Forecasting China's industrial output using a spatial Bayesian vector autoregressive model

Donald J. Lacombe, Nyakundi M. Michieka

Research output: Contribution to journalArticle

Abstract

This paper forecasts cement, steel, and TV production in China's top industrial provinces using a Bayesian prior that incorporates both time and spatial dependence as proposed by LeSage and Cashell (2015). Results indicate that growth in cement production will increase following the 3-year slump experienced between 2013 and 2016 in the five provinces in our sample. Average growth rates for steel production between 2017 and 2018 are similar to those experienced in 1999 and 2008. Our findings indicate that forecast accuracy for TV production demonstrate the superior forecasting characteristics of the hybrid prior.

Original languageEnglish
Pages (from-to)712-742
Number of pages31
JournalGrowth and Change
Volume49
Issue number4
DOIs
StatePublished - Dec 2018

Fingerprint Dive into the research topics of 'Forecasting China's industrial output using a spatial Bayesian vector autoregressive model'. Together they form a unique fingerprint.

  • Cite this