What if structural breaks were in the number of shocks?

Crawford School of Public Policy | Centre for Applied Macroeconomic Analysis

Event details

Seminar

Date & time

Tuesday 27 October 2015
12.00pm–1.00pm

Venue

Seminar Room 2, Level 1, JG Crawford Building 132, Lennox Crossing, ANU

Speaker

Luis Uzeda, PhD student, Research School of Economics, ANU.

Contacts

Rossana Bastos
6125 8108

In this seminar Luis Uzeda will provide an overview of his recent paper, ‘What if Structural Breaks Were in the Number of Shocks?’. This paper proposes a statistical framework to investigate the number stochastic processes (shocks) underlying macroeconomic aggregates. The study suggests a novel perspective on structural breaks where breaks occur in the number of innovations driving a time series while its underlying structure, as denoted by latent components, is preserved. To model breaks, Luis uses a mixture innovation approach which leads to a state space with regime switching representation. The latter enables detection of whether, where and when breaks occur. Regime identifiability is discussed within a Bayesian setting. Rank reduction of the covariance matrix from reducing the number of innovations is also addressed. In particular, a Markov Chain Monte Carlo algorithm which builds upon the properties of Toeplitz matrices and recent advances in precision-based methods is developed. In accordance with modern macroeconomic models, Luis empirical results provide support to innovations-parsimonious state space models for various macroeconomic time series. Notably, evidence in favor of reducing the number of shocks is particularly stronger during recession episodes. A forecasting application to several measures of US inflation reinforces the usefulness of his framework.

Luis Uzeda is a PhD candidate from Research School of Economics (RSE) at ANU. His doctoral dissertation is concerned with developing methods for efficient Bayesian inference and statistical testing of state space models. Methodologies developed in his thesis find applications in forecasting, structural breaks and business cycle analysis. Prior to joining the PhD program at RSE, Luis worked as a research economist at the Reserve Bank of Australia.

The CAMA Macroeconomics Brown Bag Seminars offer CAMA speakers, in particular PhD students, an opportunity to present their work in progress in front of their peers, and reputable visitors to showcase their work. Speaker: Luis Uzeda, PhD Student, Research School of Economics

Updated:  20 October 2024/Responsible Officer:  Crawford Engagement/Page Contact:  CAMA admin