Asymmetries in the response of economic activity to oil price increases and decreases?

https://doi.org/10.1016/j.jimonfin.2014.09.004Get rights and content

Highlights

  • We explore the question of asymmetry in the response of industrial production (IP) to oil price shocks.

  • We find some evidence of asymmetry for large oil exporters or importers.

  • Yet, this evidence vanishes when data-mining robust critical values are used.

  • A linear model is a good approximation for the response of IP to an oil price shock.

Abstract

It has been common to assume that the relationship between economic activity and oil prices is asymmetric. Theoretical underpinnings for this asymmetry include costly sectoral reallocation, partial equilibrium models of irreversible investment, and some version of precautionary savings. Yet, recent studies that use new methodologies to test for asymmetries in U.S. data have cast some doubts on that premise. In this paper, we use state-of-the-art techniques to evaluate the presence of asymmetries for a set of OECD countries containing both oil exporters and oil importers. We find very little support for the hypothesis that the response of industrial production to oil price increases and decreases is asymmetric. Our results have important implications for theoretical models of the transmission of oil price shocks: they point towards the importance of direct-supply and direct-demand transmission channels, as well as indirect transmission channels that imply a symmetric response.

Introduction

How does economic activity respond to oil price shocks? Does economic activity contract when oil prices increase, but no boom ensues when oil prices fall? Until recently, a consensus seemed to exist regarding the asymmetric nature of the relationship between oil prices and the macroeconomy. Indeed, discussions in academic and policy circles often refer –explicitly or implicitly– to the asymmetric nature of the relationship between GDP growth and oil price shocks (see, e.g., Bernanke et al., 1997, Bernanke, 2006). However, recent work by Kilian and Vigfusson (2011a) has called into question the view that oil price innovations have an asymmetric effect on U.S. GDP growth. In particular, they prove that the methodology commonly used in the empirical literature to assess the possible asymmetry in the response of economic activity to oil price shocks may lead to inconsistent parameter estimates due to a censoring bias. Moreover, they show that the slope based tests carried out in most studies are not informative about the presence of asymmetry in the impulse response functions.

Evaluating whether the relationship between oil prices and economic activity is symmetric constitutes a crucial step in deciding how to model oil prices, in selecting among alternative models of the transmission of oil price shocks, and in getting a good grasp on the magnitude of the macroeconomic effect that positive and negative innovations entail. Despite the large number of studies on the relationship between oil prices and the macroeconomy, almost all we know regarding the importance of the asymmetric channels of oil price transmission has been informed by U.S. data (e.g., Kilian and Vigfusson, 2011b, Kilian and Vigfusson, 2011a) or stems from estimating censored models with international data, which may lead to inconsistent estimates of the impulse response functions.4 One thus has to wonder whether the results obtained for the U.S. are just a figment of the particular data set or whether it is a result that generalizes to other countries. If it is the former, then there is no need to revise the way in which we model asymmetries; yet, if it is not, then we should re-think not only the estimation techniques, but also our theoretical models of the transmission of oil price shocks.

Our aim is to explore whether the empirical implications of different models of the transmission of oil price shocks are borne out by the country-level responses to positive and negative oil price innovations. To do so we examine the estimated responses of industrial production for 18 countries belonging to the Organization for Economic Co-operation and Development (OECD). We believe this data set provides a good testing ground for several reasons. First, theoretical explanations for an asymmetric response apply both to oil producers and oil importers, but have only been explored using impulse response based tests for U.S. data. Clearly, the theoretical underpinnings for an asymmetric response apply not only to the U.S., but also to other large net oil importers such as Japan and Germany (see Fig. 1).

Second, data for oil exporting countries is particularly valuable. In particular, crude oil production represents a large part of GDP for Norway and Canada (see Fig. 2); hence, conditions for an unexpected oil price innovation to have a large economic impact hold by construction. These data allow us to evaluate whether the presence (or absence) of asymmetry in the industrial production-oil price relationship is related to the share of crude oil production in aggregate output.

Finally, the degree of energy intensity in consumption varies greatly across OECD countries. For instance, among oil exporters, whereas energy intensity measured in total primary energy consumption (in BTU) per dollar of 2000 GDP was 13,097 for Canada in 2006, it was only 5267 for Denmark (see Fig. 3). During the same year, energy intensity was about 36% higher for the U.S. than for Japan, the largest and second largest oil importers in the sample, respectively. This disparity in the degree of energy intensity suggests that the magnitude of the response of consumption –and thus production– to positive and negative oil price shocks might vary significantly across countries.

The contribution of this paper to the ongoing debate is threefold. First, we shift the focus from the U.S. data used in the current debate (see, e.g., Kilian and Vigfusson, 2011a, Kilian and Vigfusson, 2011b, Hamilton, 2011, Herrera et al., 2011) to a larger sample including both net oil importing and exporting countries. Second, we propose a measure of the distance between impulse response functions, which allows us to evaluate the magnitude of the asymmetry between the response to a positive and a negative oil price innovation. This measure of distance is the cumulative Euclidean norm and provides additional insights into the economic significance of the asymmetry. Third, we use state-of-the-art econometric techniques to test for symmetry in the impulse response functions.5

We find very little evidence of asymmetry in the response of industrial production to positive and negative oil price innovations. We reject the null of symmetry in response to a one standard deviation –hereafter 1 s.d.– innovation for the G7 when we use the oil price increase. Similarly, we reject the null for Greece, Sweden, and the U.S. when we use the net oil price increase relative to the previous three-year maximum. For a two standard deviation –hereafter 2 s.d.– innovation, we reject the null of symmetry for the U.S. when we use the net oil price increase relative to the previous year maximum. Our results suggest that the transmission of oil price shocks to the macroeconomy takes place mainly through transmission channels that do not imply an asymmetric response.

Despite the fact that the statistical test very rarely rejects the hypothesis that the asymmetric impulse response estimates equal the impulse response estimates from the symmetric model, this finding does not imply that we should rule out alternative nonlinear models. In particular, we only consider the three particular forms of nonlinearity most commonly used in the literature. Yet, the test may have low power against alternative forms of nonlinearity such as those stemming from an asymmetric response to changes in oil uncertainty (measured as an increase in the standard deviation) of the real oil price.

The remainder of this paper is organized as follows. Section 2 discusses the theoretical underpinnings behind the transmission of oil price shocks, with particular emphasis on the empirical implications regarding asymmetry. The data on industrial production and oil prices are described in section 3. Section 4.1 describes the response of industrial production growth to positive and negative oil price shocks, Section 4.2 discusses the magnitude of the asymmetry, and Section 4.3 presents the results for an impulse response based test of symmetry. We present our conclusions in Section 5.

Section snippets

Asymmetries in the transmission of oil price shocks: theoretical underpinnings

This section briefly reviews the theoretical literature on the transmission of oil price shocks. Given our object of interest, we focus on whether a particular channel leads to an asymmetric response in economic activity.

Data

To investigate the presence of asymmetries in the relationship between oil price shocks and economic activity, we use monthly data on oil prices and industrial production (IP) indices for eighteen OECD countries. These countries are Austria, Belgium, Canada, Denmark, Finland, France, Germany, Greece, Italy, Japan, Luxembourg, Netherlands, Norway, Portugal, Spain, Sweden, the UK, and the U.S. We also consider 3 groups of countries: the G-7, OECD-Europe and OECD-Total. There are various reasons

The response of industrial production growth to oil price shocks

In this section, we evaluate the economic importance and test for the statistical significance of asymmetries in the relationship between oil price changes and industrial production growth.

Conclusions

How does economic activity respond to oil price shocks? Until recently, the consensus was that oil price increases lead to recessions in oil importing countries, but price decreases did not lead to expansions. More precisely, a number of empirical studies showed that slope based tests rejected the null of symmetry for U.S. GDP growth. These findings were confirmed by Mork et al., 1994, Cuñado and Pérez de Gracia, 2003, and Jiménez-Rodríguez and Sánchez (2005) using data for OECD countries.

Yet,

References (51)

  • Sylvain Leduc et al.

    A quantitative analysis of oil-price shocks, systematic monetary policy, and economic downturns

    J. Monetary Econ.

    (2004)
  • Kiseok Lee et al.

    On the dynamic effects of oil price shocks: a study using industry level data

    J. Monetary Econ.

    (2002)
  • Andrew Atkeson et al.

    Models of energy use: putty-putty versus putty-clay

    Am. Econ. Rev.

    (1999)
  • Yusuf Soner Başkaya et al.

    Oil price uncertainty in a small open economy

    IMF Econ. Rev.

    (2013)
  • Christiane Baumeister et al.

    What central bankers need to know about forecasting oil prices

    Int. Econ. Rev.

    (2014)
  • Christiane Baumeister et al.

    The role of time-varying price elasticities in accounting for volatility changes in the crude oil market

    J. Appl. Econ.

    (2013)
  • Christiane Baumeister et al.

    Time-varying effects of oil supply shocks on the U.S. economy

    Am. Econ. J. Macroecon.

    (2013)
  • Ben S. Bernanke

    Irreversibility, uncertainty, and cyclical investment

    Q. J. Econ.

    (1983)
  • Ben S. Bernanke

    “The Economic Outlook,” Remarks before the National Italian American Foundation, New York

    (November 28 2006)
  • Ben S. Bernanke et al.

    Systematic monetary policy and the effects of oil price shocks

    Brookings Pap. Econ. Activity

    (1997)
  • Hilde C. Bjørnland et al.

    Bloom or gloom? Examining the Dutch Disease in a Two-speed Economy

    (2013)
  • O.J. Blanchard et al.

    The Macroeconomic effects of oil price shocks: why are the 2000s so different from the 1970s?

  • Timothy F. Bresnahan et al.

    Segment shifts and capacity utilization in the U.S. Automobile Industry

    Am. Econ. Rev. Pap. Proc.

    (1993)
  • Steven J. Davis

    Allocative disturbances and specific capital in real business cycle theories

    Am. Econ. Rev.

    (1987)
  • Edelstein, Paul, Lutz, Kilian, 2007. The response of business fixed investment to changes in energy prices: a test of...
  • Cited by (109)

    View all citing articles on Scopus

    The on-line appendix is available at http://gatton.uky.edu/faculty/herrera/documents/HLWcou_appendix.pdf. We thank Lutz Kilian, Christiane Baumeister, seminar participants at the 2011 Midwest Econometrics Group Meeting, the Norges Bank workshop on Modeling and Forecasting Oil Prices, the University of Saskatchewan, the University of Windsor, Clemson University, and the University of Kentucky, and two anonymous referees for helpful comments and suggestions. All remaining errors are ours. This paper was previously circulated with the titles ”Nonlinearities in the Oil Price-Industrial Production Relationship: Evidence from 18 OECD Countries” and ”Asymmetries in the Transmission of Oil Price Shocks to Industrial Production? Evidence from 18 OECD Countries.”

    1

    Tel.: +1 859 257 1119; fax: +1 859 323 1920.

    2

    Tel.: +1 906 487 2311.

    View full text