Energy Economics 33 (2011) 975986

Contents lists available at ScienceDirect

Energy Economics
j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / e n e c o

Oil price volatility and stock price uctuations in an emerging market: Evidence from
South Korea
Rumi Masih a,1, Sanjay Peters b,2, Lurion De Mello c,
a
b
c

JP Morgan Asset Management, New York, NY 10167, USA

Department of Economics, IESE Business School, 08034, Spain
Macquarie University, Sydney, Australia

a r t i c l e

i n f o

Article history:
Received 10 April 2007
Received in revised form 26 March 2011
Accepted 27 March 2011
Available online 13 April 2011
JEL classication:
F31
C22
C52
Keywords:
Emerging markets
Real stock price
Oil price shocks
Industrial production
Generalized variance decompositions
Impulse response functions

a b s t r a c t
How important are oil price uctuations and oil price volatility on equity market performance? What are the
policy implications if volatility turns out to be signicant? We assess this issue in an economics/nance nexus
for Korea using a VEC model including interest rates, economic activity, real stock returns, real oil prices and
oil price volatility. Our main aim is to capture the effects of crude oil prices on the Korean economy thoroughly
covering the period of the Asian Financial Crisis of 1997, which heavily affected the country, and the oil price
hikes in the early 1990s after the Gulf War. South Korea was the country most hit by the nancial crisis
together with Indonesia and Thailand. Results indicate the dominance of oil price volatility on real stock
returns and emphasize how this has increased over time. Oil price volatility can have profound effect on the
time horizon of investment and rms need adjust their risk management procedures accordingly. This
increase in dependency has been found in other net oil importing emerging equity markets. We test the
relationship between oil price movements and economic activity by using modern time series techniques in a
cointegrating framework. We expand the standard error correction model by examining the dynamics of out
of sample causality through the generalized variance decomposition and impulse response function
techniques. The evidence from persistence proles also gives important guidelines based on how fast the
entire system adjusts back to equilibrium. In addition, we nd the cointegrating relationship to be stable and
nd that the linear error correction model to be more favorable than an asymmetric 2 period Markov
switching model.
2011 Elsevier B.V. All rights reserved.

1. Introduction
Since the oil price shocks of 197374 and 197980, dozens of
academics and practitioners have explored the relationships between
oil price shocks and the macroeconomic variables. The recessionary
impacts of these oil prices shocks were too close for possible causal
links to be ignored, and considerable attention has been devoted to
study the macroeconomics of these events. Policymakers have to take
serious account of the developments in the oil market, as a rise in the
world price of oil imposes macroeconomic costs in two ways. First, to
the extent that oil is both an important input to production and
consumer goods (i.e. petrol and heating oil), results in a reduction in
economic activity as energy becomes more expensive. Second, rising
oil prices contribute directly to the level of ination, particularly in

Corresponding author at: Department of Economics, Macquarie University, Sydney

NSW 2109, Australia. Tel.: + 61 412560753.
E-mail addresses: rumi.x.masih@jpmorgan.com (R. Masih), speters@iese.edu
(S. Peters), lurion.demello@mq.edu.au, lurion@gmail.com (L. De Mello).
1
Tel.: + 1 212 648 1723.
2
Tel.: + 34 93 253 4200.
0140-9883/$ see front matter 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.eneco.2011.03.015

energy dependent countries. Over time, the impact on activity and

ination will also depend on policy responses and supply-side
effects.3
The high oil prices in 2005 and 2006 reect the booming demand
from Asia (especially China and India)4 and the geopolitical risk in the
Middle East5 (the fear premium estimated to add between $4 and $8
to current prices). China and India have become two principal players
on the global energy scene. In 1990, consumption in these two
countries amounted to no less than 3.5 million barrels per day,
approximately 5% of global petroleum use. In 2003, 13 years later,
these two countries account for more than 10% of global oil
consumption. (BP Statistical Review of World Energy Markets, 2004).
It is difcult to distinguish temporary shocks from permanent
shocks; and uncertainties related to large changes in oil prices can have
signicant effects on consumer condence and therefore on growth.
The impact of these oil price shocks is likely to be signicantly greater in
oil-importing countries, especially where policy frameworks are weak,
3
4
5

See Hamilton (1983), Bohi (1989), Bernanke et al. (1997).

See Heap (2005),Radetski (2006).
See Stevens (2005).

976

R. Masih et al. / Energy Economics 33 (2011) 975986

foreign exchange reserves are low, and access to international capital

markets is limited.
Government authorities combat oil price hikes by using monetary
and scal policies. For example, in order to maintain high industrial
production and exports revenues, interest rates are kept at low levels.
However, it is difcult to evaluate the impact of the oil price shocks on
different variables of the macroeconomic environment, especially the
impact on the stock markets. It is perhaps noteworthy that it was only
in the 1990s that researchers seriously examined the impact of oil
price shocks on stock markets.6 Macroeconomics and nancial
dynamics have not been captured together in one model when
subjected to oil price shocks, especially for a net oil importing and
emerging economy such as South Korea.
The Republic of Korea (South Korea) is important to world energy
markets, as it happens to be the seventh largest oil consumer and the
fth largest net oil importer in the world. Korea relies entirely on oil
imports as there are no oil reserves in the country or surrounding
areas. Hence, being a net importer of oil, the movements or
uctuations in oil prices are of major relevance for the Korean
government when taking policy decisions that affect the national
economy. The two major oil crises of the 1970s and 1980s signicantly
affected South Korea's macroeconomic performance. By using
monetary and scal policies the country was able to weather the
rst oil price shock with some difculties, but coping with the second
crisis was much more challenging and South Korea experienced the
worst stagation between 1973 and 1990.
Korea's reliance on foreign sources for its energy requirements
drastically increased between 1960 and 1990. For instance, Korea's
total indigenous energy production to foreign imports continuously
declined from 54.77% in 1971, to 30.21% in 1980, and to 29.95% in
1990. This, in turn, increased foreign energy dependence from 71.7%
in 1971 to 77.95% in 1980 to 82.01% in 1990.7
More recently, petroleum accounted for 54 percent of South
Korea's primary energy consumption in 2002. In 2004, the country
consumed around 2.14 million barrels a day (bbl/d) of oil, all of which
was imported.8 Glasure (2002) indicates that the real oil price is the
major determinant of real income and energy consumption in South
Korea.
With the close dynamics between economic indicators and
nancial markets, many studies have used various proxies to illustrate the degree and direction of causality in a cointegrating VAR
framework. To our knowledge these dynamics have not been
observed together with exogenous oil price movements and oil
price volatilities.
This paper tries to answer the following questions. What is the
long-run relationship between oil price movements and stock
markets in an emerging market like South Korea? Did the stochastic
trends change between industrial production, interest rates, stock
markets and oil price change during the nancial crises and oil price
hikes in the early 1990s? What is the direction of causality between
these variables and what are the implications for the transmission
mechanisms of shocks? Can the domestic stock market be isolated
from oil price movements? Answers to these questions will have
serious scal and monetary policy implications not only to Korea but
also to other energy dependent countries. The negative impact of oil
price volatility on Korean industries could help the government in
looking at alternative less volatile sources of fossil fuels such as
nuclear, coal and Liqueed Natural Gas (LNG) which has continued to
grow in popularity especially in electricity generation.9 To ensure
6
Driesprong et al. (2003), argue that changes in oil prices strongly predict future
stock market returns in 12 out of 18 developed countries surveyed. South Korea is not
on the list of the countries surveyed.
7
Glasure (2002).
8
Data from Energy Information Administration (EIA), US Department of Energy.
9
According to EIA estimates in 2008, South Korea primary energy consumption
constituted of 45% Petroleum, 27% Coal and 14% Natural Gas.

energy substitution, proper infrastructure needs to be in place to

make sure industries have the means to convert from oil intensive to
gas or coal intensive processes.
The paper is structured as follows. Section 2 provides a review of
the literature and main debates surrounding the dynamics between
stock markets and economic markets, together with impacts from
energy and oil price movements. In Section 3 we describe the data
used in the analysis and briey discuss econometric concepts and
methodology surrounding multivariate cointegration analysis and the
out-of sample testing framework. The application and estimation
results are presented in Section 4 with some tables and gures
presented in Appendix A of the paper. In Section 5 we draw some
important policy conclusions with respect to monetary policy and
policies designed for stock markets to withstand oil price movements.
2. Literature review
James Hamilton's (1983) study of the role of oil price shocks in US
business cycles has had considerable inuence on research on the
macroeconomics of oil price shocks. As Mork et al. (1994) review
paper outlines, economists worked for nearly a decade on methods of
incorporating oil price shocks into macroeconomic models before a
synergy developed with real business cycle (RBC) models and oil price
shocks. This theoretical relationship between macroeconomics and oil
price movements has been applied and tested using various
econometric techniques.
Chaudhuri and Daniel (1998) use cointegration and causality to
demonstrate that nonstationary behaviour of the US dollar real
exchange rate is explained by nonstationary behaviour of real oil
prices. The authors argue that oil price shocks can have long-run
effects on real exchange rates even if perfect markets exist in the long
run.
Greene et al. (1998) assess the impact of cartels like OPEC on the
U.S. economy. They identify three main separate and additive types
of economic losses resulting from oil prices increases: the loss of the
potential to produce, macroeconomic adjustment losses and the
transfer of wealth from US oil consumers to foreign oil exporters.
Whereas, Kaneko and Lee (1995) use an eight-variable VAR model to
test the pricing inuence of economic factors on U.S. and Japanese
stock market returns and in identifying their relative importance in a
dynamic context. The eight variables used in this study are as follows:
risk premium, term premium, growth rate in industrial production,
rate of ination, changes in terms of trade, changes in oil prices,
change in exchange rates and excess stock returns. They nd the
average values of excess stock returns, rates of ination, risk
premiums and term premiums to be higher for the United States
than for Japan.
Papaetrou (2001) on the other hand tests the dynamic linkage
between crude oil price and employment in Greece using industrial
production and industrial employment as alternative measures of
economic activity. His study is modelled in a cointegrated VAR
framework and extends out by looking at the generalized variance
decomposition and impulse response functions, which is very
encouraging as most studies have not gone beyond cointegration
and error corrections modelling.
Sadorsky's (1999) research meanwhile draws attention to a
negative relationship between shocks in oil prices and real stock
returns for the US economy and a negative impact of shocks to real
stock returns on interest rates and industrial production.
In a later study, Sadorsky (2001) nds a signicant and positive
relationship between oil and gas equity index and the price of crude
oil in Canada. Furthermore the author indicates a positive relationship
between the return on the index and the return on the stock market as
a whole. Finally a negative association is found between the stock
market index value and both the premium on 3-month vs. 1-month
Government debt and the US/Canadian Dollar exchange rate.

R. Masih et al. / Energy Economics 33 (2011) 975986

977

oped by Phillips (1987), Phillips and Perron (1988), and Perron

(1988) assume as its null hypothesis that a unit root exists in the
autoregressive representation of the time series, while the null
hypothesis for the KPSS assumes the opposite (i.e., that a unit root
does not exist). The Dickey-Fuller tests (Dickey and Fuller, 1981),
attempt to account for temporally dependent and heterogeneously
distributed errors by including lagged sequences of rst differences of
the variable in its set of regressors. The PP tests try to account for
dependent and IID processes through adopting a non-parametric
adjustment, hence eliminating any nuisance parameters. Recently
these tests have been shown, by Schwert (1987) and DeJong et al.
(1992), to suffer from lack of power as they often tend to accept the
null of a unit root too frequently against a stationary alternative.

Sadorsky's (1999) previous study on the US economy shows that

oil price volatility shocks have asymmetric effects on the economy. By
analyzing the impulse response functions, he shows that oil price
movements are important in explaining movements in stock returns:
after 1986, furthermore, oil price movements explain a larger fraction
of the forecast error variance in real stock returns than do interest
rates. The results nally suggest that positive shocks to oil prices
depress real stock returns while shocks to real stock returns have
positive impacts on interest rates and industrial production.
Faff and Brailsford (1999) study the sensitivity of the Australian
industry equity returns to an oil price factor and the sensitivity to
market returns. Their results show a signicant positive sensitivity in
the oil and gas and a negative sensitivity in paper and packaging, and
transport industries.
Faff and Chan (1998) meanwhile, empirically test the returns on
gold stocks in the Australian equity market over the period 1979
1992. They nd that there are just two factors explaining the returns
on gold stock, namely, a market factor and a gold price factors.
Jones and Kaul (1996) however report that the reaction of the
United States and Canadian stock markets to oil price shocks can be
entirely explained on the basis of changes in the expected value of
future real cash ows. This evidence holds but is weaker for the UK
and Japan. Jorion (1990) on the other hand, nds that the degree of
exposure of US multinationals to be positively and reliably correlated
with the degree of foreign involvement.

where t is a white noise error term and yt is locally detrended data

process under local alternative of given by:

3. Data and econometric methodology

yt = yt zt

The data were gathered from the International Financial Statistics

of the International Monetary Fund and consist of monthly observations from May 1988 to January 2005 of the Korean stock market
index, industrial production, interest rates and oil prices. The study
thus thoroughly covers the period of the Asian Financial Crisis of 1997,
which heavily affected the country,10 and the oil price hikes in the
early 1990s after Gulf War. The data collected allowed us to calculate
the real stock returns and the volatility of oil prices. In the results we
report industrial production as ip, real stock returns as rsr, interest
rates as r, oil prices as lo, and oil price volatility as rvol.
In this paper we use a VAR model to explain the impact of oil price
changes and volatility on real stock returns, industrial production and
interest rates. This methodological framework allows us test the
endogeneity of all remaining variables when oil price shocks are
introduced as exogenous variables. We use two models to test our
dynamics. The rst model uses oil prices, while the second uses their
volatility. We test the reaction of industrial production, real stock
returns and interest rates when oil price movements and percentage
gains or drops in oil price are introduced in the model. We employ
various diagnostic tests to verify linear or non-linear effects of crude
oil prices and crude oil price shocks on the stock market. One such
approach is a bivariate Markov switching model on the residuals to
see if the speed of adjustment is asymmetric.

where zt = (1 t)' and is the regression coefcient of yt and zt for

which:

4.2. Modied DF-GLS test

In this paper, instead of the standard Augmented Dickey-Fuller test
we use the modied DickeyFuller test (DF-GLS) developed by Elliott
et al. (1995). This test is conducted using the following regression:

1Lyt1 = a0 yt1 + aj 1Lytj + t

j=1

y1 ; y2 ; ; yT = y1 ; 1Ly2 ; ; 1LyT

z1 ; z2 ; ; zT = z1 ; 1Lz2 ; ; 1LzT

The t-test of the hypothesis H0: a0 = 0 against H0: a0 b 0 gives the

ADF-GLS test statistic. Elliott et al. (1995) recommended that the
parameter c, which denes the local alternative by
=1+

c
T

be set equal to 13.5. This test can attain a signicant gain in power
over the traditional unit root tests. The critical values for Elliott et al.
(1995) in Table 1 of Appendix A are estimated from Monte Carlo
simulations. For nite sample correlations, Cheung and Lai (1995)
provide approximate critical values. In the non-deterministic case, the
use of c = 7 is recommended where the test DF-GLS basically
involves the same procedure as computing the DF-GLS test, apart
from the exception that the locally detrended process series (yt ) is
replaced by the locally demeaned series (yt ) and zt = 1. The
asymptotic distribution of the DF-GLS test is the same as that of the
conventional DF test.

4. Model specication and empirical results

4.3. Condence interval for the largest autoregressive root
4.1. Unit root tests
It is standard in the literature on VAR modelling to employ a series
of unit root tests to ensure our variables are I(1). To verify the order of
integration of the variables we test for unit root based on Perron
(1988), Phillips (1987), Phillips and Perron (1988) and Kwiatkowski
et al. (1992). The semi-parametric PhillipsPeron (PP) tests devel10
South Korea was the country most hit by nancial crisis together with Indonesia
and Thailand. The $170.9 billion fall in 1998 was equal to 33.1% of the 1997 GDP, the
exchange rate passed by 850 Won/ US$ of June 1997 to 1,290 Won/ US$ in July 1998
(34.1%).

ADF tests indicate the presence of a unit root in each series since
for no series can the null of nonstationarity be rejected. To allow us to
measure how persistent the unit root in the process is, we also
calculate a condence interval (CI) due to Stock (1991), who suggests
that reporting CIs may provide useful information regarding sampling
uncertainty. The condence interval estimates tend to suggest that
the unit root is quite persistent with all lower bounds quite clearly
above 0.80 for both ADF () and ADF (). We also supplement these
results from Sims Bayesian unit root procedure which seem to be
suggestive of a unit root with high value of a. Furthermore, Geweke
and Porter Hudack (GPH) tests for fractional integration, also quite

978

R. Masih et al. / Energy Economics 33 (2011) 975986

uniformly suggest that most estimates of d fall signicantly in the

neighbourhood of 1. To check that these series are not integrated of
higher orders, we also repeat these tests using rst differences of
each series. These results suggest that they are all stationary after
applying the difference lter only once.11 Given the consistency and
the absence of ambiguity in results from all the mentioned testing
methods, we conclude that our variables are integrated at most order
one. This provides a requisite for the forthcoming multiple cointegration analysis.
4.4. Multivariate cointegration analysis
As OLS estimates of cointegrating vectors, particularly in small
samples, may be severely biased, in this analysis we employ the wellknown Johansen and Juselius (JJ) procedure of testing for the presence
of multiple cointegrating vectors. It is demonstrated in Johansen
(1992) that the procedure involves the identication of the rank of the
m by n matrix in the specication given by:
k1

Xt = + i Xt1 + Xtk + t
i=1

where, Xt is a column vector of the m variables.

Results of cointegration rank by the JJ procedure appear in Table 2
of Appendix A. Evidence from both traces and maximal eigenvalue
tests suggest that there is at most a single cointegrating vector or
analogously 7 independent common stochastic trends within this
four-variable system.12 This nding is consistent with studies by
Corhay et al. (1993), Leachman and Francis (1995) and Jeon and
Chiang (1991) who, among others, nd that equity markets of countries belonging to the G-7 countries possess at least one cointegrating
vector. It is worth noting that an implicit assumption underlying these
tests is that events over this period such as the Asian nancial crises
did not signicantly affect the stability of this system in terms of
altering the number of common stochastic trends between the macro
and nancial variables. This issue may now be tested using procedures
advocated in the literature by Hansen and Johansen (1993), and
Quintos and Phillips (1993). However, based on evidence using
similar techniques on a system of ve OECD equity markets, Masih
and Masih (1996c) nd evidence that the crash did not affect the
number of common stochastic trends within this particular system. In
this study we nd that the Asian nancial and banking crises had very
little effect on the stochastic trends between interest rates, industrial
production, oil prices, oil price volatility and real stock returns.
Besides observing these variables we have seen that South Korea was
one of better economies that withstood the pressures of currency

11
As a means of investigating the robustness of these results derived from
conducting tests for the total sample, we also undertake a sub-sample analysis of
these tests taking the October 1987 crash as the break point. In order to save space,
these results have not been reported for pre- and post-crash samples but available
upon request. Results, in general, indicate that the unit root approximation seems to
be quite robust to the October 1987 crash, since these sub-sample results do not
change our conclusion from conducting the tests over the full sample that these
variables are integrated of at most order one. Once again, it is important to warn
readers that such results are very much vulnerable to low power due to poor
performance in small samples.
12
Due to one of the biases of the JJ procedure being the sensitivity of cointegration
rank to the order of the lag length used in the VAR, We chose the lag subject to the
Akaikes FPE criterion. In addition, results of a unique cointegrating vector were
insensitive to slight modications to lag length. Furthermore, there has been much
recent work documenting the potential for severe small sample bias in Johansen tests
(see Cheung and Lai (1995)).The scaling-up factor on the asymptotic critical values
suggested by Cheung and Lai's study does not alter our conclusion of cointegration
rank. Furthermore, their study favors the trace test in that: it shows little bias in the
presence of either skewness or excess kurtosis, and is found to be more robust to both
skewness and kurtosis than the maximal eigenvalue test. (Cheung and Lai (1995,
p.324)). In the light of this statement, the trace statistic of 215.64, further conrms our
initial conclusion of r equal to at most 1.

meltdowns in Thailand and Indonesia and the severe banking crises in

Japan in the early nineties.
In order to assess the relative strength of the long run relationship,
Johansen and Juselius (1992) point out that larger eigenvalues
are associated with the cointegrating vector being more correlated
with the stationary component of the process. To gain some insight
into the robustness of results for all ve variables, we also conducted
cointegration tests revealing r = 1 at the 95% condence level for
both models. Eigenvalues, presented in Table 2 of Appendix A are in
descending order; indicate the cointegration relationship between the
variables.
Finally in order to test that each of the variables enters the
cointegrating vector signicantly, we test for zero restrictions upon
each of the coefcients derived by the Johansen procedure. Having
established the presence of a single cointegrating vector, the Johansen
procedure allows us to test several hypotheses on the coefcients by
way of imposing restrictions and likelihood ratio tests which are,
asymptotically, chi-square distributed with one degree of freedom.
Scrutinising the cointegration vector in each model presents us with
a measure of the most important component, in terms of its relative
weight, in comparison to the remaining components. Coefcient
estimates and signicance levels associated with the tests of zeroloading restrictions appear in Table 3 of Appendix A. Normalising on
interest rates in Korea provide evidence of each of these restrictions
being rejected, for most at least at the 10% level. This implies that most
of the variables enter into the cointegrating vector at a statistically
signicant level. However, although the weights of some of the
variables are not statistically signicant, we cannot exclude this from
the cointegrating vector as it forms a part of the long-run relationship.
In general, these results indicate that almost all variables adjust in a
signicant fashion to clear any short-run disequilibrium.
4.5. Short-run dynamics and long-run relations: Vector Error-Correction
Modelling (VECM)
Given the presence of a unique cointegrating vector in the ninedimensional VAR used in the JJ cointegration tests, this then provides
us with one error-correction term for constructing our models.
Analogously, we may also extract (n r) or four common trends, (for
such an approach see Kasa (1992), Chung and Liu (1994)).
Summary results in Tables 1 and 2 offer some interesting insights.
For each of the variables, at least one channel of Granger causality is
active: either the short-run through joint tests of lagged-differences
or a statistically signicant Error Correction Term (ECT). This latter
channel is a novelty of the VECM formulation but it is noteworthy of
signicance only in the rsr equation. The economic intuition arising
from this nding implies that when there is a deviation from the
equilibrium cointegrating relationships as measured by the ECTs, it is
mainly changes in the real stock returns that adjust to clear the
disequilibrium i.e. bears the brunt of short-run adjustment to longrun equilibrium. This leaves changes in interest rates, industrial
production and oil prices, which appear to be statistically exogenous
in both models and thus represents the initial receptor of any
exogenous shocks to their long-term equilibrium relationships.
Although the ECTs are not statistically signicant for variables
other than real stock returns, one cannot assume that all other
variables are non-causal since the short-run channels are still active.
For example, uctuations in interest rates seem to explain movements
in industrial production, and the exogenous oil price shocks seem to
cause the biggest uctuations in real stock returns. These short-run
causalities are explained by the signicance of lagged differences in
Tables 1 and 2.
During the period of our study, South Korea did not have any major
policy changes that had a direct impact on the variables in our model.
The various nancial sector reforms in the 1980s and 1990s were
across the board. In addition to the nancial crises in 1997/98 and the

R. Masih et al. / Energy Economics 33 (2011) 975986

Table 1
VEC model estimates using real oil prices.

979

Table 2
VEC model estimates using real oil volatility.

Equation

rt

rolt

ipt

rsrt

Equation

rt

rolvt

ipt

rsrt

1, t 1

0.019
(0.019)
0.899
(0.913)
0.296
(0.092)
0.084
(0.088)
1.096
(1.215)
-1.183
(1.208)
0.924
(1.514)
0.643
(1.472)
0.005
(0.015)
0.012
(0.011)
0.13
0.661
0.532
21.421
0.558
344.022
0.035

0.004
(0.001)
0.216
(0.068)
0.011
(0.007)
0.003
(0.007)
0.064
(0.090)
0.132
(0.090)
0.005
(0.113)
0.171
(0.109)
0.002
(0.001)
0.000
(0.000)
0.17
0.052
0.289
16.56
0.493
6336.890
10.381

0.000
(0.000)
0.041
(0.052)
0.001
(0.005)
0.003
(0.005)
0.141
(0.069)
0.000
(0.000)
0.280
(0.085)
0.208
(0.083)
0.000
(0.000)
0.000
(0.000)
0.15
0.042
0.168
14.666
5.134
32.301
30.191

1.019
(0.160)
47.990
(7.55)
0.579
(0.759)
0.440
(0.729)
15.303
(10.043)
14.406
(9.992)
17.004
(12.516)
4.130
(12.151)
0.116
(0.122)
0.098
(0.094)
0.43
6.760
3638.812
4.819
0.059
0.714
4.463

1, t 1

0.013
(0.012)
2.035
(1.983)
0.303
(0.094)
0.110
(0.092)
0.225
(3.195)
3.032
(3.342)
0.408
(1.600)
0.456
(1.526)
0.002
(0.014)
0.010
(0.011)
0.12
0.661
52.231
19.668
0.667
366.331
0.611

0.001
(0.000)
0.212
(0.054)
0.007
(0.003)
0.000
(0.003)
0.356
(0.087)
0.076
(0.091)
0.029
(0.043)
0.011
(0.041)
0.001
(0.000)
0.000
(0.000)
0.32
0.024
0.043
28.079
2.434
7507.398
1.592

0.001
(0.001)
0.239
(0.108)
0.001
(0.005)
0.000
(0.005)
0.503
(0.174)
0.129
(0.182)
0.307
(0.087)
0.233
(0.083)
0.001
(0.000)
0.001
(0.001)
0.21
0.041
0.148
21.212
4.851
32.053
25.134

0.626
(0.108)
99.474
(17.030)
0.388
(0.807)
0.910
(0.791)
65.539
(27.441)
10.258
(28.707)
19.242
(13.747)
5.077
(13.108)
0.055
(0.117)
0.085
(0.095)
0.39
6.834
3854.178
3.262
1.512
0.369
1.795

rt 1
rt 2
rolt 1
rolt 2
ipt 1
ipt 2
rsrt 1
rsrt 2
2

R
^

RSS
2SC[12]
2FF[1]
2NOR[2]
2HET[1]

rt 1
rt 2
rolvt 1
rolvt 2
lipt 1
lipt 2
rsrt 1
rsrt 2
2

R
^

RSS
2SC[12]
2FF[1]
2NOR[2]
2HET[1]

Notes: The underlying VAR model is of order 3 and contains unrestricted intercepts and
restricted trend coefcients. Lag order was selected by Schwarz Bayesian Criterion (SBC).
Standard errors are given in parenthesis. The diagnostics are chi-squared 2 (degrees of
freedom) statistics for serial correlation (SC), functional form misspecication (FF), nonnormal error terms (NOR) and heteroskedastic error variances (HET).

Notes: The underlying VAR model is of order 3 and contains unrestricted intercepts and
restricted trend coefcients. Lag order was selected by Schwarz Bayesian Criterion (SBC).
Standard errors are given in parenthesis. The diagnostics are chi-squared 2 (degrees of
freedom) statistics for serial correlation (SC), functional form misspecication (FF), nonnormal error terms (NOR) and heteroskedastic error variances (HET).

high oil prices leading to the Gulf War in 2000, South Korea did suffer
a decline in 1989 spurred by a sharp decrease in exports and foreign
orders which affected the industrial sector. Poor export performance
resulted from structural problems embedded in the nation's economy,
including an overly strong won, increased wages and high labour
costs, frequent strikes, and high interest rates. The high labour costs
saw industries investing heavily in automation and robotics technology which resulted in an increase in industrial products through the
1990s and in early 2000 period.
It is important to test the stability of the error correction mechanism when weakly exogenous factors are present in the model. In this
study crude oil prices, interest rates and industrial production are
found to be weakly exogenous and the stability of these in the error
correction model are tested using the CUSUM and CUSUMSQ tests
given in Figs. 1 and 2 of Appendix A. These tests are based on the null
hypothesis that the cointegrating vector is the same in every period;
the alternative is simply that it (or the disturbance variance) is not.
The test is quite general in that it does not require a prior specication
of when the structural change takes place. Since our study is not based
on a denitive piece of information, namely when the structural
change takes place it is preferred to the Chow test which only works
best when a denitive piece of information is at hand.
We test the stability of crude oil prices and interest rates after
estimating the VECM model of these variables as these were found to be
exogenous in our model and shocks originating from them could have
caused some instability in the system. Both the plots of the cumulative
sum of recursive residuals fall between the critical bounds and therefore
suggest that there is no cause for alarm in the structural stability of the
VECM. The CUSUM plot is around zero and more importantly lies
between the error bounds. The supporting CUSUMSQ which is deemed
more powerful than CUSUM also supports that crude oil prices and
interest rates did not cause instability in the cointegrating vector.
In addition to the above stability tests, we ran two bivariate cointegration tests between real stock returns and oil prices and real stock

returns and oil price volatility. Both models were found to be cointegrated and we tested for asymmetric effects in the error correction
mechanism by utilising a 2 regime Markov switching model. In
regime-switching models (like TAR, STAR and SETAR; see e.g. Franses
and van Dijk, 2000) the regimes and the switching mechanism are
explicitly dened through the threshold variable and the threshold
level. However, an upfront specication of this variable and level
is not a trivial task as the regime switches in our study are likely
to result from a combination of different fundamental drivers like
industrial production, interest rates real oil price, real oil volatility. On
the other hand, in Markov Regime Switching (MRS) models the
switching mechanism between the states is assumed to be governed
by an unobserved (latent) random variable. MRS models do not
require an upfront specication of the threshold variable and level
and, hence, are less prone to modelling risk. This gives them an
advantage in terms of parsimony.
A Markov-switching mean-adjusted auto regression can be written as:
Yt X St = A + A1 Yt1 X St  + A2 Yt2 X St  +


+ AN YtN X St  + t ; t eN 0; 2

where St = {1, 2} represents a two-regime state indicator variable (we

allow for 2 regimes in this case). The parameter X(St) can take two
different values, depending on the state which is prevailing in the
system.
The transition among states is assumed to be governed by a rstorder discrete Markov process. Denoting the pi, j the probability of
switching from regime i to regime j, the two-regime transition matrix
containing these transition probabilities can be written as:

P=

p11
p12

p21
p22


=

p11
1p11

1p22
p22


7

980

R. Masih et al. / Energy Economics 33 (2011) 975986

Table 3
Generalized variance decompositions.
Horizon

rt

Rol

Shock to interest rate (rt) explained by innovations in:

1
80.55
8.30
6
79.27
6.34
12
79.57
5.73
24
79.68
5.40

ip

rsr

0.83
1.01
1.07
1.11

10.31
13.38
13.63
13.82

Shock to economic activity (ipt) explained by innovations in:

1
0.64
1.44
96.81
6
3.00
1.61
92.28
12
3.45
1.51
91.52
24
3.73
1.45
91.01

1.11
3.11
3.52
3.80

Shock to real stock return (rsrt) explained by innovation in:

1
11.04
2.69
2.27
6
16.48
3.65
2.50
12
24.09
3.67
2.32
24
34.97
3.71
2.10

84.00
77.37
69.92
59.23

[The estimation below gives you a matrix of transition probabilities

which is the transpose of P above].
The state prevailing in the system in each point in time is inferred
with an efcient algorithm. Estimating the model above using a
maximum likelihood procedure presents some computation complications as the state and space equations for this model can only be
written as conditional of each other. To estimate the model Hamilton
(1989) proposed the use of an iterative expectation maximization
algorithm which converges to the maximum value of the loglikelihood function.
We rst test for asymmetry in the error correction mechanism
between oil price and real stock returns by choosing a MSM(2)-AR(3)
model, i.e. a model with 2 regimes and 3 lags. We select 2 regimes, as
any higher would not be justied for the number of observations as
they increase signicantly the number of parameters in the model
making it less parsimonious. In some cases we get a much worse AIC
and in other cases the algorithm fails to converge due to the lack of
observations for each regime. The number of lags has been selected by
considering AIC, HQ and SBC information criteria. The results are
presented in Table 4 of Appendix A.
Comparing the values of the log-likelihood for the estimated
model and its linear counterpart suggests that the Markov-switching
model does not produce any signicant improvement with respect to
the linear model. Further analysis of the results suggests that there is
no obvious specication error in the Markov-switching model.
Examining the regime classication in the graphs suggests that
there is about 50% of probability of being in either regime 1 or 2. This is
because 1 regime is associated with the positive residuals and the
other one is associated with the negative values. So, after allowing the
model to pick any signicant change in the residuals, the inference
about regimes classication shows no sign of structural change in the
context of the AR model presented above. We conclude that the
residuals show no evidence of nonlinear behaviour that could be
incorporated into the model. Further evidence is provided in Table 5
and Fig. 3 of Appendix A.
We then estimated the same model on the bivariate relationship
between oil price volatility and stock returns and conclude that the
error correction mechanism is best represented by a linear model for
the period during the East Asian nancial crises. A summary of the
results are given in Tables 6 and 7 and Fig. 4 of Appendix A.

4.6. Generalized variance decomposition analysis

The relative strength of the Granger-causal chain amongst the
variables beyond the sample period is given by the Variance decompositions (VDCs). VDCs provide a literal breakdown of the change in

value of the variable in a given period arising from changes in the

same variable in addition to other variables in previous periods. A
variable that is optimally forecast from its own lagged values will have
all its forecast error variance accounted for by its own disturbances
(Sims, 1982).
The variance decompositions presented in Table 3 indicate that
80.55% of shocks to interest rates are self-explained in the rst month.
This weighing stays at around 79% for 6, 12 and 24 months, suggesting
that other variables inuence interest rates. We nd that stock returns
and oil prices have a greater inuence on the variance of interest rates
than industrial production. Economic theory makes the link between
industrial production and interest rates through an increase in
investment. An increase in investment results in an increase in
industrial production and then puts an increasing pressure on interest
rates. In South Korea's case industrial production has minimal impact
on interest rates and perhaps this indicates the country's capability to
enjoy high industrial growth and not worry too much about ination.
This explains why South Korea had double digit growth in the 1980s
and in early 1990s.
In Table 3 we also nd that industrial production is strongly
exogenous with only small inuences from interest rates and real
stock returns. The brunt of the variance in endogenous real stock
return variable is explained by movements in interest rates. In the rst
month, 11.04% of variance and in 6 months 16.48% variance in real
stock return is explained by changes in interest rates. This inuence
establishes a link between instruments of monetary policy and the
stock market. Movements in interest are indicative of the state of the
economy and this embeds expectations among investors. An increase
in variance of interest rates could suggest the direction of ination in
the economy which in turn reects whether economic activity has
picked up or slowed down. If interest rates are on the increase than
investors are likely to go easy on the stock market as the risk return
trade off in the bond market will become more attractive.
In Table 4 we get similar results except that oil price volatility
explains a greater proportion of variation in industrial production
than the level series of the oil price. Industrial production appears to
be less exogenous compared to Table 3 mainly because of the
uncertainty caused by oil price volatility. After 6 horizons only 82.50%
of shock is self-explained compared to 92.28 in Table 3. Oil price

Table 4
Generalized variance decompositions.
Horizon

rt

rolv

Shock to interest rate (rt) explained by innovations in:

1
82.84
6.79
6
78.92
7.52
12
78.28
7.44
24
77.89
7.41

ip

rsr

1.97
2.14
2.15
2.17

8.39
11.43
12.13
12.53

Shock to economic activity (ipt) explained by innovations in:

1
2.39
5.83
91.37
6
6.57
10.60
82.50
12
7.34
11.30
81.07
24
7.84
11.75
80.14

0.41
0.33
0.29
0.27

Shock to real stock return (rsrt) explained by innovation in:

1
7.87
1.98
0.44
6
14.45
12.36
2.87
12
19.60
17.87
2.64
24
24.43
23.32
2.45

89.71
70.33
59.88
49.81

Notes: the underlying cointegrated VAR model is of order 3, contains unrestricted

intercepts, and restricted trend coefcients. Lag order was selected using Schwarz
Bayesian Criterion (SBC). Standard errors generated from 10,000 replications are
presented in parenthesis. In Table 3 we cannot obtain VDCs for oil price because it is
introduced as an exogenous variable. In Table 4 we cannot obtain VDCs for oil price
volatility as it is introduced as an exogenous variable. We do capture the out of sample
dynamics in the subsequent impulse responses.

R. Masih et al. / Energy Economics 33 (2011) 975986

volatility explains 10.60%, while interest rates explain 6.57% of the

innovations in industrial production. The endogeneity of the real stock
returns are illustrated by the strong causal links between interest
rates and oil price volatility, with 14.45% and 12.36% of shocks in stock
returns being explained by the former and later.

4.7. Impulse response functions

The information contained in the VDCs can be equivalently
represented by graphs of the impulse response functions (IRFs). IRFs
essentially map out the dynamic response path of a variable due to a
one-period standard deviation shock to another variable. The IRFs
become crucial in our analysis of oil price shocks, as VDCs cannot be
generated for exogenous variables.
Lee et al. (1992), Pesaran et al. (1993) and Lee and Pesaran (1993)
discuss why generalized impulse responses are preferred to orthogonalized response functions. Information from application of these
tools should provide some further evidence on the patterns of linkages amongst stock markets and oil price shocks, as well as contribute
to enhancing our insights upon how other macroeconomic variables
react to system wide shocks and how these responses propagate
over time. It is important to note, however, that although derivation
of GIRFs does not suffer from the arbitrary orthogonalizations of
innovations, GIRFs should not be strictly used to isolate responses of a
particular shock, assuming that all other shocks are not present, or not
also running in conjunction with the particular shock in question. In
this respect, one should not attribute the shock, as in traditional IRF
analysis, to sole variables in the system, and thereby practice caution
when interpreting such results.
Generalized IRFs from one-standard deviation shocks to the model
using log of oil price and oil price volatility in Korea are traced out for
each individual variable in Figs. 14 (including the own shock to each
market). In general the responses show long-lasting effects and the
variables take about 6 to 7 months to return to a new equilibrium
level. Of all the variables interest rates seem to act positively in
responding to their own shocks and shocks to oil price and oil price
volatility. Shocks to industrial production and real stock returns seem
to have no upward pressure on interest rates. This shows that interest
rates in South Korea are purely driven up by expectations embedded
in interest rates (perhaps long-term interest rates) and through
shocks in oil prices. We can see that industrial production is more
susceptible to shocks in the stock market and it takes about 11 months
to reach to a new equilibrium state. The self-reactionary prole of oil
price and oil price volatility seem to settle back to their pre-shock
levels the quickest which is not surprising given the conclusions from
the within-sample causality VECM and VAR results. Responses of the
stock market are interesting when interest rates and oil prices are
shocked. In both instances the stock market increases and then reverts
back in the negative territory to its long-run level after about
9 months. This shows the lag effect of interest rates and oil price
have on stock market activity and thus shows that the Korean stock
markets is of strong character in the short term. The identical reactionary proles of real stock returns in Fig. 1 also suggest that inationary expectations are evident through oil prices and through
movements in short term interest rates.
In summary, impulse responses of the two models are very similar
in nature indicate that volatility of oil price and log of oil price have
the same impact on the Korean economy. The IRFs show that the
Korean economy is not affected adversely by oil price shocks any
differently to normal oil price movements. The long-run time path of
real stock returns in Figs. 1 and 3 when ROLV and ROL are shocked
indicate a bigger impact from oil price volatility then from the level
series. The new equilibrium for the stock market settles at a higher
negative standard deviation level (1.2) then through the impacts of
the level of oil prices.

981

5. Main ndings and conclusion

Few studies have thus far analysed the effects of oil price
uctuations on the economic and nancial variables of net importing
countries of oil. This study tries to capture the stochastic properties
and long run dynamics between the macro economy, the stock
markets, the instruments of monetary policy and the oil price
movements.
The rst result is that a long run equilibrium relationship does
exist among the four variables considered in the study. Furthermore
the system is stable throughout the entire period that is analyzed. This
means that the nancial crisis did not affect the stability of the system,
and banking has just a low effect on the stochastic trends between
interest rate, industrial production, oil prices and oil price volatility and
real stock returns. Interest rate and oil production appear to be
exogenous variables in both our models. In other words they are the
rst receptors of any external shocks. The main conclusion of our
research is that oil price movements signicantly affect the stock market.
Our analysis indicates that real stock returns are the main channel of
short-run adjustment to long-run equilibrium. After shocking oil prices
and oil price volatility (with a bigger effect in the second case), the stock
market increases and then slows down, recovering to its long-run
equilibrium level after a period of approximately 9 months. During this
period interest rates and oil prices spread their effect on the stock
market. This conclusion conrms the linkage between real economy
shocks, instruments of monetary policy and stock markets.
Results from our VECM suggest that real stock returns are the main
channel of short-run adjustment to long-run equilibrium. Oil price
shocks have two different negative effects on rm protability. First, it
has a direct negative effect because it increases the production costs of
rms. And secondly, it has an indirect negative effect because investors foresee the decline in prot margins of rms and make decisions
that affect the stock market indexes (i.e. selling shares).
For years, South Korea was one of Asia's fastest-growing economies,
but the country's rapid industrialization was accompanied by a
corresponding increase in its energy consumption. South Korea's rapid
industrialization process over the past several decades has resulted in
the country's industrial sector energy consumption increasing by more
than 300%, from 1.0 quads in 1985 to 4.2 quads just over a decade later.13
Since high and volatile oil prices can have signicant adverse
spillovers for the economy at large, there is in principle an argument
for government intervention to reduce volatility. Three areas appear
to be worthwhile for taking into consideration. First, governments of
oil-importing countries could benet from increasing their strategic
oil reserves, and thus protect themselves from the risk of supply
disruptions. Second, governments of oil-importing countries should
consider oil-saving measures very carefully. These measures include
policies to improve energy efciency, promote energy conservation
and use of alternative fuels (i.e. coal, natural gas and renewable
energy). And nally, oil-importing countries should enhance dialogue
with oil-exporting countries in order to increase multilateral
cooperation and to minimize shocks that have an adverse effect on
the national economy.

Acknowledgements
We would like to thank the two anonymous referees for their
patience, comments and suggestions that greatly improved the
paper. We are very grateful for Julian Inchauspe for research
assistance.
The views expressed in this paper are those of the authors and not
necessarily shared by JP Morgan.

13

Data from Energy Information Administration (EIA), US Department of Energy.

982

Generalized Impluse Response Paths when R is shocked

Response of R
1

Response of ROLV

0.9

Response of LIP

0.016

0.8
0.014
0.7

0.00

-0.2

0.00

-0.4

0.00

-0.6

0.012
0.6

0.00
0.01
0.008

0.4

-1.0

-0.01

0.3

0.006

0.2

0.004

-0.01

0.1

0.002

-0.01

-0.01

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31

-0.8

-0.01

-1.2

-0.01

9 11 13 15 17 19 21 23 25 27 29 31

-1.4
-1.6
-1.8
1

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31

9 11 13 15 17 19 21 23 25 27 29 31

Generalized Impluse Response Paths when ROLV is shocked

Response of R
0.4
0.3

Response of ROLV

Response of LIP

Response of RSR

0.03

0.00E+00

0.025

-2.00E-03

-0.2

-4.00E-03

-0.4

0.3
0.02
0.2

-0.6

-6.00E-03
0.015

-0.8

0.2

-8.00E-03
-1

0.01

0.1

-1.00E-02

0.1

0.005

-1.20E-02

0.0

-1.40E-02

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31

-1.2
-1.4
-1.6
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31

Fig. 1. Generalized impulse response function. Notes: The horizontal axis refers to months after shock. The vertical axis refers to standard deviations. Charts provide generalized impulse response functions (GIRF) of all variables in our model
when interest rates (R) and oil volatility (ROLV) are shocked. Dashed lines represent single standard error bounds around the point estimates. We can compare the above to Fig. 3.

R. Masih et al. / Energy Economics 33 (2011) 975986

0.5

Response of RSR
0.0

0.00

0.018

Generalized Impluse Response Paths when LIP is shocked

Response of R
0

Response of ROLV

Response of LIP

Response of RSR
0.6

0.05

0.00E+00

-0.02

0.4

0.04
-1.00E-03

-0.04

0.2

0.04
-2.00E-03

-0.06
-0.08

0.0

-3.00E-03

0.03

-0.2

-4.00E-03

0.02

-0.4

0.02

-0.6

0.01

-0.8

0.01

-1.0

-0.12
-5.00E-03

-0.14
-6.00E-03

-0.16
-0.18

-7.00E-03

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31

0.00
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31

-1.2

9 11 13 15 17 19 21 23 25 27 29 31

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31

Generalized Impluse Response Paths when RSR is shocked

Response of R
0.0

Response of ROLV
0

-0.1

0.002
5

0.0015

-0.2
-0.01

-0.2
-0.3

Response of RSR
7

0.0025
-0.005

-0.1

-0.015

0.001

0.0005

-0.3

-0.0005

-0.4

-0.02

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31

-0.025

-0.001

-0.4
-0.5

Response of LIP
0.003

R. Masih et al. / Energy Economics 33 (2011) 975986

-0.1

0.03

9 11 13 15 17 19 21 23 25 27 29 31

-0.0015

-0.002

-1

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31

Fig. 2. Generalized impulse response functions (Continued). Notes: The horizontal axis refers to months after shock. The vertical axis refers to standard deviations. Charts provide generalized impulse response functions (GIRF) of all variables
in our model when log of industrial production (LIP) and real stock returns (RSR) are shocked. Dashed lines represent single standard error bounds around the point estimates. We can compare the above to Fig. 4.

983

984

Generalized Impluse Response Paths when R is shocked

Response of R
1

Response of ROL

Response of LIP

Response of RSR
0.0

0.035

0.00

0.03

0.00

0.025

0.00

0.02

0.00

0.015

0.00

-1.2

0.01

-0.01

-1.4

0.005

-0.01

-0.2

0.9
0.8
0.7

-0.4
-0.6
-0.8

0.6

0.4
0.3

-1.6

0.2
0.1
0

-0.01

0
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31

-1.8

-2.0

9 11 13 15 17 19 21 23 25 27 29 31

9 11 13 15 17 19 21 23 25 27 29 31

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31

Generalized Impluse Response Paths when ROL is shocked

Response of R

Response of ROL

0.4

0.08

0.3

0.07

Response of LIP
0.006
0.005

0.4

0.004

0.3

0.2

0.06

0.003

0.05

0.002

0.001

-0.2

-0.4

0.2
0.04
0.2
0.1

0.03

-0.001

0.02

-0.002

-0.6
-0.8

-0.003

0.1

0.01

0.0

0
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31

Response of RSR
0.6

9 11 13 15 17 19 21 23 25 27 29 31

-0.004

-1

-0.005

-1.2
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31

Fig. 3. Generalized impulse response functions (Continued). Notes: The horizontal axis refers to months after shock. The vertical axis refers to standard deviations. Charts provide generalized impulse response functions (GIRF) of all variables
in our model when interest rates (R) and oil prices (ROL) are shocked. Dashed lines represent single standard error bounds around the point estimates. We can compare the above to Fig. 1.

R. Masih et al. / Energy Economics 33 (2011) 975986

-1.0

0.5

Generalized Impluse Response Paths when LIP is shocked

Response of R
0
-0.02
-0.04

-0.08
-0.1
-0.12

Response of LIP
0.05

0.008

0.04

0.007

0.04

0.006

0.03

0.005

0.03

0.004

0.02

0.003

0.02

0.002

0.01

0.001

0.01

0.6
0.4
0.2
0.0
-0.2

0.00

0
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31

Response of RSR
0.8

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31

-0.4
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31

Generalized Impluse Response Paths when RSR is shocked

Response of R
0.0

Response of LO
0

-0.1

6
5
4

0.001

-0.01

-0.2

0.0005

-0.3

-0.015

-0.3

-0.0005

-0.4

-0.02

-0.4
-0.5

0.0025
0.0015

-0.2

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31

-0.025

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31

Response of RSR
7

0.002

-0.005

-0.1

Response of LIP
0.003

-0.001

-0.0015

-1

-0.002

R. Masih et al. / Energy Economics 33 (2011) 975986

-0.06

Response of LO
0.009

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31

-2

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31

Fig. 4. Generalized impulse response functions (Continued). Notes: The horizontal axis refers to months after shock. The vertical axis refers to standard deviations. Charts provide generalized impulse response functions (GIRF) of all variables
in our model when log of industrial production (LIP) and oil prices (LO) are shocked. Dashed lines represent single standard error bounds around the point estimates. We can compare the above to Fig. 2.

985

986

R. Masih et al. / Energy Economics 33 (2011) 975986

Appendix A. Supplementary data

Supplementary data to this article can be found online at
doi:10.1016/j.eneco.2011.03.015.
References
Bernanke, B.S., Gertler, M., Watson, M., 1997. Systematic monetary policy and the
effects of oil price shocks. Brookings Papers on Economic Activity 1997 (1), 91142.
Bohi, D.R., 1989. Energy price shocks and macroeconomic performance. Resources for
the Future, Washington D.C.
BP, 2004. Statistical review of world energy June 2004. www.bp.com/statisticalreview
2004.
Chaudhuri, K., Daniel, B.C., 1998. Long-run equilibrium real exchange rates and oil
prices. Economics Letters 58, 231238.
Cheung, Y., Lai, K., 1995. Lag order and critical values of the augmented DickeyFuller
test. Journal of Business and Economic Statistics 13, 277280.
Chung, P.J., Liu, J.D., 1994. Common stochastic trends in Pacic Rim stock market. The
Quarterly Review of Economics and Finance 34, 241259.
Corhay, A., Tourani, A., Urbain, J.-P., 1993. Common stochastic trends in European stock
markets. Economic Letters 42, 380385.
DeJong, D., Nankervis, J., Savin, N.E., Whiteman, C.H., 1992. The power problems of unit
root tests in time series with autoregressive errors. Journal of Econometrics 53,
323343.
Dickey, D.A., Fuller, W.A., 1981. Likelihood ratio statistics for autoregressive time series
with a Unit Root. Econometrica 49, 10571072.
Driesprong, G., Jacobsen, B., Maat, B., 2003. Striking oil: another puzzle. Research paper
ERS-2003-082-F&A. Erasmus Research Institute of Management (ERIM.
Elliott, G., Rothenberg, T.J., Stock, J.H., 1995. Efcient tests for an autoregressive unit
root, 1996. Econometrica 64, 813836.
Faff, R., Brailsford, T.J., 1999. Oil price risk and the Australian stock market. Journal of
Energy Finance and Development 49, 6987.
Faff, R., Chan, H., 1998. A multifactor model of gold industry stock returns: evidence
from the Australian equity market. Applied Financial Economics 8, 2128.
Franses, P.H., van Dijk, D., 2000. Non-linear time series models in empirical nance.
Cambridge University Press, United Kingdom.
Glasure, Y.U., 2002. Energy and national income in Korea: further evidence on the role
of emitted variables. Energy Economics 24, 355365.
Greene, D.L., Jones, D.W., Leiby, P.N., 1998. The outlook of US oil dependence. Energy
Policy 26, 5569.
Hamilton, J.D., 1983. Oil and the macroeconomy since World War II. Journal of Political
Economy 91, 228248.
Hamilton, J.D., 1989. A new approach to the economic analysis of nonstationary time
series and the business cycle. Econometrica 57 (2), 357384.
Hansen, H., Johansen, S., 1993. Recursive estimation in cointegrated VAR-Models.
Institute of Mathematical Statistics Working paper 1. University of Copenhagen,
Copenhagen.
Heap, A., 2005. China_the engine of a commodities super cycle. Citigroup Global
Markets Paper, March 31.
Jeon, B., Chiang, T., 1991. A system of stock prices in world stock exchanges: common
stochastic trends for 19751990? Journal of Economics and Business 43, 329338.

Johansen, S., 1992. Determination of cointegration rank in the presence of a linear

trend. Oxford Bulletin of Economics and Statistics 54 (3), 383397.
Johansen, S., Juselius, K., 1992. Testing structural hypotheses in a multivariate cointegration analysis of PPP and the UIP for UK. Journal of Econometrics 53 (13), 211244.
Jones, C.M., Kaul, G., 1996. Oil and the stock markets. The Journal of Finance 51, 463491.
Jorion, P., 1990. The exchange-rate exposure of U.S. multinationals. The Journal of
Business 63, 331345.
Kaneko, T., Lee, B.S., 1995. Relative importance of economic factors in the U.S. and
Japanese stock markets. Journal of the Japanese and International Economies 9,
290307.
Kasa, K., 1992. Common Stochastic Trend in International Stock Markets. Journal of
Monetary Economics 29, 95124.
Kwiatkowski, D., Phillips, P.C.B., Schmidt, P., Shin, Y., 1992. Testing the null hypothesis
of stationarity against the alternative of a unit root: how sure are we that economic
time series have a unit root? Journal of Econometrics 54, 159178.
Leachman, L., Francis, B.B., 1995. Long run relations among the G-5 and G-7 equity
markets: evidence on the Plaza and Louvre Accords. Journal of Macroeconomics 17,
551579.
Lee, K.C., Pesaran, M.H., 1993. Persistence proles and business cycle uctuations in a
disaggregate model of U.K. output growth. Ricerche Economiche 47, 293322.
Lee, K.C., Pesaran, M.H., Pierse, R.G., 1992. Persistence of shocks and its sources in a
multisectoral model of UK output growth. The Economic Journal 102, 342356.
Masih, A.M.M., Masih, R., 1996. Energy consumption, real income and temporal
causality: results from a multi-country study based on cointegration and errorcorrection modelling techniques. Energy Economics 18, 165183.
Mork, K.A., Olsen, O., Mysen, H.T., 1994. Macroeconomic responses to oil price increases
and decreases in seven OECD countries. The Energy Journal 15, 1935.
Papaetrou, E., 2001. Oil Price shocks, stock market, economic activity and employment
in Greece. Energy Economics 23, 511532.
Perron, P., 1988. Trends and random walks in macroeconomic time series: further
evidence from a new approach. Journal of Economic Dynamics and Control 12,
297332.
Pesaran, M.H., Pierse, R.G., Lee, K.C., 1993. Persistence, cointegration and aggregation: a
disaggregated analysis of output uctuations in the U.S. economy. Journal of
Econometrics 56, 5788.
Phillips, P.C.B., 1987. Time series regression with unit roots. Econometrica 55, 277302.
Phillips, P.C.B., Perron, P., 1988. Testing for a unit root in time series regression.
Biometrika 75, 335346.
Quintos, C.E., Phillips, P.C.B., 1993. Parameter constancy in cointegrating regressions.
Empirical Economics 18, 675706.
Radetski, M., 2006. The anatomy of three commodity booms. Resources Policy 31,
5664.
Sadorsky, P., 1999. Oil price shocks and stock market activity. Energy Economics 21,
449469.
Sadorsky, P., 2001. Risk factors in stock returns of Canadian oil and gas companies.
Energy Economy 23, 1728.
Schwert, G.W., 1987. Effects of model specication on tests for unit roots in
macroeconomic data. Journal of Monetary Economics 20, 73103.
Sims, C.A., 1982. Policy analysis with econometric models. Brookings Papers on
Economic Activity 1, 107152.
Stevens, P., 2005. Oil markets. Oxford Review of Economic Policy 21, 1942.
Stock, J.H., 1991. Condence intervals for the largest autoregressive root in U.S.
macroeconomic time series. Journal of Monetary Economics 28, 435459.