The predictive power of industrial electricity usage revisited: evidence from non‐parametric causality tests

• Date: 01 June 2018
• Author: Riza Demirer,Matteo Bonato,Rangan Gupta
• Published date: 01 June 2018
• DOI: http://doi.org/10.1111/opec.12119

The predictive power of industrial electricity

usage revisited: evidence from non-

parametric causality tests

1

Matteo Bonato*

,

****, Riza Demirer** and Rangan Gupta***

,

****

*Department of Economics and Econometrics, University of Johannesburg, Auckland Park, South Africa.

**Department of Economics and Finance, Southern Illinois University Edwardsville, Edwardsville, IL 62026-

1102, USA. Email: rdemire@siue.edu

***Department of Economics, University of Pretoria, Pretoria, South Africa.

****IPAG Business School, Paris, France.

Abstract

Recent research shows that the industrial electricity usage growth rate carries predictive ability

over stock market returns up to 1 year. Using the recently developed non-parametric causality tests

we show that the predictive power of industrial electricity usage can be explained by an ‘industry

effect’that is transmitted via the volatility channel. We argue that the countercyclical premium

associated with industrial electricity usage growth is driven by the industry components that drive

stock reversals, thus resulting in the negative relationship between today’s industrial electricity

usage and stock market returns in the future. The findings are in line with the notion that the returns

on industry portfolios are informative about macroeconomic fundamentals and suggest that the

informational value of industrial electricity usage as a business cycle variable may be an artefact of

return reversals driven by past industry performance.

1. Introduction

The predictive power of energy prices over the stock market has been the focus of

numerous studies in the literature. Recently, Da et al. (2017) propose a new, energy

market-based predictor for stock returns and show that the industrial electricity usage

growth rate can predict stock market returns up to 1 year, while it tracks the output of the

most cyclical sectors. Given that electricity cannot easily be stored, Da et al. (2017)

suggest that industrial electricity usage can be used to track production and output in real

time and thus, provides a reliable measure of economic growth, which in turn drives its

JEL classification: C22, C32, G1.

1

We would like to thank an anonymous referee for many helpful comments. However,

any remaining errors are solely ours.

©2018 Organization of the Petroleum Exporting Countries. Published by John Wiley & Sons Ltd, 9600 Garsington

Road, Oxford OX4 2DQ, UK and 350 Main Street, Malden, MA 02148, USA.

93