Applied Economics

ISSN: 0003-6846 (Print) 1466-4283 (Online) Journal homepage: https://www.tandfonline.com/loi/raec20

Impact of oil price shocks on output, inflation and

the real exchange rate: evidence from selected
ASEAN countries

Hem C. Basnet & Kamal P. Upadhyaya

To cite this article: Hem C. Basnet & Kamal P. Upadhyaya (2015) Impact of oil price shocks on
output, inflation and the real exchange rate: evidence from selected ASEAN countries, Applied
Economics, 47:29, 3078-3091, DOI: 10.1080/00036846.2015.1011322

To link to this article: https://doi.org/10.1080/00036846.2015.1011322

Published online: 12 Feb 2015.

Submit your article to this journal

Article views: 1233

View Crossmark data

Citing articles: 14 View citing articles

Full Terms & Conditions of access and use can be found at

https://www.tandfonline.com/action/journalInformation?journalCode=raec20
Applied Economics, 2015
Vol. 47, No. 29, 3078–3091, http://dx.doi.org/10.1080/00036846.2015.1011322

Impact of oil price shocks on

output, inflation and the real
exchange rate: evidence from
selected ASEAN countries
Hem C. Basneta and Kamal P. Upadhyayab,*
a
Chadron State College, Chadron, NE 69337, USA
b
Department of Economics, University of New Haven, West Haven, CT
06516, USA

This article analyses the impact of oil price shocks on real output, inflation
and the real exchange rate in Thailand, Malaysia, Singapore, the Philippines
and Indonesia (ASEAN-5) using a Structural VAR model. The cointegra-
tion tests indicate that the macroeconomic variables of these countries are
cointegrated and share common trends in the long run. The impulse
response functions reveal that oil price fluctuations do not impact the
ASEAN-5 economies in the long run and much of its effect is absorbed
within five to six quarters. The variance decomposition results further assert
that with a few exceptions oil price shocks do not explain a significant
variation in any of the variables under consideration. We also identify a very
unique pattern of response to oil price fluctuations between Malaysia and
Singapore and between the Philippines and Thailand. The pairs exhibit a
high degree of similarity in their responses; they do not share any common-
alities across the group.

Keywords: ASEAN-5; oil price; inflation; exchange rate; real GDP

JEL Classification: C32; E30; F41

I. Introduction disruption in oil supplies began. This led to a start

in oil price rise, which continued until it peaked to
Oil has been a major source of energy since last $106.36 (inflation adjusted) per barrel in 1980. With
century. The importance of oil as a source of energy the election of Reagan as the President of the United
has increased significantly after World War II, parti- States and the end of the Iranian hostage crisis, it
cularly with the increased use of automobiles as the however started declining. The start of the oil price
main source of transportation. Until the mid-1970s, decline began in 1981, which continued for several
the international price of petroleum was more or less years. This continued decline in the oil price can be
stable. With the Iranian crisis in 1979, a major attributed to several factors. First, in the early 1980s,

*Corresponding author. E-mail: Kupadhyaya@newhaven.edu

3078 © 2015 Taylor & Francis

Impact of oil price shocks on output, inflation and the RER 3079
most industrialized nations went through the eco- affect the heating as well as manufacturing costs.
nomic slowdown, which caused for a decrease in Increase in production costs can lead to a fall in
demand for oil. Second, because of high oil price output and ultimately a decrease in aggregate supply
for a long period businesses and firms adopted a in the economy. The demand side effect stems from
more energy-efficient operation. Third, during the consumption and investment. Fluctuations in oil
same time small Japanese cars became popular in prices directly influence the consumption and expen-
the US auto market as they were more fuel efficient diture patterns of households. Since the price elasti-
that resulted in replacing the gas guzzling big cars. city of demand for oil is generally low, higher oil
All these factors reduced the demand for oil signifi- prices may force the households to cut down on their
cantly, forcing the market to lower the price. expenditures on other goods and services, which lead
Additionally, during the same period the supply of to lower the demand for those goods and services. In
oil increased because of the finding of new oil this regard, Sill (2007) suggests that higher oil prices
reserves such as North Sea in Europe. A combination can reduce demand for other goods because they
of decrease in demand and an increase in supply reduce wealth and at the same time inducing uncer-
dramatically reduced the world price of petroleum tainty about the future. Fernald and Trehan (2005)
all the way down to $17.10 (inflation adjusted) per argue that higher oil prices are like an increase in tax
barrel in 1998. on consumers because the extra payment that con-
The price of oil, however, started rising beginning sumers make to foreign oil producers can no longer
1999, which peaked in the summer of 2008. The be spent on other kinds of consumption goods.
average price of crude oil reached $99.07 per barrel Bernanke (1983) notes that higher oil prices will
(inflation adjusted) in 2008. There are several reasons have an adverse effect on short-run economic perfor-
for this increase. First, in emerging economies the mance precisely because of their ability to disrupt
use of modern modes of transportation such as auto- purchases of large ticket consumption and invest-
mobiles is increasing, which increases the demand ment goods.
for oil. Second, the increase in industrialization in the Another variable that oil price can affect is the
emerging as well as developing countries has put overall price level in the economy. As mentioned
upward pressure on oil demand. Third, even in the above, oil is widely used as an input for production
developing countries the use of modern modes of as well as for distribution of the goods and services.
transportation has increased significantly and at the An increase in the price of oil, therefore, can increase
same time their primary economic sector, agriculture, the production as well as the distribution cost of goods
has become more mechanized, which also raises the and services. This eventually feeds into the price level
demand for oil. These are some of the important in the economy, leading to an increase in the inflation
factors that have helped increase oil price in recent rate. An increase in oil prices not only affects the
period. For the last several years, it has been hovering output and the prices in an economy, it also affects
around $100 per barrel, which is significantly higher the currency exchange rate of a country (see Cebula
than the normal price level of around $20 per barrel and Michael Frewer, 1980; Hamilton, 1996; Amano
until the late 1990s. and Norden, 1998; Cebula, 2000; Issa et al., 2008). It
A change in oil price affects different macroeco- is well known that the standard theory of exchange
nomic variables in an economy. The transmission rate determination suggests that an increase in the
mechanism through which the oil price impacts the oil price causes the currency of an oil-exporting
real economy includes both the demand and the country to appreciate as the demand for its currency
supply channels. The supply side effects are related increases in the foreign exchange market. And con-
to the firm’s cost of production. For instance, an versely, an increase in oil price depreciates the cur-
increase in oil price raises production costs because rency of an oil-importing country because the
crude oil is used as a key intermediate input in the supply of its domestic currency in the foreign
production processes. This is more importantly true exchange market increases.
in case of the goods made with petroleum products. There is a considerable amount of empirical litera-
Oil price increase also raises the distribution costs of ture that studies the effect of oil price shocks on
the products of the firms via an increase in transpor- different macroeconomic variables (e.g. Hamilton,
tation cost. In addition, oil prices also indirectly 1996, 2000; Hooker, 1996, 1999; Huntington,
3080 H. C. Basnet and K. P. Upadhyaya
1998). In a seminal paper, Hamilton (1983) docu- Most of the studies mentioned above are either
mented a significant systematic relationship between based on the data from industrialized nations or
oil prices and output in the US economy. He attrib- oil-exporting nations. There is a dearth of similar
uted oil price shocks to be the major contributing studies in the context of ASEAN nations. One of
factors in 9 out of 10 US recessions after World War the studies related to ASEAN countries is by Chang
II. The similar findings to that of Hamilton (1983) and Wong (2003). In their study of Singapore econ-
with respect to the impact of oil price shocks on omy, they find an adverse impact of an oil price shock
macroeconomic variables can also be found in on Singapore’s macroeconomic performance. But the
Cebula et al. (2002), Cunado and Garcia (2005), impact is marginal. Rafiq et al. (2009) examined the
Cologni and Manera (2008) and Kilian (2008). In a impact of oil price volatility on key macroeconomic
study of eight developed countries, Backus and variables in Thailand. Their empirical findings sug-
Crucini (2000) show that a large part in the variabil- gest that, in most of the cases, oil price volatility has
ity of the terms of trade is associated with extreme an impact in the short run only and most notably on
movements in oil prices. Their findings suggest that investment and the unemployment rate. Another
the increased volatility in the terms of trade since the study is by Ahmed and Wadud (2011) for Malaysia.
Bretton Woods system seems largely due to the Their findings suggest an asymmetric effect of oil
increased fluctuations of oil price rather than the price shocks on Malaysian industrial production and
volatility of nominal or real exchange rates (RERs). inflation. Their variance decomposition analysis
Jiménez-Rodríguez and Sánchez (2005) assess the reconfirms that volatility in the oil price is the second
effect of oil price shocks on real economic activity of most important factor to explain the variance of
the major industrialized countries (a mix of oil industrial production after its own shocks.1 This find-
exporting and importing). Their findings report that ing is consistent with Mehrara (2008), who report a
the real GDP growth of an oil-importing country is nonlinear and asymmetric relationship between oil
negatively affected by any increases in oil price. With prices and economic growth of oil-exporting coun-
regards to the oil-exporting countries, the effect, tries. Cunado and Garcia (2005) measured the impact
however, appears to be ambiguous. Applying VAR of oil price shocks on economic activity and CPIs for
approach, Farzanega and Markwardt (2009) analyse six Asian countries and documented that the impact
the dynamic relationship between oil price shocks was mainly limited to the short run.
and major macroeconomic variables in the Iranian The purpose of this study is to estimate and ana-
economy (an oil-exporting country). They find a lyse the effect of oil price shock on output, inflation
positive relationship between oil price increases and and the exchange rates of the major ASEAN coun-
industrial output growth. But they also detected an tries – Indonesia, Malaysia, the Philippines,
inflationary effect and an evidence of the Dutch Singapore and Thailand (ASEAN-5) using quarterly
Disease because of the appreciation of the domestic time series data from 1970 to 2010. ASEAN region is
currency. one of the fastest growing regions in the world. It will
There are some studies that have investigated the require a sustained increase in the energy supply to
sources of macroeconomic fluctuations in the context maintain its growing economy. Even though the
of purely oil-exporting countries and found that oil ASEAN region controls almost 40% of the entire
price shocks are the main source of output fluctua- oil and gas reserve of the Asia Pacific region,
tions (see Mehrara and Oskoui, 2007; El Anshasy Singapore, Thailand and the Philippines are still
and Bradley, 2012). In contrast, Iwayemi and highly dependent on oil imports (www.asean.org/
Fowowe (2011) show that the effect of oil price resources). On the other hand, Indonesia ranks as
shocks on most of the Nigerian (an oil-exporting one of the world’s top liquid natural gas (an alter-
country) macroeconomic variables is miniscule. native to oil) producers and Malaysia is leading the

1
Park et al. (2011) classified 15 metropolitan cities and provinces of South Korea into 4 major regions (Capital, Central,
Honam and Gyeongsang) and analysed the effects of oil price fluctuations on the economy of these region. Their results
show that one region (Capital) is less affected by oil price fluctuations than the other three provincial regions.
Some studies examine the link between oil prices and stock market volatility among various countries (e.g. see Filis
et al., 2011; Arouri and Rault, 2012).
Impact of oil price shocks on output, inflation and the RER 3081
path of improving its energy technologies to enhance Johansen’s (1988) cointegration test in order to
energy efficiency. In 2010, each of these five coun- establish the long-run relationship between real
tries recorded an economic growth of more than 6%. GDP, inflation and RERs, which allows us to iden-
This indicates that these countries have already tify the number of common trends among the vari-
rebounded from the global economic crisis. In fact, ables. The SVAR model that we developed is as
with economic recovery underway across the region follows:
Singapore grew at a rate of 14% in 2010. According
to the OECD development centre’s 2010 Southeast A0 Xt ¼ A1 Xt1 þ . . . þ Ap Xtp þ εt (1)
Asian Economic Outlook report, the average growth
rate of six ASEAN (including Viet Nam) is expected
to grow at a rate of 6% over the 2011–2015 period. where Xt represents an (n × 1) vector of relevant
Given the spectacular economic growth of the region variables as follows: Xt ¼ ½ΔoilPt ; Δyt ; Δπ t ; ΔrExt 0 ,
in the recent years and its importance in the Asia Ai represents 4 × 4 matrices of coefficients for i = 0,
1…P and εt is the vector of structural disturbances
Pacific region, the purpose of this article is to esti-
y π rEx 0
mate and analyse the impact of oil price fluctuations denoted by εoilPt ; εt ; εt ; εt and they are assumed
on different macroeconomic variables in Thailand, to be independent and identically distributed (iid)
0
Malaysia, Indonesia, Singapore and the Philippines. (i.e. Eðεt Þ ¼ 0 & Eðεt εt Þ ¼ ε Þ. The coefficient of
Since the increase in oil price corresponds to negative the structural equation in Equation 1 can be
external shocks for oil-importing countries, this obtained through the estimated reduced form:
study measures how sensitive are these emerging
countries to oil price fluctuations over time. It is A0 Xt ¼ A1 Xt1 þ . . . þ Ap Xtp þ εt
expected that this study will provide some important
policy implications for the academicians, business Xt ¼ A1 1 1
0 A1 Xt1 þ . . . þ A0 Ap Xtp þ A0 εt
community, as well as the policymakers.2 Xt ¼ BðLÞXt þ μt
The rest of the article is organized as follows. (2)
Section II provides the methodology and the data
used in this research. Section III presents the empiri-
cal results and discussion. Finally, in Section IV, where BðLÞXt ¼ A01 A1 ðLÞ with A0 μt ¼ εt . The resi-
summary, conclusion and policy implications are dual μt is the reduced form VAR and assumed to be
reported. iid as well. A1 ðLÞ is a matrix of polynomial in the lag
operator (for detailed explanations of SVAR, see
Enders, 2004; Park et al., 2011).

II. Methodology and Data Identification and contemporaneous restrictions

Since this study does not intend to test any particular
The primary purpose of this article is to estimate the model, we base our identification restrictions on dif-
impact of oil price shocks on real GDP, the ferent economic theories and intuitions. We employ
exchange rate and the price level of the countries only short-run restrictions on the contemporaneous
under study. The theoretical link of oil price with relations because the structural VAR based on short-
output, inflation and the RER was already discussed run restrictions perform well (Christiano et al.,
in Section I. In order to estimate the effect of oil 2007). The short-run restrictions primarily control
price change (fluctuation) on these variables, we the contemporaneous feedback effects among the
develop a structural VAR (SVAR) model which variables in the model.
includes oil price (oilP) and three macroeconomic The identification in Equation 2 is obtained by
variables – output (y), inflation (π) and the RER imposing restrictions on contemporaneous relation
(rEx). Before estimating the model, we conduct among variables included in SVAR:

2
While the entirety of the article may be taken as general policy prescriptions for emerging economies around the world,
given some unique characteristics (see on page 10) of the ASEAN countries, the findings of this article may not be
applicable to other countries or regions.
3082 H. C. Basnet and K. P. Upadhyaya
2 32 3 2 3
1 0 0 0 uoilP 1 0 0 0 Estimation of Equations 4–6 allows us to measure the
6a 7 6 7 6 07 influence of oil price shock on the macroeconomic
6 21 1 0 0 76 u y 7 6 0 1 0 7
6 76 7 ¼6 7 fundamentals of the ASEAN countries more
4 a31 a32 1 0 54 u π 5 4 0 0 1 05 precisely.
a41 a42 a43 1 urEx 0 0 0 1
2 3 Data
εoilP
6 ε 7 This article uses quarterly data from 1970Q1 to
6 y 7
6 7 2010Q2 for the ASEAN-5 countries. Variables that
4 επ 5
enter into the VAR model are oil price, real output,
εrEx inflation and RERs. For oil price, we use the spot
market price of West Texas Intermediate (WTI) crude
The identification restrictions are motivated by the oil, which is considered as the benchmark for world
following economic reasoning: first, since each of the oil prices. The WTI crude oil price is measured in US
ASEAN-5 countries are small open economies, they dollar and is obtained from the US Energy
are price takers in the oil market. And because the price Information Agency website. The required series on
of oil is determined by the global demand and supply output (real GDP), constant in 2000 US dollars, are
conditions, the domestic output level, inflation rates obtained from the World Bank’s World Development
and the exchange rates in any of our sample country Indicators, 2012. With respect to Inflation, the CPI is
will have a negligible effect on it. Therefore, the oil used to measure each country’s inflation rate. RER is
price is assumed to be exogenous. Thus, the reduced defined as the nominal exchange rate times the US
error term for oil price can be expressed as follows: CPI divided by the domestic price index (i.e. CPI).
The data for CPIs and the nominal exchange rates are
uoilP ¼ εoilP (3) derived from International Financial Statistics (CD-
ROM) provided by IMF. All variables are expressed
in natural logarithmic form.
Equation 3 shows that the error term of international
oil price will be equal to its structural error term.
However, in case of other variables in our model, a
change in oil price can have a contemporaneous effect III. Empirical Results and Discussion
on them as an increase (decrease) in oil price can raise
(reduce) the cost of production because oil is used as Unit root and cointegration tests
one of the important inputs in the production as well as Macroeconomic time series data are usually nonsta-
the distribution of goods and services. Second, we tionary, and the use of nonstationary data produces
assume that the output does not respond contempor- spurious results (Nelson and Plosser, 1982). In order
aneously to any changes in the domestic variables to ensure the stationarity of the data series, ADF and
(second row in the identification matrix). Third, the Phillips–Perron test (1988) are conducted. The test
domestic price level is assumed to be influenced by all results are reported in Table 1. As seen in the table, all
the variables except the exchange rate. Finally, we do the data series are found nonstationary in the level.
not impose any restrictions with respect to exchange The CPI of Malaysia and Singapore, although
rate, suggesting that the exchange rate responds to appears stationary in the ADF and the Phillips–
changes in all variables. The reduced form error Perron tests, however, indicate that they are nonsta-
terms of the domestic variables (real GDP, inflation tionary. Since the Phillips–Perron tests are known to
and the RER) are as follows: be more robust than the ADF tests (Brooks, 2008),
we proceed with the analysis under the assumption
uy ¼ a21 uoilP þ εy (4) that oil prices are nonstationary in levels. After estab-
lishing the stationarity properties of the data, we
proceed with the examination of common trends
uπ ¼ a31 uoilP  a32 uy þ επ (5) among the variables using an unrestricted cointegra-
tion test. The lag-lengths of the unrestricted cointe-
urEx ¼ a41 uoilP  a42 uy  a43 þ εrEx (6) gration systems are determined by minimizing the
Impact of oil price shocks on output, inflation and the RER 3083
Table 1. Unit root tests 5 and 3 lags, respectively. The cointegration test
results are reported in Table 2.
ADF PP
As reported in Table 2, the null hypothesis of no
First First cointegration is rejected at 5% critical level and both
Variables Levels difference Levels difference the λ  trace and the λ  max statistics indicate the
IDN presence of two cointegrating vectors, that is, r = 2.
rgdp −1.57 −2.90** −2.01 −5.59* This implies that there is a presence of n  r ¼ 5 
CPI −1.51 −7.21* −1.33 −7.23* 2 ¼ 3 common trends between the real GDP of the
rEx −1.46 −11.12* −0.73 −11.10* ASEAN-5 countries. The cointegration test results
MYS also indicate the existence of 2 and 1 cointegrating
rgdp −1.47 −3.21* −1.61 −5.89*
CPI −2.92** −5.36* −2.66 −7.88* vectors for CPI and the RERs, respectively, implying
rEx −1.41 −11.05* −1.49 −11.04* 3 and 4 common trends between the CPI and the
PHL RERs, respectively, in ASEAN-5. These results
rgdp 0.02 −2.78*** −0.59 −4.23* further suggest that these macroeconomic variables
CPI −2.65 −5.46* −3.31* −7.07* of the ASEAN-5 cannot swing independently for a
rEx −1.31 −7.12* −1.30 −13.80*
SGP long time; they eventually move together in the long
rgdp −1.24 −3.70* −2.15 −5.49* run. We know that the existence of cointegrating
CPI −4.88* −3.97* −2.90** −5.93* relationship does not imply that these countries do
rEx −1.97 −11.27* −2.03 −11.49* not have policy differences; it simply means that the
THA short-run deviation will be corrected by internal
rgdp −1.51 −2.74*** −1.68 −4.78*
CPI −2.80 −6.15* −2.75 −6.28* dynamics within the system that corrects misalign-
rEx −1.21 −12.06* −1.21 −12.08* ment and brings these economies back to the equili-
OilP −1.97 −10.34* −1.78 −10.14* brium path in the long run (Darrat and Al-
Notes: *, ** and *** indicate significant at 1, 5 and 10%
Shamsi, 2005).
level, respectively.
IDN, Indonesia; MYS, Malaysia; PHL, Philippines; SGP, Impulse response
Singapore; THA, Thailand. In order to analyse the effect of oil price shocks, we
conducted the impulse response test of a onetime oil
AIC starting with a maximum lag-length of eight price shock on real GDP, inflation and the RERs.
lags. The AIC criteria indicate that the appropriate The impulse response test results are reported in
lag-lengths for real GDPs, inflations and RERs are 6, Fig. 1 (a–c) through Fig. 5 (a–c). Each figure traces

Table 2. Cointegration test results

Critical values (5%)
Variables Eigenvalues Null hypothesis λ-trace λ-max λ-trace λ-max
rgdp 0.287 r=0 110.98 53.26 76.97 34.80
0.183 r≤1 57.71 31.92 54.07 28.58
0.087 r≤2 25.79 14.44 35.19 22.29
0.042 r≤3 11.34 6.79 20.26 15.89
0.028 r≤4 4.55 4.55 9.16 9.16
CPI 0.257 r=0 97.46 46.48 76.97 34.80
0.141 r≤1 50.98 23.70 54.07 28.58
0.077 r≤2 27.27 12.57 35.19 22.29
0.061 r≤3 14.70 9.77 20.26 15.89
0.031 r≤4 4.93 4.93 9.16 9.16
rEx 0.213 r=0 78.04 37.63 60.06 30.43
0.108 r≤1 40.40 18.02 40.17 24.15
0.095 r≤2 22.38 15.81 24.27 17.79
0.029 r≤3 6.56 4.63 12.32 11.22
0.012 r≤4 1.93 1.93 4.12 4.12
Note: rgdp = real GDP and rEx = Real exchange rate.
3084 H. C. Basnet and K. P. Upadhyaya

(a) the horizontal line, the response becomes statisti-

.0020 cally insignificant.
Figure 1 (a–c) displays the response of the
.0015
Indonesian output, inflation and the RER, respec-
.0010 tively, to 1 SD structural innovation in oil price.
The Indonesian output shows a statistically positive
.0005
response to innovations for up to the first four quar-
.0000 ters (Fig. 1a). After the fourth quarter, the economy
begins to experience a contractionary effect. This
–.0005
negative impact, however, dies out after nine quar-
–.0010 ters, leading to again an expansionary phase in the
1 2 3 4 5 6 7 8 9 10 longer horizon. The effects of positive oil price
(b) shocks are different for oil-importing and oil-
.004 exporting countries. Oil-exporting countries may
.003 experience a positive impact on the economy at
.002 least in the short run. The positive impact may fade
away over time due to, among others, higher wage
.001
demands, which feed into the cost of production and
.000
consumer prices. Among the ASEAN-5, Indonesia is
–.001 an oil-exporting country and hence the country
–.002 experiencing an expansionary effect due to a positive
–.003 oil price shock is consistent with the conventional
–.004 wisdom. As expected, a positive oil price shock leads
1 2 3 4 5 6 7 8 9 10 to an inflationary effect contemporaneously (Fig.1b).
(c) It is also observed that the price level tends to
.015 decrease quickly after the first quarter, followed by
a sustained rise. The response is statistically signifi-
.010
cant for all the 10 quarters of time horizon. Our
.005 finding concurs with the evidence found by
Mehrara and Oskoui (2007), where oil price shocks
.000 have a very small effect on Indonesian output.
–.005 The contemporaneous effect of oil price shock on
the RER depicts a decline on it. It, however, starts
–.010 rising beginning the next quarter. The response of the
–.015
RER is more volatile than output and inflation as is
1 2 3 4 5 6 7 8 9 10 seen in Fig. 1(c). In the long run, Indonesian currency
suffers from depreciation. The impulse responses of
Fig. 1. Impulse response graphs. (a) Response all macroeconomic variables are statistically signifi-
of IDN_Real GDP to Shock1. (b) Response of IDN_
Inflation to Shock1. (c) Response of IDN_Exchange cant, suggesting that the Indonesian economy seems
rate to Shock1. Shock1 = oil price shock to have adverse impacts in its real and financial
sectors due to oil price shocks.
The impulse responses of Malaysia, Singapore, the
the effect of onetime positive shock to the oil price Philippines and Thailand display a very interesting
on the present and future course of each of the economic interdependency (see Figs 2–5). The
variables under consideration. The confidence results indicate that these four countries can be paired
bands are reported at the 95% confidence level. into two groups in terms of their similarity in
The solid line in the figures represents the actual responses. Malaysia and Singapore constitute the
impulse response. The dotted lines represent the first pair with the homogenous response to the oil
confidence intervals (upper and lower bands). If price shocks, while the Philippines and Thailand
the upper and lower bands coincide with or cross make another pair with identical responses. Note
Impact of oil price shocks on output, inflation and the RER 3085

(a) (a)
.00150 .0020
.00125
.0015
.00100

.00075
.0010
.00050

.00025 .0005

.00000
.0000
–.00025

–.00050 –.0005
1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10
(b) (b)
.00150 .0020
.00125
.0015
.00100

.00075 .0010

.00050
.0005
.00025
.0000
.00000

–.00025 –.0005
–.00050
1 2 3 4 5 6 7 8 9 10 –.0010
1 2 3 4 5 6 7 8 9 10
(c)
.004
(c)
.004

.002
.002

.000
.000

–.002
–.002

–.004
–.004

–.006
1 2 3 4 5 6 7 8 9 10 –.006
1 2 3 4 5 6 7 8 9 10
Fig. 2. Impulse response graphs. (a) Response
of MYS_Real GDP to Shock1. (b) Response of Fig. 3. Impulse response graphs. (a) Response
MYS_Inflation to Shock1. (c) Response of MYS_ of SGP_Real GDP to Shock1. (b) Response of
Exchange rate to Shock1. Shock1 = oil price shock SGP_Inflation to Shock1. (c) Response of SGP_
Exchange rate to Shock1. Shock1 = oil price shock

that the responses across the group are somewhat is positive. The growth of output, however, begins to
asymmetrical. This study intends to interpret the slow down right after the first quarter and dies out in
findings in the spirit of response similarities between the fifth quarters. The possible reason for experien-
these two groups. cing a brief positive effect could be companies may
Figures 2 (a–c) and 3 (a–c) display the impulse embed the oil price hike in their sale prices before it
response of Malaysian and Singaporean macro- passes on to the cost of production. This action will
variables. It is observed that the immediate response boost up industrial production. The effects are statis-
of real output to a onetime positive shock to oil price tically significant up until the sixth quarters. The
3086 H. C. Basnet and K. P. Upadhyaya

(a) (a)
.0008 .0008

.0006
.0004
.0004

.0002 .0000
.0000

–.0002 –.0004

–.0004
–.0008
–.0006

–.0008 –.0012
1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10
(b) (b)
.004 .004

.003 .003

.002 .002

.001 .001

.000 .000

–.001 –.001
1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10
(c) (c)
.004 .004

.003 .003

.002 .002

.001 .001

.000 .000

–.001 –.001

–.002 –.002

–.003 –.003

–.004 –.004
1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10

Fig. 4. Impulse response graphs. (a) Response Fig. 5. Impulse response graphs. (a) Response
of PHL_Real GDP to Shock1. (b) Response of of THI_Real GDP to Shock1. (b) Response of
PHL_Inflation to Shock1. (c) Response of PHL_ THI_Inflation to Shock1. (c) Response of THI_
Exchange rate to Shock1. Shock1 = oil price shock Exchange rate to Shock1. Shock1 = oil price shock

reaction of inflation to the shock in oil price in both price level, as a result of oil price shock, the RERs
countries is very similar and positive. As oil- of both countries fall but its effect becomes insignif-
importing countries, as expected, the price level icant after the fifth quarter. The findings from
shoots up contemporaneously (Figs 2b and 3b). Malaysia and Singapore indicate that the oil price
Prices begin to show a lesser impact as time passes, shock is not significant in affecting any of the macro-
which suggests that the oil price hike slowly assim- economic variables under study in the longer hori-
ilates into the prices of other goods and services until zon. Dynamic responses of Malaysia and Singapore
the new equilibrium is achieved. Contrary to the to oil price shock in this study differ from the past
Impact of oil price shocks on output, inflation and the RER 3087
studies. For instance, Ahmed and Wadud (2011), in a that the countries in each group tend to respond to
single-country (i.e. Malaysia) analysis, document the oil price shocks symmetrically. Further, a careful
that oil price uncertainty caused a significant decline observation of the figures indicates that the effect of
in Malaysian aggregate output. Likewise, in case of an oil price shock on inflation is quite similar in all
Singapore, Chang and Wong (2003) also find an four countries (except Indonesia) during the entire 10
adverse impact of an oil price shock on Singapore’s quarters horizon. One possible reason of such sym-
macroeconomic performance. metric response of inflation could be that these
Unlike the case of Malaysia and Singapore, a AEAN countries were successful in keeping the
positive oil price shock has a contractionary effect inflation rates within 5–6% range beginning the
on the output of the Philippines and Thailand (Figs 4a early 1990s by adopting several adjustment pro-
and 5a). This effect is rather brief and is completed in grammes in their policies (Upadhyaya and
the first quarter showing a positive effect momenta- Upadhyay, 1999). Indonesia, however, still suffered
rily. Afterwards the shock generates a contractionary from close to the double-digit inflation during the
impact, which is statistically significant throughout 1990s and 2000s; it recorded an almost 21% inflation
the 10 quarters horizon. While this effect becomes during the Asian financial crisis.
trivial in the longer horizon, a significant negative The similar responses between these two groups –
effect is observed for a prolonged period. Figures 4 Malaysia–Singapore, and the Philippines–Thailand,
(b) and 5(b) suggest that a positive oil price shock has warrant further economic intuitions. First, presum-
an inflationary impact in both the Philippines and ably the economies of Malaysia and Singapore are
Thailand’s economy but the effect begins to subside more integrated with each other relative to other
right after the first quarter. While inflation in countries in the ASEAN-5. Second, these two econo-
Thailand witnesses a sharp drop right after the first mies are driven by manufacturing, financial and ser-
quarter (Fig. 5b), there is a sustained negative reac- vice sectors. Third, these two economies have a well-
tion in the Philippines (Fig. 4b). The significant effect developed capital market, have attracted large scale
dissipates in both countries in the longer horizon. Our of FDI and have successfully implemented financial
findings echo the findings of Rafiq et al. (2009), that and economic reforms leading to macroeconomic
is, no significant impact of oil price volatility in the stability. On the other hand, the economies of the
Thai economy in the longer horizon. Further to the Philippines and Thailand largely consist of agricul-
point, Cunado and Garcia (2005) find evidence that tural sectors and they have histories of financial dis-
the impact of oil shocks on macroeconomy of Asian tress and crises. With relatively less stable and less
countries is limited to the short term. It is fair to say investment-friendly financial and economic system,
that oil-importing countries are found to have a nega- the two countries, unlike the former group, have not
tive impact of oil price hike on economic activities. been able to secure the same level of foreign invest-
The ASEAN countries are oil-importing countries ment. Above all, Indonesia stands distinctly out.
and a negative impact, only in the short run though, Among the ASEAN-5, Indonesia lacks sound eco-
is consistent with the conventional wisdom. In a nomic fundamentals. It suffered several banking,
study of some of the OECD oil-importing countries, financial and political crises during this study period.
Jiménez-Rodríguez and Sánchez (2005) also find a Indonesian currency experienced the most severe
negative effect of oil shocks on economic activity in devaluations during and before the Asian crisis, just
all cases (individual G-7 countries, Norway and the to provide an example. The differences in macroeco-
Euro area as a whole) except in Japan. As explained nomic fundamentals among these countries corrobo-
before, the higher oil price is transmitted into the rate the findings of our article. Countries may
prices of goods and services, leading to a new equili- experience a differential impact of oil price shocks
brium in the longer horizon. The RER (Fig. 5c) due to some variables such as their sectoral composi-
seems to be unaffected contemporaneously in tion or their relative position as oil importer or expor-
Thailand, rises temporarily after the third quarter, ter (Cunado and Garcia, 2005). Therefore, the similar
reaches to its peak around the fourth quarter, after response between Malaysia and Singapore and
which the impact is no longer significant. Almost an between the Philippines and Thailand and the stan-
identical phenomenon can be observed in the dalone response of Indonesia can be explained in
Philippines (see Fig. 4c). These findings suggest light of these contrasts.
3088 H. C. Basnet and K. P. Upadhyaya
Variance decompositions almost the same forecast error variation (see Table 3)
for each group mentioned in the previous section.
The variance decompositions (VDCs) of real out-
The estimated results indicate that oil price explains
puts, inflations and the RERs in the ASEAN-5 coun-
approximately the same variations in the macroeco-
tries because of oil price shock are reported in
nomic variables throughout the 10 quarters in
Table 3. This study intends to focus on the influence
Indonesia, Malaysian and Singapore. The results
of oil price shock on the real output, inflation and
show that oil price innovations explain almost 3
RER. Therefore, we limit our discussion of the fore-
and 5% forecast error in every quarter of the 10
cast error variances explained by oil price only. The
quarters in the real GDP of Malaysia and
VDC results largely confirm the findings from
Singapore, respectively. With respect to inflation
impulse response, that is, with a few exceptions, oil
and RERs, approximately the same variation can be
price shock has a much weaker influence on the
explained by uncertainty in the oil price during the 10
ASEAN economies. Interesting enough, the empiri-
quarters horizon. It accounts for 5.52 and 4.06%
cal results suggest that the oil price shock explains
variations in the inflation rate in Malaysia and
Singapore, respectively, in the tenth quarter. Almost
Table 3. VDCs of domestic variables due to oil price 5% of variance in the RER in Singapore is explained
shock by oil price shocks, whereas it explains only about
Real GDP Inflation Exchange rate 3% of the fluctuations in the RER in Indonesia. Oil
price contributes about the same variations in the
Period Indonesia Indonesian economy as in Malaysian and Singapore.
1 2.93 3.18 2.48
2 3.62 2.66 2.96
In the case of the Philippines and Thailand, the oil
4 4.62 2.46 3.01 price explains approximately the same level of fore-
6 4.68 2.36 3.50 cast error variation in real output and exchange rates
8 4.84 3.03 3.64 in both countries. It contributes very minimum varia-
10 4.82 3.22 3.74 tions (1.7 and 0.5% at most during the 10 quarters
Period Malaysia horizon for the Philippines and Thailand, respec-
1 3.11 4.15 2.12 tively) to the real GDP. Similarly, it does not exert a
2 3.87 5.78 3.21
4 3.29 5.52 3.26
significant impact in RER fluctuations. It explains
6 3.18 5.55 3.27 less than 1% of variation in the RER of these two
8 3.17 5.53 3.27 countries. This suggests that RER fluctuations in the
10 3.17 5.52 3.27 ASEAN-5 countries are not mostly due to oil price
Period Singapore shocks. But oil price shocks explain relatively a sig-
1 5.77 4.95 4.41 nificant portion of forecast errors in the inflation in
2 7.07 5.33 4.31 Thailand. It induces more than 21% variation in the
4 5.81 4.22 4.64
6 5.48 4.11 4.68
second quarters, which drops to 17.84% in the sixth
8 5.36 4.06 4.68 quarters and 17.42% in the tenth quarters, indicating
10 5.32 4.04 4.68 a persistent impact during the entire time horizon.
Period Philippines Like in Thailand, oil price shocks contribute a sig-
1 1.59 5.61 0.092 nificant fluctuation to the inflation in the Philippines.
2 1.34 6.54 0.092 It explains 6.54% of variation in the second quarters,
4 1.26 7.31 0.163 which increases to 7.61% in the sixth quarters and
6 1.51 7.61 0.163
8 1.68 7.71 0.164
7.73 in the tenth quarters. However, in Indonesia,
10 1.77 7.73 0.165 Malaysia and Singapore, it hovers around 3, 5 and
Period Thailand 4%, respectively, during the 10 quarters of the
1 0.031 20.96 0.173 horizon.
2 0.081 21.15 0.219 The impulse response and the VDC results suggest
4 0.343 18.45 0.433 that the oil price shocks do not have a major impact
6 0.462 17.84 0.446 on output, inflation and exchange rate in the
8 0.492 17.55 0.446
10 0.504 17.42 0.446
ASEAN-5 in the long run. Given that these econo-
mies are fairly open and very competitive in export, it
Impact of oil price shocks on output, inflation and the RER 3089
is possible that any negative effect of oil price Malaysia and Singapore display identical responses
increase in exchange rate, inflation and output is contemporaneously as well as in the longer horizon.
neutralized by their export demand. In addition, The same holds true for the Philippines and Thailand.
these countries are the favourite destination of But the response across these two groups is different.
resource-seeking as well as market-seeking FDIs, We attribute this finding to the similarities as well as
which also could have helped to dampen the negative dissimilarities to their macroeconomic fundamentals.
effect of oil price on their macroeconomic variables. However, while sectoral compositions are very much
Further to this point, the significant presence of FDI alike between the two groups, more research in future
in these economies can be related to the asymmetric is needed to further confirm such pairwise sym-
effect of oil price shocks. For example, when there metric/asymmetric responses.
are large oil price shocks, the foreign companies may The findings of this study provide a number of
find technologies that use alternate energy such as policy implications for the ASEAN-5 countries.
gas or coal, rather than cutting production. First, while macro-variables under consideration did
Equivalently, when oil price declines, the multina- not show any significant changes following oil price
tional companies do not give up their investment on shocks in the longer horizon, each of the five coun-
alternative technologies to minimize sunk costs. tries should put a defensive mechanism in place to
With respect to the impact of oil price shocks, our prevent the immediate negative impact. Second, our
findings are in accordance with the findings of findings suggest that oil price shocks do not have any
Cunado and Garcia (2005), who suggest that the significant effect on the macroeconomic perfor-
impact of oil price shock is limited to the short run mances of these countries presumably because of
for Asian countries that include four of the five the large inflow of FDI as well as the thriving export
sample countries in this study. sectors. Therefore, the ASEAN countries should
implement policies that attract more FDI as well as
promote the foreign sector of the economy. Third,
IV. Conclusion and Policy Implication since our estimation indicates that Thailand is more
prone to higher inflation rate because of the oil price,
This study empirically examines the impact of oil the Thai government should be more vigilant in con-
price shock on key macroeconomic variables of the trolling the price level that may increase with the
ASEAN-5 by using SVAR analysis. The cointegra- increase in the oil price. Finally, a policy coordina-
tion tests show that the macroeconomic variables of tion, not only the macroeconomic but also regulatory,
these countries are cointegrated, suggesting that they as well as social policy might be more helpful to deal
share a common trend in the long run. Our findings with the external shocks, including the oil price
suggest that an oil price shock is not the major shocks.
impediment of economic growth in ASEAN-5 coun-
tries as the impulse response function indicates that
oil price volatility does not impact in the long run, Disclosure Statement
much of its effect is absorbed within 5–6 quarters. No potential conflict of interest was reported by the
The VDC results also reveal that oil price volatility authors.
does not explain any significant portion of forecast
errors of any of the macroeconomic variables, with
exception to the inflation rate in Thailand. Based on References
overall findings, we conclude that the oil price shock
Ahmed, H. J. and Wadud, I. K. M. (2011) Role of oil price
does not induce much of the fluctuations in the shocks on macroeconomic activities: an SVAR
ASEAN-5 economies; and the impact is a short- approach to the Malaysian economy and monetary
term one. One of the interesting findings of this responses, Energy Policy, 39, 8062–9. doi:10.1016/j.
study is that while Indonesia does not exhibit any enpol.2011.09.067
similarities in responding to oil price shocks with Amano, R. and Norden, S. (1998) Exchange rates and oil
prices, Review of International Economics, 6,
Thailand, Malaysia, Singapore and the Philippines, 683–94. doi:10.1111/1467-9396.00136
the latter four can be paired into two groups with Arouri, M. E. H. and Rault, C. (2012) Oil prices and stock
respect to their responses to oil price shocks. markets in GCC countries: empirical evidence from
3090 H. C. Basnet and K. P. Upadhyaya
panel analysis, International Journal of Finance and International Review of Financial Analysis, 20,
Economics, 17, 242–53. doi:10.1002/ijfe.443 152–64. doi:10.1016/j.irfa.2011.02.014
Backus, D. and Crucini, M. (2000) Oil prices and the Hamilton, J. (1996) This is what happened to the oil price–
terms of trade, Journal of International Economics, macroeconomy relationship, Journal of Monetary
50, 185–213. doi:10.1016/S0022-1996(98)00064-6 Economics, 38, 215–20. doi:10.1016/S0304-3932
Bernanke, B. (1983) Irreversibility, uncertainty, and cycli- (96)01282-2
cal investment, Quarterly Journal of Economics, 98, Hamilton, J. (2000) What is an oil shock?, NBER Working
85–106. doi:10.2307/1885568 Paper, No. 7755. doi:10.3386/w7755
Brooks, C. (2008) Introductory Econometrics for Finance, Hamilton, J. D. (1983) Oil and the macroeconomy since
2nd edn, Cambridge University Press, Cambridge. World War-II, Journal of Political Economy, 91,
Cebula, C. J., Saadatmand, Y. and Piccone, Y. (2002) An 228–48. doi:10.1086/261140
empirical note on the inflation impact of the price of Hooker, M. (1996) What happened to the oil price–macro-
imported crude oil: the case of Germany, Risec, 49, economy relationship?, Journal of Monetary
531–7. Economics, 38, 195–213. doi:10.1016/S0304-3932
Cebula, R. J. (2000) A brief empirical note on impact of (96)01281-0
crude oil prices on domestic inflation: the case of the Hooker, M. (1999) Oil and the macroeconomy revisited,
United States, 1965–1999, Economia Internazionale, Federal Reserve Board, FEDS Working Paper No.
LIII, 449–54. 99–43. Federal Reserve Board, Washington, DC.
Cebula, R. J. and Michael Frewer, M. (1980) Oil imports Huntington, H. (1998) Crude oil prices and US economic
and inflation: an empirical international analysis of performance: where does the asymmetry reside?,
the ‘imported inflation’ thesis, Kyklos, 33, 615–22. Energy Journal, 19, 107–32. doi:10.5547/
doi:10.1111/j.1467-6435.1980.tb00716.x ISSN0195-6574-EJ-Vol19-No4-5
Chang, Y. and Wong, J. F. (2003) Oil price fluctuations Issa, R., Lafrance, R. and Murray, J. (2008) The turning
and Singapore economy, Energy Policy, 31, 1151– black tide: energy prices and the Canadian dollar,
65. doi:10.1016/S0301-4215(02)00212-4 Canadian Journal of Economics, 41, 737–59.
Christiano, J., Eichenbaum, M. and Vigfusson, R. (2007) doi:10.1111/j.1540-5982.2008.00483.x
Assessing structural VARs, NBER Macroeconomics Iwayemi, A. and Fowowe, B. (2011) Impact of oil price
Annual 2006, 21, 1–106. shocks on selected macroeconomic variables in
Cologni, A. and Manera, M. (2008) Oil prices, inflation Nigeria, Energy Policy, 39, 603–12. doi:10.1016/j.
and interest rates in a structural cointegrated VAR enpol.2010.10.033
model for the G-7 countries, Energy Economics, 30, Jiménez-Rodríguez, R. and Sánchez, M. (2005) Oil price
856–88. doi:10.1016/j.eneco.2006.11.001 shocks and real GDP growth: empirical evidence for
Cunado, J. and Garcia, F. (2005) Oil prices, economic some OECD countries, Applied Economics, 37, 201–
activity and inflation: evidence for some Asian coun- 28. doi:10.1080/0003684042000281561
tries, The Quarterly Review of Economics and Johansen, S. (1988) Statistical analysis of cointegration
Finance, 45, 65–83. doi:10.1016/j.qref.2004.02.003 vectors, Journal of Economic Dynamics and Control,
Darrat, A. and Al-Shamsi, F. (2005) On the path of inte- 12, 231–54. doi:10.1016/0165-1889(88)90041-3
gration in the Gulf region, Applied Economics, 37, Kilian, L. (2008) Exogenous oil supply shocks: how big
1055–62. doi:10.1080/00036840500109027 are they and how much do they matter for the US
El Anshasy, A. A. and Bradley, M. D. (2012) Oil prices economy?, Review of Economics and Statistics, 90,
and the fiscal policy response in oil-exporting coun- 216–40. doi:10.1162/rest.90.2.216
tries, Journal of Policy Modeling, 34, 605–20. Mehrara, M. (2008) The asymmetric relationship between
doi:10.1016/j.jpolmod.2011.08.021 oil revenues and economic activities: the case of oil-
Enders, W. (2004) Applied Econometrics Time Series, exporting countries, Energy Policy, 36, 1164–8.
John Wiley & Sons, New York, NY. doi:10.1016/j.enpol.2007.11.004
Farzanega, M. R. and Markwardt, G. (2009) The effects of Mehrara, M. and Oskoui, K. N. (2007) The sources of
oil price shocks on the Iranian economy, Energy macroeconomic fluctuations in oil exporting coun-
Economics, 31, 134–51. doi:10.1016/j.eneco. tries: a comparative study, Economic Modelling, 24,
2008.09.003 365–79. doi:10.1016/j.econmod.2006.08.005
Fernald, J. and Trehan, B. (2005) Why hasn’t the jump in Nelson, C. R. and Plosser, C. R. (1982) Trends and ran-
oil prices led to a recession?, in FRBSF Economic dom walks in macroeconomic time series: some evi-
Letter, Federal Reserve Bank of San Francisco, dence and implications, Journal of Monetary
2005–31. Economics, 10, 139–62. doi:10.1016/0304-3932
Filis, G., Degiannakis, S. and Floros, C. (2011) Dynamic (82)90012-5
correlation between stock market and oil prices: the Park, C., Chung, M. and Lee, S. (2011) The effects of oil
case of oil-importing and oil-exporting countries, price on regional economies with different
Impact of oil price shocks on output, inflation and the RER 3091
production structures: a case study from Korea using investigation in the Thai economy, Resources Policy,
a structural VAR model. Energy Policy, 39, 8185–95. 34, 121–32. doi:10.1016/j.resourpol.2008.09.001
10.1016/j.enpol.2011.10.018 Sill, K. (2007) The macroeconomics of oil shocks,
Phillips, P. and Perron, P. (1988) Testing for a unit root in Business Review, Q1, 21–31.
time series regression, Biometrika, 75, 335–46. Upadhyaya, K. and Upadhyay, M. (1999) Output effects
doi:10.1093/biomet/75.2.335 of devaluation: evidence from Asia, Journal of
Rafiq, S., Salim, R. and Bloch, H. (2009) Impact of crude Development Studies, 35, 89–103. doi:10.1080/
oil price volatility on economic activities: an empirical 00220389908422603