Estimating High Frequency Regional Output Growth Using Mixed Frequency Methods

cube-no-animation-1

Estimating High Frequency Regional Output Growth Using Mixed Frequency Methods

Research Seminar

Tuesday 12 June 2018, 16:30 — 17:30

Presented by: James Mitchell, Warwick Business School

with Gary Koop, Stuart McIntyre and Aubrey Poon (University of Strathclyde, ESCoE)

National Institute of Economic and Social Research, 2 Dean Trench Street, Smith Square, London SW1P 3HE

In this seminar James Mitchell discusses new work, undertaken with Gary Koop, Stuart McIntyre and Aubrey Poon (University of Strathclyde, ESCoE), to produce quarterly estimates of regional output growth in the UK from 1967 to 2016. Using these estimates they analyse how connected the different regions of the UK are, how this has changed over time and how the different regions of the UK respond to economic shocks.

They develop a mixed-frequency Vector Autoregressive (MF-VAR) model to produce their estimates from the available annual regional estimates and quarterly data for the UK as a whole. Temporal and cross-sectional restrictions are imposed on the MF-VAR to ensure that the quarterly regional estimates produced are consistent with the annual regional observations and the UK totals. The hierarchical Dirichlet-Laplace prior is used to ensure optimal shrinkage and, thus, parsimony in the MF-VAR.

James Mitchell is Professor of Economic Modelling and Forecasting at Warwick Business School, University of Warwick. As part of the research programme of the Economic Statistics Centre of Excellence he is involved in projects on both regional nowcasting and measuring and communicating data uncertainty. He holds a PhD in Economics from Cambridge, and is currently an Associate Editor at the Journal of Business and Economics Statistics, the Journal of Applied Econometrics and the International Journal of Forecasting.