Modelling Italian Potential Output and the Output Gap

58 Pages Posted: 23 Mar 2011

Multiple version iconThere are 2 versions of this paper

Date Written: September 15, 2010

Abstract

The aim of the paper is to estimate a reliable quarterly time-series of potential output for the Italian economy, exploiting four alternative approaches: a Bayesian unobserved component method, a univariate time-varying autoregressive model, a production function approach and a structural VAR. Based on a wide range of evaluation criteria, all methods generate output gaps that accurately describe the Italian business cycle over the past three decades. All output gap measures are subject to non-negligible revisions when new data become available. Nonetheless they still prove to be informative about the current cyclical phase and, unlike the evidence reported in most of the literature, helpful at predicting inflation compared with simple benchmarks. We assess also the performance of output gap estimates obtained by combining the four original indicators, using either equal weights or Bayesian averaging, showing that the resulting measures (i) are less sensitive to revisions; (ii) are at least as good as the originals at tracking business cycle fluctuations; (iii) are more accurate as inflation predictors.

Keywords: potential output, business cycle, Phillips curve, output gap

JEL Classification: E37, C52

Suggested Citation

Bassanetti, Antonio and Caivano, Michele and Locarno, Alberto, Modelling Italian Potential Output and the Output Gap (September 15, 2010). Bank of Italy Temi di Discussione (Working Paper) No. 771. Available at SSRN: https://ssrn.com/abstract=1788071 or http://dx.doi.org/10.2139/ssrn.1788071

Antonio Bassanetti (Contact Author)

Bank of Italy ( email )

Via Nazionale 91
Rome, 00184
Italy

Michele Caivano

Bank of Italy ( email )

Via Nazionale 91
Rome, 00184
Italy

Alberto Locarno

Bank of Italy ( email )

Via Nazionale 91
Rome, 00184
Italy

Here is the Coronavirus
related research on SSRN

Paper statistics

Downloads
47
Abstract Views
534
PlumX Metrics