the IO matrix long-run prior conveys relevant information about the data. We identify three key shocks – an oil price shock, a cereal price shock, and a monetary policy shock – using instrumental variables and find that sectoral linkages play a significant role in their transmission. Sectoral asymmetries are crucial for evaluating the macroeconomic consequences of energy price shocks, as industries with slower price adjustments amplify inflation persistence, even after the shock dissipates. And the impact of a sector-specific shock like the cereals price shock is non-negligible once the production network is accounted for.

Francesca Monti is Professor of Economics at the Université catholique de Louvain and a research affiliate of CEPR. She holds a PhD in Economics and Statistics from the Université Libre de Bruxelles. Francesca worked at the Bank of England for 10 years, focusing mainly on macro modelling and monetary policy strategy issues, before moving full-time to academia. She spent a couple of years at King’s Business School in London, before taking on her current position. Her research spans the fields of macroeconometrics, empirical macro and monetary economics.