Bayesian Macroeconometric Modelling

Advances in Econometrics Volume 45 workshop

  • When9am to 5pm

  • WhenThursday-Friday
    16-17 April 2020

  • WhereTerrace Room, Level 6
    Sir Llew Edwards Building

This workshop aims to bring together researchers working in applied and theoretical macroeconometrics to discuss recent developments in modelling and inference using Bayesian methods.

This event is supported by the University of Queensland, Australian Research Council,  Economic Society of Australia (Queensland Chapter) and the EFaB (Economics, Finance and Business) Section of ISBA (The International Society for Bayesian Analysis).

About the workshop

Bayesian methods have become core to the exploration of a range of issues in macroeconometrics. Improved computing power and computational algorithms have made computation of posterior distributions more accessible. We have seen an expanding range of model specifications and techniques for model selection over the past few years. These advances have improved inference and thereby improved our understanding of significant macroeconomic issues. The dimensions of models or the systems that can be modelled have grown from very few – often at most 10 variables – to systems with 200 or more variables and potentially hundreds of thousands of parameters and states. Forecasting has also benefited from advances including increased dimensions of VARs, better understanding of the values of variable selection versus shrinkage (the dense versus sparse model issue) for dealing with high dimensional models, and improved estimation and inference in VARMAs.

Several challenges in macroeconometrics remain. Despite recent progress, estimation of VARMAs with the attending issues of identification, specification and implied reduced sources of errors remains problematic. More generally, economic theory is a delicate flower in that when confronted with empirical evidence, the data decisively reject the economics. Features of reduced from models such as restrictions implied by aspects or components of these theories, however, often have significant empirical support. Time variation has proven important for learning about the evolution of economic behaviour, but with larger time varying parameter models, it has become more important to identify which parameters vary and which do not.

In response to these outstanding questions, this special issue will provide a venue for researchers to discuss these issues and present solutions.

Learn more about Advances in Econometrics

Keynote speakers

Professor Gary Koop, University of Strathclyde

Professor Sylvia Kaufmann, Study Centre Gerzensee (Foundation of the Swiss National Bank)

Transport and accommodation

Please view the transport options and list of hotels and apartments which the UQ School of Economics recommends. 

Organising/Program committee

The program and local organising committees are chaired by Rodney Strachan (The University of Queensland), a Senior Co-Editor of AIE. Other members of the committee are Joshua Chan (Purdue) and Eric Eisenstat (The University of Queensland).

 

Register

Registration is free and now open. Contact Rodney Strachan at r.strachan@uq.edu.au with any enquiries.

Register now