Markov Switching Oil Price Uncertainty
| DOI | http://doi.org/10.1111/obes.12300 |
| Published date | 01 October 2019 |
| Author | Libo Xu,Apostolos Serletis |
| Date | 01 October 2019 |
1045
©2019 The Department of Economics, University of Oxford and JohnWiley & Sons Ltd.
OXFORD BULLETIN OF ECONOMICSAND STATISTICS, 81, 5 (2019) 0305–9049
doi: 10.1111/obes.12300
Markov Switching Oil Price Uncertainty*
Apostolos Serletis† and Libo Xu‡
†Department of Economics, University of Calgary, Calgary, AB T2N 1N4, Canada
(e-mail: serletis@ucalgary.ca)
‡Department of Economics, University of San Francisco, San Francisco, CA 94117, USA
(e-mail: lxu37@usfca.edu)
Abstract
We investigate whether the United States economy responds negatively to oil price un-
certainty and whether oil price shocks exert asymmetric effects on economic activity. In
doing so, we relax the assumption in the existing literature that the data are governed by
a single process, modifying the Elder and Serletis (2010) bivariate structural GARCH-in-
Mean VAR to accommodate Markov regime switching in order to account for changing
oil price dynamics over the sample period. We find evidence of asymmetries, against those
macroeconomic theories that predict symmetries in the relationship between real aggregate
economic activity and the real price of oil.
I. Introduction
There is an ongoing debate in macroeconomics about how oil price shocks and oil price
uncertainty affect the level of economic activity. Those of the view that positive oil price
shocks havebeen the major cause of recessions in the United States (and other oil-impor ting
countries) as, for example, Hamilton (1983, 1996, 2011), Hooker (1996), and Herrera, La-
galo and Wada(2011), appeal to models that imply asymmetric responses of real output to
oil price increases and decreases.These models are able to explain larger economic contrac-
tions in response to positive oil price shocks and smaller economic expansions in response
to negative ones. On the other hand, those of the view that positive oil price shocks do not
cause recessions as, for example, Kilian (2008), Edelstein and Kilian (2009), and Kilian
and Vigfusson (2011a,b), appeal to theoretical models of the transmission of exogenous
oil price shocks that imply symmetric responses of real output to oil price increases and
decreases. These models cannot explain large declines in the level of economic activity in
response to positive oil price shocks.
A series of recent papers by Elder and Serletis (2010, 2011), Bredin, Elder and Fountas
(2011), Rahman and Serletis (2011, 2012), Pinno and Serletis (2013), Jo (2014), Elder
(2018), and Serletis and Mehmandosti (2019) also look at the relationship between the price
of oil and the level of economic activity, focusing on the role of uncertainty about the price
JEL Classification numbers: C32, E32, G31.
*Wethank Jonathan Temple and an anonymous referee for comments that greatly improvedthe paper.
1046 Bulletin
of oil. They appeal to the real options theory, also known as investment under uncertainty,
which predicts that firms are likely to delay making irreversibleinvestment decisions in the
face of uncertainty about the price of oil, particularly when the cash flow from investments
is contingent on the oil price – see, for example, Bernanke (1983), Brennan and Schwartz
(1985), Majd and Pindyck (1987), Brennan (1990), Gibson and Schwartz (1990), and
Dixit and Pindyck (1994). In doing so, they utilize internally-consistent simultaneous
equations empirical models that accommodate an independent role for the effects of oil
price uncertainty. They find that oil price uncertainty has had a negative and statistically
significant effect on several measures of investment, durables consumption, and aggregate
output. They also find that accounting for the effects of oil price uncertainty tends to
exacerbate the negative dynamic response of economic activity to a negative oil price
shock, while dampening the response to a positive one.
Most of the recent literature on the macroeconomic effects of oil price shocks and oil
price uncertainty assumes that the data are governed by a single process. However, as re-
cently noted by Serletis and Mehmandosti (2019, p. 9), ‘the crude oil market in the United
States (and globally) has undergone very significant structural changes over the past 150
years due to technological innovation, revisions in regulatory regimes and market struc-
tures, the emergence of new players, such as OPEC, and changes in the sources of supply
and demand’. Motivated by these considerations, we assume a more complex process to
allow for different types of relationship at differenttimes. Twoclasses of models that allow
this to occur have been used so far in the literature – the ‘time-varying coefficient model’
and the ‘Markov switching model’. The former assumes that the relationship changes in
every period (which might not be the case in the real world), whereas the latter assumes a
switching mechanism from one state to another that is controlled by an unobserved variable
governed by a Markov process.
In this paper, we take the Markovswitching approach, associated with Hamilton (1989,
1990), which has been widely followed in the analysis of economic and financial time
series – see, for example, Garcia and Perron (1996) and Sims and Zha (2006). In doing
so, we modify the Elder and Serletis (2010) methodology by assuming Markov regime
switching to account for changing oil price dynamics over the sample period. In particular,
we use a Markov switching (identified) structural GARCH-in-Mean VAR in real output
growth and the change in the real price of oil, associate the oil price change VAR residual
with exogenous oil price shocks, use the conditional standard deviation of the forecast
error for the change in the real oil price as a measure of uncertainty about the impending
real price of oil, and investigate the relationship between the real price of oil and the
level of real economic activity in the United States using quarterly data since the early
1970s.
In the context of a mean-square stable structural GARCH-in-Mean VAR with Markov
regime switching, we find that oil price uncertainty has a negative and statistically signifi-
cant effect on the real output growthrate, and that this effect is asymmetric over contractions
and expansions in the business cycle, being significantly larger during contractions and
periods when large oil price changes occur. We also find that oil price uncertainty tends
to amplify the negative dynamic response of the growth rate to positive oil price shocks,
and to dampen the positive response to negative oil price shocks during expansions in the
business cycle.
©2019 The Department of Economics, University of Oxford and JohnWiley & Sons Ltd
To continue reading
Request your trial