www.ccsenet.org/mas Modern Applied Science Vol. 5, No. 2; April 2011
ISSN 1913-1844 E-ISSN 1913-1852
192
Oil Price Shocks: A Comparative Study on the Impacts in Purchasing
Power in Pakistan
Muhammad Bachal Jamali
Associate Professor, Deptt of Commerce
Shah Abdul Latif University-Khairpur
Asif Shah
Assistant Professor, Deptt of Commerce
Shah Abdul Latif University-Khairpur
Hassan Jawad Soomro
Assistant Professor
Department of Commerce, SALU-Khairpur
E-mail: hassanjawadsoomro@yahoo.com
Kamran Shafiq
COMSATS-Abbott bad
E-mail: kamranhashmi@ciit.net.pk
Faiz M.Shaikh
Assistant Professor
SZABAC-Dokri, Sindh-Pakistan
Email: faizanmy2000@hotmail.com
Received: November 23, 2010 Accepted: February 28, 2011 doi:10.5539/mas.v5n2p192
Abstract
The current research investigates the relationship between changes in crude oil prices and Pakistan and the
macro-economy. A multivariate VAR analysis is carried out among five key macroeconomic variables: real
gross domestic product, short term interest rate, real effective exchange rates, long term interest rate and money
supply. From the VAR model, the impulse response functions reveal that oil price movements cause significant
reduction in aggregate output and increase real exchange rate. The variance decomposition shows that crude oil
prices significantly contribute to the variability of real exchange rate long term interest rate in the Pakistan
economy while oil price shocks are found to have significant effects on money supply and short term interest
rate in the economy. Despite these macro econometric results, caution must be exercised in formulating energy
policies since future effects of upcomming oil shocks will not be the same as what happened in the past.
Explorations and development of practicable alternatives to imported fuel energy will cushion the economy from
the repercussions of oil shocks. Oil price shock has negative impact on the GDP and as well as economy of
Pakistan.
Keywords: Oil price fluctuations, Macroeconomic performance, Pakistan
1. Introduction
The global economy is now facing its worst prospects in more than half a century, with increasing financial
losses, falling asset prices, and a deep downturn in real economic activity. Several developed economies
including the US, UK, Japan and Germany are already in recession. Overall global GDP growth is projected to
decline by the World Bank, from 2.5 per cent in 2008 to 0.9 per cent in 2009, the weakest since records became
available in 1970. International trade would decelerate sharply, with global export volumes declining for the first
time since 1982. As labor market conditions have deteriorated, consumer spending, business investment, and
industrial production have also declined, the Federal Reserve lowered the target for its benchmark interest rate
and established a target range for the federal funds rate of 0 per cent and 0.25 per cent. The European
Commission in November unveiled an economic recovery plan worth €200bln.with aims to save further job
losses, stimulate spending and boost consumer confidence. The ECB has reduced its key policy rate from 4.25
per cent in September to 2 per cent by mid-January. The British central bank has reduced interest rates four times
www.ccsenet.org/mas Modern Applied Science Vol. 5, No. 2; April 2011
Published by Canadian Center of Science and Education
193
since April from 5.0 per cent to 1.5 per cent. The latest half-percentage point cut in January 2009, that brings the
rate to its lowest level in the central bank's 315-year history, was necessitated by weakening consumer spending,
a tightening credit market for households and businesses, and a deteriorating business and residential investment
outlook. Bank of Japan has cut interest rates from 0.5 per cent to 0.3 and then to 0.1 per cent by mid-December,
and has adopted several liquidity enhancing measures. The Bank of Korea lowered its Base Rate from 5.00 per
cent in early October to 2.50 per cent by January 2009. China has cut lending rates considerably since
mid-September and unveiled a 4 trillion-yuan fiscal stimulus package in early November to rejuvenate the
weakening economy. The Bank of Thailand cut the benchmark interest rate by 75 basis points to 2 per cent in
January, the decision, which follows a 1-percentage point reduction in December, is more aggressive than
expected.
A permanent oil price shock would clearly have a major impact on the world economy. This should send a
message to policymakers around the world to consider ways to tackle demand and improve energy efficiency, in
order to reduce the vulnerability of their economies to an oil price shock. Oil price shocks would normally affect
macroeconomic performance through a number of channels. First, higher oil prices transfer income from
oil-importing countries to oil-exporting countries through a shift in the terms of trade. This results in a loss of
real income for oil-importing countries. Second, higher oil prices reduce industry output through higher costs of
production. Third, they directly increase inflation via higher prices of imported goods and petroleum products. If
higher inflation leads to an upward spiral in wages, central banks would be forced to raise interest rates.
Higher oil price causes different impacts to both net oil importers and net oil exporters in this world (combining
both crude and products). The effect of the oil shock is expected to lower world GDP because of the reduced
purchasing power by the oil importers to balance higher oil import costs will not fully offset by increased
demand for imports from oil exporters. Therefore, GDP of most oil importing countries fall as their exports of
other goods will fall as well. As a net exporter of oil, oil price shocks will impede the growth of trade between
Pakistan and other countries, especially for oil importing countries like U.S., China, Japan and Europe.
Economic slowdown in these countries will limit their demand of consumers’ and thus affect Pakistan exports of
goods and services.
2. Literature Review
Since the first oil shock in 1973/74, much research has been undertaken into the oil price–macro economy nexus.
These studies have reached different conclusions over time. Earlier works (Hamilton, 1983; Burbidge and
Harrison, 1984) have achieved statistically significant empirical relationships between oil prices and aggregate
economic performance, principally GDP/ GNP growth. Hamilton (1983) propounded three hypotheses for
oil-shock and output correlation: (1) historical coincidence,(2)endogeneity of crude oil prices, and (3) causal
influence of an exogenous increase in the price of crude petroleum. Econometric results showed that there was
insignificant evidence that the correlation was neither a consequence of coincidence nor a set of influences that
triggered oil shocks and recessions. The causal interpretation leads to the conclusion that the characteristics of
the pre-1973 recessions would have been different if such energy shocks and disruptions did not come about
(Hamilton 1983).
Meanwhile Burbidge and Harrison (1984) tested the effects of increases in oil 9prices using a seven-variable
vector auto regression (VAR) model for five countries (United States, Japan, Germany, United Kingdom and
Canada) in the Organization for Economic Cooperation and Development (OECD) using monthly data from
January 1961 to June 1982. They found out that substantial effects of oil-price shocks on price level were evident
on the U.S. and Canadian economies and with great pressure on industrial production on U.S. and U.K. They
also pointed out that the oil shock in 1973 only worsened the incoming recession of that period. Following the
collapse of oil prices in 1986, it was argued that the oil price–macro economy relationship has weakened. In
addition, an asymmetric oil price–macro economy relationship was established (Mork, 1989). Mork (1989)
extended Hamilton’s study by using a longer data sample and taking into account oil price controls existed
during the 1970s. Furthermore, he looked into the possibility of an asymmetric response to oil price increases as
well as decreases. The results showed that GNP growth was correlated with the circumstances of the oil market
and that oil price declines were not statistically significant as oil price increases.
Mork, Olsen and Mysen (1994) applied essentially the same model as Mork (1989) to the experience of seven
OECD countries over the period 1967:3-1992:4. Their model also included the contemporaneous oil price and
five quarterly lags for price increases and decreases separately. For the United States, the contemporaneous price
increase and the first and second lags were significant, and of negative sign. Five of the other six countries;
Japan, West Germany, France, Canada, and the United Kingdom had roughly similar patterns of coefficients,
while Norway had positive, statistically significant elasticities for both price increases and decreases.
When almost all researches dealt with the effects of oil prices, as measured in 10levels or in logarithmic form,
on key macroeconomic variables, (Ferderer, 1996) used oil price volatility (monthly standard deviations of daily
oil prices) to assess movements in U.S. aggregate output. He also took note of the monetary channel through
which oil prices affect the economy by including federal funds rate and non-borrowed reserves to capture the
www.ccsenet.org/mas Modern Applied Science Vol. 5, No. 2; April 2011
ISSN 1913-1844 E-ISSN 1913-1852
194
monetary policy stance during oil shocks. Results showed that contractionary monetary policy in reaction to oil
price increases partly explains the correlation between oil and output. However, sectoral shocks and uncertainty
channels, but not monetary policy channel, provide partial explanation to the asymmetric relationship between
oil price changes and output growth (Ferderer, 1996).
3. Theoretical Framework
Volatility of oil prices has negative repercussions on the aggregate economy as abundantly shown by economic
literature. An oil price shock, as a classic example of an adverse supply shock, i.e. an increase in oil prices shifts
the aggregate supply upward, results to a rise in price level and a reduction in output and employment
(Dornbusch, Fisher and Startz 2001). On the other hand, aggregate demand decreases as higher commodity
prices translate to lower demand for goods and services, resulting to contraction in aggregate output and
employment level. The macroeconomic effects of oil shocks are transmitted via supply and demand side
channels and are potentially minimized by economic policy reactions.
3.1 Supply Side Channel
Since oil is a factor of production in most sectors and industries, a rise in oil prices increases the companies’
production costs and thus, stimulates contraction in output (Jimenez-Rodriguez and Sanchez 2004). Given a
firm’s resource constraints, the increase in the prices of oil as an input of production reduces the quantity it can
produce. Hunt, Isard and Laxton (2001) add that an increase in input costs can drive down non-oil potential
output supplied in the short run given existing capital stock and sticky wages. Moreover, workers and producers
will counter the declines in their real wages and profit margins, putting upward pressure on unit labor costs and
prices of finished goods and services.
In addition, oil price volatility shrinks investment activities in production of oil and gas (Verleger 1994 as cited
from Raguindin & Reyes, 2005). Verleger (1994) as cited from Raguindin & Reyes (2005) adds that a
“permanent increase in volatility might lead to a situation where future capacity will always be a little lower than
in a world of zero price volatility and prices a little higher”. Hamilton (1996) shares the same point and stresses
that concerns on oil prices variability and oil supply disruptions could cause postponement of investment
decisions in the economy.
3.2 Demand Side Channel
As presented earlier, oil price increases translate to higher production costs, leading to commodity price
increases at which firms sell their products in the market. Higher commodity prices then translate to lower
demand for goods and services, therefore shrinking aggregate output and employment level. Furthermore, higher
oil prices affect aggregate demand and consumption in the economy. The transfer of income and resources from
an oil-importing to oil-exporting economies is projected to reduce worldwide demand as demand in the former is
likely to decline more than it will rise in the latter (Hunt, Isard and Laxton 2001). The resulting lower purchasing
power of the oil-importing economy translates to a lower demand. Also, oil price shocks pose economic
uncertainty on future performance of the macro economy. People may postpone consumption and investment
decisions until they see an improvement in the economic situation. In sum, an increase in oil prices causes a
leftward shift in both the demand and supply curve, resulting to higher prices and lower output.
3.3 Economic Policy Reactions
The effects of oil price increases on headline and core inflation may stimulate the tightening of monetary policy
(Hunt, Isard and Laxton 2001). Authorities have the policy tools to minimize, if not totally eliminate, the adverse
effects of such shock. The Central Bank (CB) has its key policy interest rates that can influence demand and
inflation directions in the economy. However, pursuing 12one policy can be counterproductive; when CB cuts
its interest rate, demand rises, but at the expense of higher inflation, and vice versa.
The credibility of the monetary authorities in responding to oil shocks is at stake when monetary policy reactions
appeared inconsistent with the announced policy objectives. As a result, inflation expectation and process are
disrupted (Hunt, Isard and Laxton 2001). Money supply plays a role on the negative correlation between oil
prices and economic activity. By means of the real money balances channel, increases in oil prices cause
inflation which, in turn, reduces the quantity of real balances in the economy (Ferderer 1996). Ferderer (1996)
further noted that “counter inflationary monetary policy responses to oil price shocks are responsible for the real
output losses associated with these shocks”.
4. Empirical Method
This section presents the empirical method used in this paper to assess the oil price- macro economy relationship
of the Pakistan. First, data definition and limitation are discussed. Second, a vector auto regression (VAR) model
was constructed using historical data to capture the behavior of the macro economy given oil price fluctuations.
Impulse response functions were examined to trace out the response of the dependent variable in the VAR model
to shocks in the error terms. Variance decomposition technique was done to evaluate the relative importance of
oil price fluctuations on the volatility of the other variables in the model.
www.ccsenet.org/mas Modern Applied Science Vol. 5, No. 2; April 2011
Published by Canadian Center of Science and Education
195
4.1 The Data
This paper used quarterly data for the period 1992:2 to 2006:4 of five macroeconomic variables and oil price
variables to capture economic behavior. The model includes output and exchange rate variables (real gross
domestic product (RGDP) and real effective exchange rate (REER), three monetary variables namely money
supply (M1), long term interest rate (GBOND5) and short term interest rate (TBILLS3) and the oil price variable
(ROIL). RGDP, ROIL and REER were expressed in logarithmic form while M1, GBONDS5 and TBILLS3 were
expressed in levels. The data sets were obtained from the International Finance Statistics (IFS), Economic
Planning Unit (EPU), Statistics Department of Pakistan and the Economic and Social Research Council (ESRC)
of the W.
4.2 Definition of Terms
Five of the most commonly used terms in this research are defined as follows:
1. Gross Domestic Product (RGDP) is a measure of total output within the geographic limits of the country,
regardless of the nationality of the producers of output.
2. Real Effective Exchange Rate (REER) index of the Ringgit Pakistan 13(RM) and the British Pound Sterling
are the Nominal Effective Exchange Rate Index (NEERI) of the RM and Pound adjusted for inflation rate
differentials with the countries whose currencies comprise the NEERI basket.
3. Short Term Interest Rate (TBILLS) interest rates on loan contracts-or debt
4. Long Term Interest Rate (GBONDS) is the interest rate earned by a note or bond that matures in 10 or more
years.
5. Money supply (M1) is currency plus demand deposits.
4.3 Oil Price Variable
A number of studies used different oil price variables to account for the effects of these shocks on economic
activity. Hamilton (1983) used the quarterly changes in nominal Producer’s Price Index (PPI) for crude
petroleum. Burbidge and Harrison (1984) employed a relative price of oil computed as the ratio of Saudi Arabian
crude cost (US$) to the CPI of the country under studied. Mork (1989) used the refiner acquisition cost (RAC)
for crude oil and PPI. Ferderer (1996) used the monthly means and standard deviations of prices for refined
petroleum products (deflated by CPI) as the real oil price and oil price volatility, respectively. Abeysinghe (2001)
proposed different definitions of oil price variables4 and finally modeled the oil price in first-log-difference of
oil price (in US$) multiplied by the country’s exchange rate. He pointed out that the other real oil price definition
appears to be a poor proxy for the relative oil price because of the direct dependence of CPI to oil price. Hooker
(1996a) and Jimenez-Rodriguez and Sanchez (2004) both used oil prices in real terms but the former also
included nominal PPI for crude petroleum in his regression model.
Most of the international cross-country analysis used the US$ world oil price in real terms (PPI for crude oil
divided by PPI for all commodities) or the world oil price transformed into each country’s currency through the
exchange rate. However, only the latter recognizes the different effects of oil prices on each country due to
exchange rate volatility or level of inflation. Furthermore, as noted by Cunado and de Gracia (2004), oil prices
converted into each country’s currency produced more significant impacts on variables under study.
4.4 The Vector Auto regression (VAR) Model
A number of the studies cited made use of vector auto regression models. This technique treats all variables in
the system as endogenous and regresses each current (non-lagged) variable in the model on all the variables in
the model lagged a certain number of times.
The study employs the following VAR model of order p (VAR (p)):
Yt = c + Σ AiYt-1 + εt ,
where Yt is a (n x 1) vector of endogenous variables, c is the intercept vector of the VAR, Ai is the ith matrix of
autoregressive coefficients and εt is the generalization of a white noise process. The study estimated two sets of
VAR models which incorporated the linear and nonlinear specifications of oil price response to economic
activity. The first VAR model used the oil price variable measured as the log - first-difference of crude oil.
4.5 VAR Applications
A six-variable vector auto regression model is presented to examine the sources of variations and fluctuations in
the Pakistan and south Asian countries economies triggered by oil prices. The first step of our analysis is to test
for stationary – to investigate the existence of unit roots in our statistical series by calculating the Augmented
Dickey-Fuller Test (ADF Test). This test is based on autoregressive models that always include an intercept and
generally a trend component. A large negative test statistic rejects the null hypothesis and implies that the time
series is stationary.
The Akaike information criterion (AIC) will be used to compare the performance of the VAR with various lag
length specifications. Both variance decomposition and impulse responses will be utilized to assess the
www.ccsenet.org/mas Modern Applied Science Vol. 5, No. 2; April 2011
ISSN 1913-1844 E-ISSN 1913-1852
196
relationship between oil price shocks and aggregate economic activity. A variance decomposition provide the
variance of forecast errors in a given variable to its own shocks and those of the other variables in the VAR. It
allows us to assess the relative importance of oil price shocks to the volatility of the other variables. Impulse
response functions allow us to examine the dynamic effects of oil price shocks on Pakistan and the British
macroeconomics. It traces over time the expected responses of current and future values of each of the variables
to a shock in one of the VAR equations.
5. Results and Discussion
In this section, the preliminary tests and data transformations are presented. Moreover, the empirical results
obtained from the estimated VAR models using linear oil price specifications are discussed. The impulse
response functions and variance decompositions obtained from the estimated VAR models are also expounded.
5.1 Presentation of Results
Tests of Stationary Econometric analysis using time-series data necessitates stationary. To have stationary
representations of the VAR models, each variable was tested for unit roots specification using the augmented
Dickey-Fuller (ADF) test. Table 1 and 2 provide the unit root regression results in levels and first-differences of
the variables entered in the model and the corresponding critical value of 10%, 5% or 1% to reject the null
hypothesis of the presence of a unit root.
Integration Test for Pakistan
The ADF statistics in Table 1 suggest that all six variables are integrated of order one, whereas the
first-differenced are integrated of order zero. These non-stationary variables were transformed by taking their
first differences in order to exhibit stationary, indicating that the mean, variance and covariance of the time series
are independent of time.
Notes: We denote with one/two/three asterisks the rejection of the null hypothesis of the presence of unit root at
10% / 5% / 1% critical levels. The calculated statistics are those computed in MacKinnon (1991).
Integration Test
Table 2 provides the unit root regression results for the W. Only TBILLS was stationary in levels. The remaining
variables, namely REER, RGDP, ROIL, GBONDS and M1 are observed to be non-stationary at all significance
levels but exhibit stationary after the variables were transformed by taking their first difference, indicating that
the mean, variance and covariance of the time series are independent time 16
Optimal Lag Length
Next, the Akaike Information Criterion (AIC) was used to assess the performance of the VAR model with
varying lag length specifications. The optimal lag length is the one that minimizes the AIC. The AIC showed
that the optimal lag length is six (6) for VAR models of Pakistan and the (refer to Table 3 and Table 4)
Impulse Response Function
An impulse response function (IRF) was computed from the coefficients of vector regression using
orthogonalzed set of residuals. IRF traces the effect of one standard deviation shock to one of the innovations on
current and future values of each of the endogenous variables in the system.
IRF: Pakistan
Generally, most of the variables show an increase during the first few quarters, with the exception of real GDP,
GBONDS and ROIL. Chart 1 presents the IRFs generated from the VAR model using the linear specification of
crude oil prices and show that a positive oil price shock leads to a decline in real GDP, long term interest rate
and real oil price, persisting for three (3) quarters after which, the three variables recover. Money supply and
short term interest rate increase a quarter (with the exception of real exchange rate which increases for three
consecutive periods) after an oil price shock. However, such increase do not last long (i.e., M1 and TBILLS3 go
back to its pre-shock level between the third and fourth quarters) while REER goes back to pre-shock level
between four and fifth quarter.
VCOM: Pakistan
Table 5 shows the variance decomposition of the VAR model specification for Pakistan. It suggests that oil price
shocks contribute a relatively large share on the long-term interest rate and reel effective exchange rate. In most
cases, if not at all times, the variable itself are the largest source of its own variation in succeeding periods.
The largest effect of an oil shock to a variable’s variation is on long-term interest rate (GBONDS5), accounting
for approximately 18 percent in the third, fourth and the fifth period. Likewise, crude oil prices account for 11
percent of real exchange rate volatility.
Meanwhile, crude oil prices are marginal sources of variation of short-term interest rate (TBILLS3). Volatility of
money supply (M1) due to oil price fluctuations is accounted for 8 percent. Changes in real GDP and TBILLS3
are nominal, accounting for only 5 percent and 4 percent respectively.
