Does oil price volatility influence real sector growth? Empirical evidence from Pakistan

Yasmeen, Humaira; Wang, Ying; Zameer, Hashim; Solangi, Yasir Ahmed Article Does oil price volatility influence real sector growth? Empirical evidence from Pakistan Energy Reports Provided in Cooperation with: Elsevier Suggested Citation: Yasmeen, Humaira; Wang, Ying; Zameer, Hashim; Solangi, Yasir Ahmed (2019) : Does oil price volatility influence real sector growth? Empirical evidence from Pakistan, Energy Reports, ISSN 2352-4847, Elsevier, Amsterdam, Vol. 5, pp. 688-703, https://doi.org/10.1016/j.egyr.2019.06.006 This Version is available at: https://hdl.handle.net/10419/243622 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by/4.0/ Energy Reports 5 (2019) 688–703 Contents lists available at ScienceDirect Energy Reports journal homepage: www.elsevier.com/locate/egyr Research paper Does oil price volatility influence real sector growth? Empirical evidence from Pakistan Humaira Yasmeen ∗, Ying Wang, Hashim Zameer, Yasir Ahmed Solangi College of Economics and Management, Nanjing University of Aeronautics and Astronautics, Nanjing 211106, China article info Article history: Received 26 December 2018 Received in revised form 12 May 2019 Accepted 19 June 2019 Available online xxxx Keywords: Oil price volatility Real economic sectors Economic growth ARDL Pakistan abstract The study investigates the short-run and long-run relationship between oil price fluctuation and real sector growth in Pakistan. Four major sectors of the economy (Manufacturing, electricity, transport and communication, and livestock) were analyzed to find any relation. Similar studies can be found in the existing literate, however, the distinguish feature of present study is that it investigates each individual sector’s linkage to oil price changes. Annual time series data of selected sectors ranging from 1976 to 2017 is selected for the study. Classical normal linear regression models under auto regressive distributed lag (ARDL) were employed to study the relationship between economic sectors and oil price fluctuation. Empirical results indicate that changes in oil price adversely affect manufacturing, livestock and electricity sectors in short-run and long-run, while significant positive impact was found on transportation and communication. Consequently, the sectors prone to oil price changes require special attention of policy makers. An expansionary monetary policy can be a short-run solution to reduce the impact of increasing oil price, whereas the government can introduce a policy framework to counter this effect in long-run. ©2019 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). 1. Introduction Oil is one of the major energy sources for both the developed and the developing economies of the world. It serves as a backbone for economic development, as it satisfies the industrial and domestic energy requirements of any country. The data of June 2017 indicate that United States along with China, Japan, India and Russia are the five major oil consuming countries in the world. Collectively, these five countries consume 40 million barrels per day. However, consumption is expected to further aggravate in future due to rising demand for energy in these countries. The rise in demand is expected to occur because the rapid economic development of export oriented China and India, it has challenged the global trade dominance of the United States, Japan and the European Union. While, these global players compete, a sudden oil price shock can have a detrimental impact on their economic growth which can consequently cause a global economic recession. As such, oil prices can influence the performance of numerous macroeconomic factors. For example, oil price fluctuations influence economic policy uncertainty (Ahmed et al.,2018;Kang et al.,2017;Kang and Ratti,2015;Wesseh Jr and Lin,2018), human development (Marza et al.,2018), movements ∗Corresponding author. E-mail addresses: [email protected] (H. Yasmeen), [email protected] (Y. Wang), [email protected] (H. Zameer), [email protected] (Y.A. Solangi). in stock and bond prices (Bastianin et al.,2016;Waheed et al., 2018), inflation (Naser,2019), interest rates (Nazlioglu et al., 2019), portfolio optimization (Sarwar et al.,2019) and business cycle (Pönkä and Zheng,2019). The scientific literature from the past support the argument that upward oil price shock has a negative influence on economic growth of both developed and the developing economies (Kilian, 2008;Kilian and Vigfusson,2011;Narayan et al.,2014). The study of Hamilton (2009) highlights that the oil price shock of 2007– 2008 were due to high global demand for oil while the preceding oil shocks are attributed to the supply disruptions; but, the consequence of both types have been the same which results in causing economic recessions. According to Lescaroux and Mignon (2008) rising oil prices have an adverse effect on the economic growth of net oil importing countries because both consumer and producers of goods and services suffer. The producers have to suffer because the marginal cost of production undergoes an increment resulting into a decline in profitability; whereas, the consumer has to cut down their consumption of goods other than necessary if disposable income lags behind inflation. All this adds up to having an adverse impact on economic output that consequently results into an adverse impact on real wages, employment, profitability, investment and price level. It was generally perceived that developed economies were the only who were drastically affected by the oil price surge. But, in reality, the developing countries are more affected by these oil price shocks because of inefficient energy utilization and wastage. The price surge in the https://doi.org/10.1016/j.egyr.2019.06.006 2352-4847/©2019 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). H. Yasmeen, Y. Wang, H. Zameer et al. / Energy Reports 5 (2019) 688–703 689 2000’s have been a matter of great distress for the economists of developing countries because the oil price increments had interrupted the economic growth by creating an inflationary pressure in the economy, large government budget deficits and problems in balance of payments. Moreover, the recent fall in oil prices has opened a new paradigm of discussion in contrast to the general view which indicates that downward fall in oil price is better for economic activity worldwide. It has been stated that in industrialized economies, the near zero interest rate has changed the traditional channels through which benefits of lower oil prices get transferred to real economic output (Obstfeld et al.,2016). Additionally, the positive association among oil price fluctuations and equity markets has also provided the evidence of a slowdown of global economic activity, as a relaxing of aggregate demand has condensed profitability of firms and the overall demand for oil (Bernanke,2016). Previous research reveals that energy plays an important role in improving economic productivity (Shahbaz et al.,2017). Further, the economic progress in any country stimulates the demand of energy due to the demand in consumption patterns (Sadorsky,2013). The study of Komal and Abbas (2015) highlighted that EIA estimates have shown 56%increase in energy consumption in 2010–2040 worldwide. Mostly, the increase in energy consumption will occur in non-OECD countries, where energy consumption is encouraged by strong economic growth (Islam et al.,2013;Khan and Ahmad,2008). Population growth and industrial sector expansion has increased the use of energy consumption in Asian countries, specifically in Pakistan (Zaman et al.,2017,2012); and the country belongs to the group of low middle income countries. Pakistan is facing severe energy crises from last two decades (Zameer and Wang,2018). The country has 6.5% long-run growth potential, while energy crises have reduced this potential to 2% (Komal and Abbas,2015). It indicates that the economic growth of the country is largely suppressed by the energy crises. The electricity shortage in the country is creating negative effect on exports, international competitiveness, poverty alleviation and employment in Pakistan (Kessides,2013). The overall evaluations of previous research indicates that mostly studies have emphasized on overall economic performance of the countries, but, studies that focus on the behavior of individual sectors in response to oil price fluctuations are scarce in literature. Since Pakistan is a developing country and oil is considered as a major source for energy production and similarly it drives the economic growth. Further, country’s most of the energy production system is dependent on thermal electricity (Solangi et al., 2018;Zameer and Wang,2018). The country fulfills most of its energy requirement using imported furnace oil (Wakeel et al., 2016). The undiversified energy production mix and huge reliance on imported oil has made Pakistan more reactive to oil prices. Most of the oil consumption requirements of the country is based upon imported furnace oil from the Gulf countries. Therefore, an upward trend in oil price is expected to negatively affect the economic growth of Pakistan like other developing countries in the world. The rise in oil will result into a rise in the production cost, the balance of payment problems, exchange rate depreciation, government budget deficits, fall in aggregate demand, fall in real wages, unemployment, overly contractionary or expansionary monetary policy, and even an economic recession (Malik,2008). Pakistan’s energy mix is highly imported furnace oil dependent. The upward rise in oil prices can influence only oil dependent countries like Pakistan, while the global consumption may fall as the trading partners are also adversely affected by oil shocks. Therefore, it is highly significant to point out that oil prices will not only influence domestically but the affects becomes more severe when exports decline. The previous studies in the context of Pakistan by (Khan and Ahmed,2011b;Malik, 2010;Syed,2010a) have revealed that oil price has a negative relationship with GDP growth; but so far we have not come across any work that could signify the relationship between oil prices and growth of individual economic sectors contributing to Pakistan’s gross domestic production. As, the role of individual economic sectors is highly important for the overall economy. Accordingly, it is vital to signify the relation between oil prices and growth of individual economic sectors contributing to Pakistan’s gross domestic production. Therefore, the aim of this study is to explore the effects of oil price changes on the growth of the real sector in Pakistan. The real sector of Pakistan comprises of agricultural, industrial, and services sector. Within these sectors, the study highlighted over four sub-sectors (Manufacturing, electricity, transport and communication, and livestock) that are instrumental to Pakistan’s gross domestic product. This study will put-forward how the growth of an individual sector responds to the fluctuations in oil price. The study fills the research gap by exploring the impact of oil price fluctuations on the growth of individual economic sectors. Similarly, the study put-forward the policy framework, so that the policy maker can avoid over or under-reacting to oil price shocks by developing a sound monetary policy that is neither over contractionary nor over expansionary. Further, it pitches the vulnerable sectors those are not immune to oil price shocks. It will help the policy makers in redirecting their attention to the vulnerable sectors and facilitate them according to their unique requirements. Similarly, the investment in such sectors is encouraged and they can withstand the detrimental impacts of oil price shocks. 2. Materials and methods 2.1. Data Broadly, the study has three sectors under investigation that are the agricultural, industrial and the services sector. Within these sectors, the study selected four sub-sectors that are critically important to gross domestic production. Data of all variables is annual data and encompasses the period from fiscal year 1976 to 2017. All the economic data for Pakistan was obtained from ‘‘handbook of statistics on Pakistan economy 2015’’ that is available at the State Bank of Pakistan’s website (Pakistan,2017) and Pakistan Economic Survey 2016–17 available at the website of the ministry of finance (Finance,2018), whereas UK Brent crude oil price data was obtained from commodity price data available at World Bank’s website. The real output in million rupees from all the selected subsectors has been transformed into growth by calculating a quantum index. The quantum index has been calculated at a constant factor cost of year 1980–81 by using the Laspeyer’s formula (1) following the study of Biggeri et al. (2017). The calculated growth or quantum indexes for the selected sub-sectors are the dependent variables. After initial transformation variables were converted into log form. Quantum Index =Y(n) Y(0) ∗100 (1) Where, Y (n) =Real output in million Rupees in a year (n) at a constant factor cost of 1980–81 and Y (0) =Real output in million Rupees in base year 1980–81. For oil prices, the study used UK Brent crude oil prices which were USD/BL. Following the study of Lee and Chiu (2011), authors has converted them into real oil prices in PKR by multiplying 690 H. Yasmeen, Y. Wang, H. Zameer et al. / Energy Reports 5 (2019) 688–703 Table 1 Acronym and complete names of variables. Acronym Complete name FDI Foreign direct investment WPI Wholesale price index ROP Real oil prices MC Money in circulation GOVEXP Government expenditures REXR Real exchange rate GDEXP Government development expenditures LVSTKG Livestock sector growth MANUFG Manufacturing sector growth ELEC Electricity sector growth TRC Transportation and communication sector growth ARDL Autoregressive distributed lag ADF Augmented Dickey–Fuller PP Phillips–Perron CUSUM Cumulative Sum CUSUMSQ Cumulative Sum of Squares AGDP Agricultural gross domestic product GDP Gross domestic product them with exchange rate and then deflating them by consumer price index based on the 1980–81 price level. ROP =IOP ∗FEX CPI (2) Where, ROP =Real oil price in PKR, IOP =international oil price in USD/BL, FEX =foreign exchange in PKR/USD, and CPI = consumer price index based on the 1980–81 price level. The other variables used in the model as explanatory variables include wholesale price index, foreign direct investment, money in circulation, real foreign exchange rate and the government development expenditures. The data of these variables are also collected from ‘‘handbook of statistics on Pakistan economy 2015’’ that is available at the State Bank of Pakistan’s website and Pakistan Economic Survey 2016–17 available at the website of the ministry of finance. The detailed list of variables and their acronyms is given in Table 1. 2.2. Methods To discover the relation between oil price fluctuations and growth of economic sectors, the multifactor classical normal linear regression models have been estimated that are based on the open economy IS function for output or production. As in the succeeding section we will observe that the estimated models incorporate independent variables that are actually proxies for consumption, investment, government spending and international trade. As the basic objective is to explore the effects of oil price changes on sectors growth; each model will necessarily contain oil price as an explanatory variable, whereas the rest of the independent variables are incorporated to improve the model fitting. These independent variables include foreign direct investment, money in circulation, government expenditures and wholesale price index. Initially, the data was transformed in log form as a logarithm format of data provide better results. The selection of the independent variables is supported by previous studies (Hunt et al.,2002;Jo,2014;Khan and Ahmed,2011b;Malik,2008,2010;Montgomery,2017;Mussa, 2000;Syed,2010a). In preliminary models estimation by trial and error, we resolved the problem of multicollinearity without inducing specification bias by dropping out variables. The study employs ARDL model introduced by Pesaran et al. (2001a) to estimate the influence of oil price changes on sector growth. ARDL is an advanced approach and it has many advantages. Such as the technique of Engle and Granger (1987), this method is useful in estimating the relation among two variables. Whereas, when more than two variables are in the model, Johansen Cointegeration test (Johansen,1988) is used. Therefore, it can be argued that Johansen Cointegeration test has certain advantages over Engle–Granger technique. The study of Johansen and Juselius (1990) has extended the VAR (vector auto regression) model. But, this model is merely useful under specific conditions. First, it is merely used when large sample size data is under evaluation. Second, the precondition for co-integrated vector auto regression is that variables being estimated must have the same order integration. ARDL modeling technique not merely overcomes aforesaid problems but it also has numerous added benefits. ARDL approach is more appropriate as compare to other techniques in case of small size (Pesaran et al.,2001b). Further, ARDL technique can be applied even if variables are purely stationary at level I(0) or at first difference I(1) or the mixture of both I(0) and I(1) (Hasem and Pesaran,1997). The study of Laurenceson and Chai (2003) indicate that in data generating process ARDL technique can capture proper number of lags. Based on bound testing, error correction model can be obtained through transforming OLS. Without even losing the long-run information ECM show the adjustment mechanism of the model both in short-run and long-run (Pesaran and Shin,1999). ARDL technique makes the econometric model more dynamic. Ouattara (2004) argued that ARDL approach cannot be used if any variable being used in the model is stationary at second difference, as bound testing method is based on merely I(0), I(1) or mix of these. In this study, to ensure that all the variables under investigation in this study are stationary at I(0), I(1) or mix of these, we employed ADF (Dickey and Fuller,1979) and PP (Phillips and Perron,1988) unit root tests. Even though, ADF and PP methods are widely used in academic literature to explore unit root in data. But, many researchers in economics criticize that these tests have low power and give ambiguous results for unit root testing. The critic also debate that these tests do not have capability to report the evidence regarding structural breaks in the time series data. Therefore, to determine the reasonable outcomes, we followed the previous studies (Balcilar et al.,2017;Smith et al.,2019), and complemented unit root evaluation with Bai and Perron (2003) multiple structural breaks test. Basically, to analyze the long-run relation, one can use ARDL approach following two steps. These steps include, first, a researcher needs to analyze the presence of the long-run relation using F-statistic. If the value of F-statistics confirms the existence of cointegeration, then researcher can move forward and check and interpret the estimated coefficients for short-run and long-run. This technique postulates null hypothesis that there is no cointegeration among variables. According to Pesaran et al. (2001a), ARDL model reports the critical values for lower bounds and upper bounds. The variables are taken as I(0) and I(1) at lower and upper bound respectively. To conclude that cointegeration exists among the variables, the value of computed Fstatistic should be greater than upper bounds. In other words, null hypothesis is rejected and it can be inferred that long-run cointegeration exists. In contrast, if the calculated F-statistic value is below the lower bounds, it means, we could not found enough evidence to reject the null hypothesis, and cannot proceed with ARDL model. In a situation when the estimated value of F-statistic is found between the I(0) and I(1), this situation is regarded as an inconclusive. The critical bounds introduced by Pesaran et al. (2001a) are useful in case of large sample size. Ahmad and Du (2017) argued that it can provide biased results in the context of small sample size. The mechanism introduced by Narayan (2005) is useful for small size i.e. 30–80 observations. As the sample size of this study is between 30–80 observations, therefore, the study H. Yasmeen, Y. Wang, H. Zameer et al. / Energy Reports 5 (2019) 688–703 691 followed the mechanism introduced by Narayan (2005). In order to employ ARDL model for individual sectors, the following ECMs are being estimated. The mathematical representation of these ECMs is as follows. ∆LVSTKGt=β0+ n ∑ i=1 β1i∆FDIt−i+ n ∑ i=1 β2i∆WPIt−i + n ∑ i=1 β3i∆ROPt−i+ n ∑ i=1 β4i∆MCt−i+δ1∆FDIt−i +δ2∆WPIt−i+δ3∆ROPt−i+δ4∆MCt−i +dummy1999 +dummy2005 +µt(3) ∆FDIt=β0+ n ∑ i=1 β1i∆LVSTKGt−i+ n ∑ i=1 β2i∆WPIt−i + n ∑ i=1 β3i∆ROPt−i+ n ∑ i=1 β4i∆MCt−i +δ1∆LVSTKGt−i+δ2∆WPIt−i+δ3∆ROPt−i +δ4∆MCt−i+dummy1999 +dummy2005 +µt(4) ∆WPIt=β0+ n ∑ i=1 β1i∆FDIt−i+ n ∑ i=1 β2i∆LVSTKGt−i + n ∑ i=1 β3i∆ROPt−i+ n ∑ i=1 β4i∆MCt−i+δ1∆FDIt−i +δ2∆LVSTKGt−i+δ3∆ROPt−i+δ4∆MCt−i +dummy1999 +dummy2005 +µt(5) ∆ROPt=β0+ n ∑ i=1 β1i∆FDIt−i+ n ∑ i=1 β2i∆WPIt−i + n ∑ i=1 β3i∆LVSTKGt−i+ n ∑ i=1 β4i∆MCt−i +δ1∆FDIt−i+δ2∆WPIt−i+δ3∆LVSTKGt−i +δ4∆MCt−i+dummy1999 +dummy2005 +µt(6) ∆MCt=β0+ n ∑ i=1 β1i∆FDIt−i+ n ∑ i=1 β2i∆WPIt−i + n ∑ i=1 β3i∆ROPt−i+ n ∑ i=1 β4i∆LVSTKGt−i +δ1∆FDIt−i+δ2∆WPIt−i+δ3∆ROPt−i +δ4∆LVSTKGt−i+dummy1999 +dummy2005 +µt (7) ∆MANUFGt=β0+ n ∑ i=1 β1i∆ROPt−i+ n ∑ i=1 β2i∆WPIt−i + n ∑ i=1 β3i∆GDEXPt−i+ n ∑ i=1 β4i∆REXRt−i + n ∑ i=1 β5i∆MCt−i+β6∆ROPt−i+β7∆WPIt−i +β8∆GDEXPt−i+β9∆REXRt−i+β10∆MCt−i +dummy1999 +dummy2008 +µt(8) ∆ROPt=β0+ n ∑ i=1 β1i∆MANUFGt−i+ n ∑ i=1 β2i∆WPIt−i + n ∑ i=1 β3i∆GDEXPt−i+ n ∑ i=1 β4i∆REXRt−i + n ∑ i=1 β5i∆MCt−i+β6∆MANUFGt−i+β7∆WPIt−i +β8∆GDEXPt−i+β9∆REXRt−i+β10∆MCt−i +dummy1999 +dummy2008 +µt(9) ∆WPIt=β0+ n ∑ i=1 β1i∆ROPt−i+ n ∑ i=1 β2i∆MANUFGt−i + n ∑ i=1 β3i∆GDEXPt−i+ n ∑ i=1 β4i∆REXRt−i + n ∑ i=1 β5i∆MCt−i+β6∆ROPt−i+β7∆MANUFGt−i +β8∆GDEXPt−i+β9∆REXRt−i+β10∆MCt−i +dummy1999 +dummy2008 +µt(10) ∆GDEXPt=β0+ n ∑ i=1 β1i∆ROPt−i+ n ∑ i=1 β2i∆WPIt−i + n ∑ i=1 β3i∆MANUFGt−i+ n ∑ i=1 β4i∆REXRt−i + n ∑ i=1 β5i∆MCt−i+β6∆ROPt−i+β7∆WPIt−i +β8∆MANUFGt−i+β9∆REXRt−i+β10∆MCt−i +dummy1999 +dummy2008 +µt(11) ∆REXRt=β0+ n ∑ i=1 β1i∆ROPt−i+ n ∑ i=1 β2i∆WPIt−i + n ∑ i=1 β3i∆GDEXPt−i+ n ∑ i=1 β4i∆MANUFGt−i + n ∑ i=1 β5i∆MCt−i+β6∆ROPt−i+β7∆WPIt−i +β8∆GDEXPt−i+β9∆MANUFGt−i+β10∆MCt−i +dummy1999 +dummy2008 +µt(12) ∆MCt=β0+ n ∑ i=1 β1i∆ROPt−i+ n ∑ i=1 β2i∆WPIt−i + n ∑ i=1 β3i∆GDEXPt−i+ n ∑ i=1 β4i∆REXRt−i + n ∑ i=1 β5i∆MANUFGt−i+β6∆ROPt−i+β7∆WPIt−i +β8∆GDEXPt−i+β9∆REXRt−i+β10∆MANUFGt−i +dummy1999 +dummy2008 +µt(13) ∆ELECt=β0+ n ∑ i=1 β1i∆ROPt−i+ n ∑ i=1 β2i∆WPIt−i + n ∑ i=1 β3i∆FDIt−i+ n ∑ i=1 β4i∆MCt−i + n ∑ i=1 β5i∆GOVEXPt−i+β6∆ROPt−i+β7∆WPIt−i +β8∆FDIt−i+β9∆MCt−i 692 H. Yasmeen, Y. Wang, H. Zameer et al. / Energy Reports 5 (2019) 688–703 +β10∆GOVEXPt−i+dummy2012 +µt(14) ∆ROPt=β0+ n ∑ i=1 β1i∆ELECt−i+ n ∑ i=1 β2i∆WPIt−i + n ∑ i=1 β3i∆FDIt−i+ n ∑ i=1 β4i∆MCt−i + n ∑ i=1 β5i∆GOVEXPt−i+β6∆ELECt−i+β7∆WPIt−i +β8∆FDIt−i+β9∆MCt−i +β10∆GOVEXPt−i+dummy2012 +µt(15) ∆WPIt=β0+ n ∑ i=1 β1i∆ROPt−i+ n ∑ i=1 β2i∆ELECt−i + n ∑ i=1 β3i∆FDIt−i+ n ∑ i=1 β4i∆MCt−i + n ∑ i=1 β5i∆GOVEXPt−i+β6∆ROPt−i+β7∆ELECt−i +β8∆FDIt−i+β9∆MCt−i +β10∆GOVEXPt−i+dummy2012 +µt(16) ∆FDIt=β0+ n ∑ i=1 β1i∆ROPt−i+ n ∑ i=1 β2i∆WPIt−i + n ∑ i=1 β3i∆ELECt−i+ n ∑ i=1 β4i∆MCt−i + n ∑ i=1 β5i∆GOVEXPt−i+β6∆ROPt−i+β7∆WPIt−i +β8∆ELECt−i+β9∆MCt−i +β10∆GOVEXPt−i+dummy2012 +µt(17) ∆MCt=β0+ n ∑ i=1 β1i∆ROPt−i+ n ∑ i=1 β2i∆WPIt−i + n ∑ i=1 β3i∆FDIt−i+ n ∑ i=1 β4i∆ELECt−i + n ∑ i=1 β5i∆GOVEXPt−i+β6∆ROPt−i+β7∆WPIt−i +β8∆FDIt−i+β9∆ELECt−i +β10∆GOVEXPt−i+dummy2012 +µt(18) ∆GOVEXPt=β0+ n ∑ i=1 β1i∆ROPt−i+ n ∑ i=1 β2i∆WPIt−i + n ∑ i=1 β3i∆FDIt−i+ n ∑ i=1 β4i∆MCt−i + n ∑ i=1 β5i∆ELECt−i+β6∆ROPt−i+β7∆WPIt−i +β8∆FDIt−i+β9∆MCt−i +β10∆ELECt−i+dummy2012 +µt(19) TRCt=β0+ n ∑ i=1 β1i∆WPIt−i+ n ∑ i=1 β2i∆ROPt−i + n ∑ i=1 β3i∆FDIt−i+ n ∑ i=1 β4i∆MCt−i +β5∆WPIt−i+β6∆ROPt−i+β7∆FDIt−i +β8∆MCt−i+dummy2005 +µt(20) ∆WPIt=β0+ n ∑ i=1 β1i∆TRCt−i+ n ∑ i=1 β2i∆ROPt−i + n ∑ i=1 β3i∆FDIt−i+ n ∑ i=1 β4i∆MCt−i +β5∆TRCt−i+β6∆ROPt−i+β7∆FDIt−i +β8∆MCt−i+dummy2005 +µt(21) ROPt=β0+ n ∑ i=1 β1i∆WPIt−i+ n ∑ i=1 β2i∆TRCt−i + n ∑ i=1 β3i∆FDIt−i+ n ∑ i=1 β4i∆MCt−i +β5∆WPIt−i+β6∆TRCt−i+β7∆FDIt−i +β8∆MCt−i+dummy2005 +µt(22) FDIt=β0+ n ∑ i=1 β1i∆WPIt−i+ n ∑ i=1 β2i∆ROPt−i + n ∑ i=1 β3i∆TRCt−i+ n ∑ i=1 β4i∆MCt−i +β5∆WPIt−i+β6∆ROPt−i+β7∆TRCt−i +β8∆MCt−i+dummy2005 +µt(23) MCt=β0+ n ∑ i=1 β1i∆WPIt−i+ n ∑ i=1 β2i∆ROPt−i + n ∑ i=1 β3i∆FDIt−i+ n ∑ i=1 β4i∆TRCt−i +β5∆WPIt−i+β6∆ROPt−i+β7∆FDIt−i +β8∆TRCt−i+dummy2005 +µt(24) In Eq. (3), the operator β0is constant, β1−β4are used as error correction dynamics in the model. Dummy variables are added into the model according to the structural breaks identified using Bai and Perron (2003) multiple structural break unit root test. The operator µtindicate white noise error-term in the model. In the second part of Eq. (3), the operator δ1−δ4represent the long-run association among variables. ARDL model is based on the value of Wald F-statistic which shows the long-run cointegeration among variables with a null of no cointegeration as H0:δ1=δ2=δ3= δ4=0. And, the alternative H1:δ1#δ2#δ3#δ4#0. Following the same mechanism, the other Eqs. (4)–(24) can be explained. After the evaluations on the long-run relation among variables (using F-statistic), application of model and finding the long-run coefficients (from bound testing), and the next step is to find short-run coefficients. Therefore, to find the short-run relations, the following short-run models are being estimated: ∆LVSTKGt=β0+ n ∑ i=1 β1i∆FDIt−i+ n ∑ i=1 β2i∆WPIt−i + n ∑ i=1 β3i∆ROPt−i+ n ∑ i=1 β4i∆MCt−i +η1ECTt−i+dummy1999 +dummy2005 +µt(25) H. Yasmeen, Y. Wang, H. Zameer et al. / Energy Reports 5 (2019) 688–703 693 ∆FDIt=β0+ n ∑ i=1 β1i∆LVSTKGt−i+ n ∑ i=1 β2i∆WPIt−i + n ∑ i=1 β3i∆ROPt−i+ n ∑ i=1 β4i∆MCt−i+η1ECTt−i +dummy1999 +dummy2005 +µt(26) ∆WPIt=β0+ n ∑ i=1 β1i∆FDIt−i+ n ∑ i=1 β2i∆LVSTKGt−i + n ∑ i=1 β3i∆ROPt−i+ n ∑ i=1 β4i∆MCt−i+η1ECTt−i +dummy1999 +dummy2005 +µt(27) ∆ROPt=β0+ n ∑ i=1 β1i∆FDIt−i+ n ∑ i=1 β2i∆WPIt−i + n ∑ i=1 β3i∆LVSTKGt−i+ n ∑ i=1 β4i∆MCt−i+η1ECTt−i +dummy1999 +dummy2005 +µt(28) ∆MCt=β0+ n ∑ i=1 β1i∆FDIt−i+ n ∑ i=1 β2i∆WPIt−i + n ∑ i=1 β3i∆ROPt−i+ n ∑ i=1 β4i∆LVSTKGt−i+η1ECTt−i +dummy1999 +dummy2005 +µt(29) ∆MANUFGt=β0+ n ∑ i=1 β1i∆ROPt−i+ n ∑ i=1 β2i∆WPIt−i + n ∑ i=1 β3i∆GDEXPt−i+ n ∑ i=1 β4i∆REXRt−i + n ∑ i=1 β5i∆MCt−i+η2ECTt−i+dummy1999 +dummy2008 +µt(30) ∆ROPt=β0+ n ∑ i=1 β1i∆MANUFGt−i+ n ∑ i=1 β2i∆WPIt−i + n ∑ i=1 β3i∆GDEXPt−i+ n ∑ i=1 β4i∆REXRt−i + n ∑ i=1 β5i∆MCt−i+η2ECTt−i+dummy1999 +dummy2008 +µt(31) ∆WPIt=β0+ n ∑ i=1 β1i∆ROPt−i+ n ∑ i=1 β2i∆MANUFGt−i + n ∑ i=1 β3i∆GDEXPt−i+ n ∑ i=1 β4i∆REXRt−i + n ∑ i=1 β5i∆MCt−i+η2ECTt−i+dummy1999 +dummy2008 +µt(32) ∆GDEXPt=β0+ n ∑ i=1 β1i∆ROPt−i+ n ∑ i=1 β2i∆WPIt−i + n ∑ i=1 β3i∆MANUFGt−i+ n ∑ i=1 β4i∆REXRt−i + n ∑ i=1 β5i∆MCt−i+η2ECTt−i+dummy1999 +dummy2008 +µt(33) ∆REXRt=β0+ n ∑ i=1 β1i∆ROPt−i+ n ∑ i=1 β2i∆WPIt−i + n ∑ i=1 β3i∆GDEXPt−i+ n ∑ i=1 β4i∆MANUFGt−i + n ∑ i=1 β5i∆MCt−i+η2ECTt−i+dummy1999 +dummy2008 +µt(34) ∆MCt=β0+ n ∑ i=1 β1i∆ROPt−i+ n ∑ i=1 β2i∆WPIt−i + n ∑ i=1 β3i∆GDEXPt−i+ n ∑ i=1 β4i∆REXRt−i + n ∑ i=1 β5i∆MANUFGt−i+η2ECTt−i+dummy1999 +dummy2008 +µt(35) ∆ELECt=β0+ n ∑ i=1 β1i∆ROPt−i+ n ∑ i=1 β2i∆WPIt−i + n ∑ i=1 β3i∆FDIt−i+ n ∑ i=1 β4i∆MCt−i + n ∑ i=1 β5i∆GOVEXPt−i+η3ECTt−i +dummy2012 +µt(36) ∆ROPt=β0+ n ∑ i=1 β1i∆ELECt−i+ n ∑ i=1 β2i∆WPIt−i + n ∑ i=1 β3i∆FDIt−i+ n ∑ i=1 β4i∆MCt−i + n ∑ i=1 β5i∆GOVEXPt−i+η3ECTt−i +dummy2012 +µt(37) ∆WPIt=β0+ n ∑ i=1 β1i∆ROPt−i+ n ∑ i=1 β2i∆ELECt−i + n ∑ i=1 β3i∆FDIt−i+ n ∑ i=1 β4i∆MCt−i + n ∑ i=1 β5i∆GOVEXPt−i+η3ECTt−i +dummy2012 +µt(38) ∆FDIt=β0+ n ∑ i=1 β1i∆ROPt−i+ n ∑ i=1 β2i∆WPIt−i + n ∑ i=1 β3i∆ELECt−i+ n ∑ i=1 β4i∆MCt−i 694 H. Yasmeen, Y. Wang, H. Zameer et al. / Energy Reports 5 (2019) 688–703 + n ∑ i=1 β5i∆GOVEXPt−i+η3ECTt−i +dummy2012 +µt(39) ∆MCt=β0+ n ∑ i=1 β1i∆ROPt−i+ n ∑ i=1 β2i∆WPIt−i + n ∑ i=1 β3i∆FDIt−i+ n ∑ i=1 β4i∆ELECt−i + n ∑ i=1 β5i∆GOVEXPt−i+η3ECTt−i +dummy2012 +µt(40) ∆GOVEXPt=β0+ n ∑ i=1 β1i∆ROPt−i+ n ∑ i=1 β2i∆WPIt−i + n ∑ i=1 β3i∆FDIt−i+ n ∑ i=1 β4i∆MCt−i + n ∑ i=1 β5i∆ELECt−i+η3ECTt−i +dummy2012 +µt(41) ∆TRCt=β0+ n ∑ i=1 β1i∆WPIt−i+ n ∑ i=1 β2i∆ROPt−i + n ∑ i=1 β3i∆FDIt−i+ n ∑ i=1 β4i∆MCt−i +η4ECTt−i+dummy2005 +µt(42) ∆WPIt=β0+ n ∑ i=1 β1i∆TRCt−i+ n ∑ i=1 β2i∆ROPt−i + n ∑ i=1 β3i∆FDIt−i+ n ∑ i=1 β4i∆MCt−i +η4ECTt−i+dummy2005 +µt(43) ∆ROPt=β0+ n ∑ i=1 β1i∆WPIt−i+ n ∑ i=1 β2i∆TRCt−i + n ∑ i=1 β3i∆FDIt−i+ n ∑ i=1 β4i∆MCt−i +η4ECTt−i+dummy2005 +µt(44) ∆FDIt=β0+ n ∑ i=1 β1i∆WPIt−i+ n ∑ i=1 β2i∆ROPt−i + n ∑ i=1 β3i∆TRCt−i+ n ∑ i=1 β4i∆MCt−i +η4ECTt−i+dummy2005 +µt(45) ∆MCt=β0+ n ∑ i=1 β1i∆WPIt−i+ n ∑ i=1 β2i∆ROPt−i + n ∑ i=1 β3i∆FDIt−i+ n ∑ i=1 β4i∆TRCt−i +η4ECTt−i+dummy2005 +µt(46) Here, in Eq. (25), the mathematical form of short-run model is presented for livestock sector. The ECT is used for error correction term, which is basically used to explain if there is a disturbance in the model, how much time it will take to reach back to its equilibrium path in the long-run. Whereas, η1is used to represent the coefficient of error correction term. Dummy is used to incorporate structural breaks in the model. Lastly, the stability of the coefficients in the short-run and long-run is checked using CUSUM and CUSUMSQ introduced by Brown et al. (1975). Following the same mechanism, the other Eqs. (26)–(46) can be explained. 3. Results and findings This section provides the detailed estimation and discussion of results from unit root testing, structural breaks exploration and how selected exploratory variables influence the growth of four selected sector of the economy. The mathematical models designed in the previous part were applied to test unit root, structural breaks and to get short-run and long-run coefficients. In this part, we elaborated in detail about the unit root testing, structural breaks, model appropriateness, residuals normality and stability, heteroscedasticity and autocorrelation. Following this, we will discuss how oil price fluctuations influence the individual sector of the economy. Initially, ADF and PP unit root test were employed to check the stationarity of data being used for the evaluation of the effects of oil price fluctuations on the real sector growth. The results from ADF and PP test are presented in Table 2. The results from unit root testing from both tests (i.e.) confirms that all variable under evaluation are stationary at I(0), I(1) and none of the variable is stationary at I(2). Hence, it confirms the precondition of ARDL approach that all the variable must be stationary at I(0), I(1) or mix of these. As discussed in the previous part that the critics of ADF and PP test debate that these tests do not have capability to report structural breaks. Therefore, to determine the reasonable outcomes, we followed the studies of (Balcilar et al.,2017;Smith et al., 2019) and used Bai and Perron (2003) multiple structural breaks evaluation test. Using this test, we identified multiple structural breaks in the variables followed by the structural breaks in our models. Table 3 summarize the results from Bai and Perron (2003) multiple structural breaks test. It can be seen that all the independent variables contains structural breaks. Following the exploration of structural breaks in the variables, we used the same mechanism of Bai and Perron (2003) to find structural breaks in the model. The results presented in Table 4 suggest that for every model we need to incorporate structural breaks. For livestock sector and manufacturing sector we need to incorporate two dummy variables. Whereas, for electricity and transportation & communication sector, the results from Bai and Perron (2003) multiple structural breaks exploration test indicate that we need to add one dummy variable in the model. 3.1. Oil price fluctuations and livestock sector growth The livestock sector in Pakistan contributes about 56.3% of the total value of agriculture and almost 11% toward AGDP (agricultural gross domestic product) (Rehman et al.,2017). Milk is the single important commodity of Pakistan’s livestock sector. Pakistan is 4th largest producer of milk worldwide after USA, China and India. Due to overall contribution of livestock sector toward agricultural GDP, it is considered as most significant sector of agriculture based economy like Pakistan. According to Bettencourt et al. (2015) the livestock sector plays a significant role in wellbeing of rural households. This sector helps family nutrition, family income, food supply, soil productivity, asset savings and agricultural traction and diversification (Moyo and Swanepoel, 2010). It is widely acknowledged that energy and agricultural H. Yasmeen, Y. Wang, H. Zameer et al. / Energy Reports 5 (2019) 688–703 695 Table 2 Summary of unit root testing. Source: Authors’ estimation using E-Views 10. Variables Augmented Dickey–Fuller Phillips–Perron I(0) I(1) I(0) I(1) C C&T C C&T C C&T C C&T FDI −2.3098 −2.8005 −5.2304 −5.2620 −2.2709 −2.9507 −5.2257 −5.2628 WPI −0.3653 −4.8847 −4.0626 −4.0130 −0.3711 −2.4540 −5.4953 −5.4381 ROP −1.7946 −1.9252 −6.1472 −6.0918 −1.8319 −2.0090 −6.1449 −6.0857 MC −1.9446 −3.7533 −7.4198 −7.8303 −2.1780 −4.1628 −7.4179 −7.8303 GOVEXP −1.3075 −2.8391 −8.4684 −8.4244 −1.4178 −2.7438 −8.5268 −8.4934 REXR −0.7418 −0.9864 −4.4984 −3.7117 −0.7015 −1.5421 −4.4953 −4.4481 GDEXP 0.8549 −0.9329 −5.1321 −5.2487 0.6069 −1.3066 −5.1795 −5.2801 LVSTKG −3.6387 −6.1912 −6.4503 −6.5775 −3.5729 −6.1926 −24.917 −32.633 MANUFG −4.4324 −4.9212 −9.2713 −9.1486 −4.4077 −4.8946 −29.010 −28.570 ELEC −5.7727 0.9949 −7.0853 −7.6497 −5.9301 −5.8075 −24.565 −28.923 TRC −3.7459 −5.6948 −6.7888 −6.7112 −3.7540 −5.6952 −19.040 −19.580 Test critical values 1% level −3.6009 −4.1985 −3.6267 −4.2349 −3.6009 −4.1985 −3.6055 −4.2050 5% level −2.9350 −3.5236 −2.9458 −3.5403 −2.9350 −3.5236 −2.9369 −3.5266 10% level −2.6058 −3.1929 −2.6115 −3.2024 −2.6058 −3.1929 −2.6068 −3.1946 Table 3 Summary of Bai–Perron structural breaks testing. Source: Authors’ estimation using E-Views 10. Variables F-statistic Critical value** Break dates FDI 25.06278 11.14 1985, 1992, 2004 WPI 21.44766 11.83 1987, 1995, 2002, 2008 ROP 60.35027 8.58 2004 MC 33.12461 11.83 1984, 1995, 2001, 2007 GOVEXP 44.39183 11.14 1987, 1999, 2006 REXR 64.30310 11.14 1985, 1996, 2009 GDEXP 52.80955 11.14 1983, 1991, 2011 LVSTKG 36.72945 8.58 1999 MANUFG 13.77003 8.58 1999 ELEC – – – TRC 43.55637 8.58 1999 * Significant at the 0.05 level. ** Bai–Perron (Econometric Journal, 2003) critical values. commodity markets are closely integrated due to the expansion of biofuel sector (Fabiosa,2009). The major impact of said integration is on grains such as corn which is primary feedstock. So, the changes in feed cost will affect livestock sector. As 59% of total cost belongs to feed cost (Fabiosa,2009). The study of Patton et al. (2012) also indicated the association of oil prices and livestock sector performance. Most of the previous studies in the context of oil price and livestock sector performance have been done in context of developed countries, but no study has been done in context of developing country like Pakistan. First time, this study is used to explore the impact of oil price fluctuations on the growth of livestock sector. The study has developed ARDL model to estimate the influence of oil price fluctuations on the livestock sector. The diagnostic test and estimated coefficients for short-run and long-run are given in the subsequent part. Prior to the discussion of short-run and long-run coefficients, it is vital to explore the structural breaks in the data if any. Following this, one would be in a better position to elaborate Table 5 Diagnostic test results for estimated model (1). Source: Authors’ estimation using E-Views 10. Diagnostic test Statistics R20.944517 F-statistic 22.34341 (0.000000) Durbin–Watson‘ 1.277087 Serial Correlation 1.637345 (0.2108) Normality 3.317675 (0.190360) Heteroscedasticity 0.853680 (0.6217) Note: The value in the parenthesis is p-value. Jarque–Bera test null is normality, Breauch–Godfrey serial-correlation LM test null is no serial correlation, Breusch–Pagan–Godfrey heteroscedasticity null is no heteroscedasticity. the results of model goodness of fit. Using the multiple structural breaks test of Bai and Perron (2003), we found that there are two structural breaks in the model. One structural break is found during year 1999 and other during year 2005. Results indicated that both structural breaks has significantly influenced our model. Therefore, to incorporate the influence of said structural breaks, we have added two dummy variables in the model. Table 5 provide summary of the results from model goodness of fit. With reference to diagnostic test results, it can be seen from Table 5 that different stability and diagnostic test were performed to confirm the goodness of fit of the estimated model. These tests include overall functional form of the model, heteroscedasticity, normality, distribution of residuals and serial correlation. The stability coefficients were analyzed through CUSUM and CUSUMSQ introduced by Brown et al. (1975). The value of R2and Durban– Watson test indicate the overall goodness of fit of the model. These values indicate that in overall evaluation, the model fitting is appropriate. The residuals follow a normal distribution, and the calculated Jarque–Bera test in insignificant which confirms the normality. The Breusch–Pagan–Godfrey test was used to check Table 4 Identification of structural breaks in the models. Source: Authors’ estimation using E-Views 10. Sector model F-statistic Critical value** Break dates Livestock Sector 13.59159 18.11 1999, 2005 Manufacturing Sector 10.67932 19.91 1999, 2008 Electricity Sector 4.383707 18.23 2012 Transportation and communication sector 4.190507 16.19 2005 * Significant at the 0.05 level. ** Bai–Perron(Econometric Journal, 2003) critical values. 702 H. Yasmeen, Y. Wang, H. Zameer et al. / Energy Reports 5 (2019) 688–703 importing nations; consumption of oil exporting countries goes unharmed during global recessions caused by oil prices. Further, the government shall take initiatives to diversify the country’s energy mix with greater emphasis on renewable energy sources. It shall invest in exploring and utilizing in house energy resources like natural gas, coal and crude oil. All such government initiatives will help in reducing energy wastage and decrease reliance on imported fuel, which will consequently strengthen the sustainable supply of energy for industry, reduce the outflow of foreign exchange and will immune the real sector from the detrimental impact of oil price shocks. Thirdly, the negative impact of government development expenditures gives an alarming signal to the policy makers because all the development expenditures are being consumed on the projects those are not necessary for the growth of manufacturing sector. The government is putting a huge amount of money in building roads & bridges, irrigation and urban development & transport. Even though these projects are important but, these are not the necessity of manufacturing sector. Thus, it is suggested that the policy makers needs to divert their attention and build projects like electricity generation from cheap energy sources that is the need of manufacturing and electricity sector, and it will give some relief to these sectors. Finally, the study also found that increase in money supply adversely affects most of the sectors. An increase in money supply means that money demand is being met which brings down the interest rates, and lending becomes less profitable, as a result, the sector’s growth is negatively affected by increasing quantity of money in the circulation stream. Moreover, the increased money supply also minimizes the options for firm saving and investment because the interest rates in the economy go down. Thus, based upon the findings, the study provides policy suggestion to the government to control the money supply to improve the sectors growth. Funding This study is supported by the National Natural Science Foundation of China (Grant No. 71873064) and the General Projects of Humanities and Social Sciences of the Ministry of Education, China (Grant No. 18YJA790085). Declaration of competing interest The authors declare no conflict of interest. References Ahmad, N., Du, L., 2017. Effects of energy production and CO2 emissions on economic growth in Iran: ARDL approach. Energy 123, 521–537. Ahmed, K., Bhutto, N.A., Kalhoro, M.R., 2018. Decomposing the links between oil price shocks and macroeconomic indicators: Evidence from SAARC region. Resources policy. Ang, J.B., 2009. Financial development and the FDI-growth nexus: the Malaysian experience. Appl. Econom. 41 (13), 1595–1601. Azman-Saini, W., Law, S.H., Ahmad, A.H., 2010. FDI And economic growth: New evidence on the role of financial markets. Econom. Lett. 107 (2), 211–213. Bai, J., Perron, P., 2003. Critical values for multiple structural change tests. Econom. J. 6 (1), 72–78. Balcilar, M., Bouri, E., Gupta, R., Roubaud, D., 2017. Can volume predict bitcoin returns and volatility? A quantiles-based approach. Econ. Model. 64, 74–81. Bastianin, A., Conti, F., Manera, M., 2016. The impacts of oil price shocks on stock market volatility: Evidence from the G7 countries. Energy Policy 98, 160–169. Bernanke, B., 2016. The relationship between stocks and oil prices. Ben Bernanke’s Blog on Brookings posted on February 19 (2016). Bettencourt, E.M.V., Tilman, M., Narciso, V., Carvalho, M.L.d.S., Henriques, P.D.d.S., 2015. The livestock roles in the wellbeing of rural communities of Timor-Leste. Rev. Econom. Sociol. Rural 53, 63–80. Biggeri, L., Ferrari, G., Zhao, Y., 2017. Estimating cross province and municipal city price level differences in China: Some experiments and results. Soc. Indicat. Res. 131 (1), 169–187. Brown, R.L., Durbin, J., Evans, J.M., 1975. Techniques for testing the constancy of regression relationships over time. J. R. Stat. Soc. Ser. B Stat. Methodol. 149–192. Chowdhury, A., Mavrotas, G., 2006. FDI And growth: What causes what?. World Economy 29 (1), 9–19. Dickey, D.A., Fuller, W.A., 1979. Distribution of the estimators for autoregressive time series with a unit root. J. Am. Statist. Associat. 74 (366a), 427–431. Engle, R.F., Granger, C.W., 1987. Co-integration and error correction: representation, estimation, and testing. Econom.: J. Econom. Soc. 251–276. Fabiosa, J.F., 2009. The Impact of the Crude Oil Price on the Livestock Sector under a Regime of Integrated Energy and Grain Markets. In: Paper presented at the Agricultural & Applied Economics Association 2009 AAEA & ACCI Joint Annual Meeting. Milwaukee (USA) Cerca con Google. Finance, M.o., 2018. Pakistan Economic Survey 2016-17 Retrieved 25-November2018, from . http://www.finance.gov.pk/survey/chapters_17/Pakistan_ES_ 2016_17_pdf.pdf. Hamilton, J.D., 2009. Causes and Consequences of the Oil Shock of 2007-08. National Bureau of Economic Research. Hasem, P.M., Pesaran, B., 1997. Working with Microfit 4.0: Interactive Econometric Analysis. Oxford University Press. Hunt, B., Isard, P., Laxton, D., 2002. The macroeconomic effects of higher oil prices. Nat. Inst. Econom. Rev. 179 (1), 87–103. Iamsiraroj, S., Ulubaşoğlu, M.A., 2015. Foreign direct investment and economic growth: A real relationship or wishful thinking?. Econ. Model. 51, 200–213. Islam, F., Shahbaz, M., Ahmed, A.U., Alam, M.M., 2013. Financial development and energy consumption nexus in Malaysia: a multivariate time series analysis. Econ. Model. 30, 435–441. Jawad, M., 2013. Oil price volatility and its impact on economic growth in Pakistan. J. Financ. Econom. 1 (4), 62–68. Jo, S., 2014. The effects of oil price uncertainty on global real economic activity. J. Money Credit Bank. 46 (6), 1113–1135. Johansen, S., 1988. Statistical analysis of cointegration vectors. J. Econom. Dyn. Control 12 (2–3), 231–254. Johansen, S., Juselius, K., 1990. Maximum likelihood estimation and inference on cointegration—with applications to the demand for money. Oxford Bull. Econom. Statist. 52 (2), 169–210. Kang, W., de Gracia, F.P., Ratti, R.A., 2017. Oil price shocks, policy uncertainty, and stock returns of oil and gas corporations. J. Int. Money Financ. 70, 344–359. Kang, W., Ratti, R.A., 2015. Oil shocks, policy uncertainty and stock returns in China. Econom. Trans. 23 (4), 657–676. Kessides, I.N., 2013. Chaos in power: Pakistan’s electricity crisis. Energy Policy 55, 271–285. Khan, M.A., Ahmad, U., 2008. Energy demand in Pakistan: a disaggregate analysis. Pak. Develop. Rev. 437–455. Khan, M.A., Ahmed, A., 2011b. Macroeconomic effects of global food and oil price shocks to the Pakistan economy: A structural vector autoregressive (SVAR) analysis. Pak. Develop. Rev. 491–511. Kilian, L., 2008. Exogenous oil supply shocks: how big are they and how much do they matter for the US economy?. Rev. Econ. Stat. 90 (2), 216–240. Kilian, L., Vigfusson, R.J., 2011. Are the responses of the US economy asymmetric in energy price increases and decreases?. Quant. Econom. 2 (3), 419–453. Komal, R., Abbas, F., 2015. Linking financial development, economic growth and energy consumption in Pakistan. Renew. Sustain. Energy Rev. 44, 211–220. Laurenceson, J., Chai, J.C., 2003. Financial Reform and Economic Development in China. Edward Elgar Publishing. Lee, C.-C., Chiu, Y.-B., 2011. Oil prices, nuclear energy consumption, and economic growth: New evidence using a heterogeneous panel analysis. Energy Policy 39 (4), 2111–2120. Lescaroux, F., Mignon, V., 2008. On the influence of oil prices on economic activity and other macroeconomic and financial variables. OPEC Energy Rev. 32 (4), 343–380. Malik, A., 2008. Crude oil price, monetary policy and output: The case of Pakistan. Pak. Develop. Rev. 425–436. Malik, A., 2010. Oil prices and economic activity in Pakistan. South Asia Econom. J. 11 (2), 223–244. Marza, M., Shaaibith, S.J., Daly, S.S., 2018. Impact of oil price fluctuations on human development: A standard study of Iraq. J. Soc. Sci. Res. 395, 396–399. Montgomery, H., 2017. Energy Price Shocks and Macroeconomic Performance. Routledge. Moyo, S., Swanepoel, F., 2010. Multifunctionality of livestock in developing communities. The role of livestock in developing communities: Enhancing multifunctionality. p. 1. Mussa, M., 2000. The Impact of Higher Oil Prices on the Global Economy. International Monetary Fund, p. 8. Nandha, M., Hammoudeh, S., 2007. Systematic risk, and oil price and exchange rate sensitivities in Asia-Pacific stock markets. Res. Int. Bus. Finance 21 (2), 326–341. H. Yasmeen, Y. Wang, H. Zameer et al. / Energy Reports 5 (2019) 688–703 703 Narayan, P.K., 2005. The saving and investment nexus for China: evidence from cointegration tests. Appl. Econom. 37 (17), 1979–1990. Narayan, P.K., Sharma, S., Poon, W.C., Westerlund, J., 2014. Do oil prices predict economic growth? New global evidence. Energy Econom. 41, 137–146. Naser, H., 2019. Oil price fluctuation, gold returns and inflationary pressure: An empirical analysis using cointegration approach. Appl. Econom. Financ. 6 (2), 71–78. Nazlioglu, S., Gormus, A., Soytas, U., 2019. Oil prices and monetary policy in emerging markets: Structural shifts in causal linkages. Emerg. Market. Financ. Trade 55 (1), 105–117. Obstfeld, M., Milesi-Ferretti, G., Arezki, R., 2016. Oil prices and the global economy: It’s Complicated. iMFdirect (short version) and VoxEU (long version). Ouattara, B., 2004. The impact of project aid and programme aid inflows on domestic savings: A case study of Cô In: Paper presented at the Centre for the Study of African Economies Conference on Growth, Poverty Reduction and Human Development in Africa. Pakistan, S.B.o., 2017. Handbook of statistics on Pakistan economy 2015. Retrieved 25-October-2018, 2018, from http://www.sbp.org.pk/departments/ stats/PakEconomy_HandBook/index.htm. Patton, M., Binfield, J.C., Kim, I.S., Zhang, L., Davis, J., 2012. Linkages between the energy, biofuel and agricultural sectors. In: Paper presented at the 86th Annual Conference, April (2012), pp. 16–18, Warwick University, Coventry, UK. Pesaran, H., Shin, Y., 1999. An Autoregressive Distributed Lag Modelling Approach to Cointegration chapter 11 Paper presented at the Econometrics and Economic Theory in the 20th Century: The Ragnar Frisch Centennial Symposium. Pesaran, Shin, Y., Smith, R.J., 2001a. Bounds testing approaches to the analysis of level relationships. J. Appl. Econom. 16 (3), 289–326. Pesaran, M.H., Shin, Y., Smith, R.J., 2001b. Bounds testing approaches to the analysis of level relationships. J. Appl. Econom. 16 (3), 289–326. Phillips, P.C., Perron, P., 1988. Testing for a unit root in time series regression. Biometrika 75 (2), 335–346. Pönkä, H., Zheng, Y., 2019. The Role of Oil Prices on the Russian Business Cycle. Research in International Business and Finance. Pradhan, R.P., Arvin, M.B., Nair, M., Mittal, J., Norman, N.R., 2017. Telecommunications infrastructure and usage and the FDI–growth nexus: evidence from Asian-21 countries. Inf. Technol. Dev. 23 (2), 235–260. Regan, A., Brazys, S., 2018. Celtic phoenix or leprechaun economics? The politics of an FDI-led growth model in Europe. New Political Economy 23 (2), 223–238. Rehman, A., Jingdong, L., Chandio, A.A., Hussain, I., 2017. Livestock production and population census in Pakistan: Determining their relationship with agricultural GDP using econometric analysis. Inform. Process. Agricul. 4 (2), 168–177. Sadorsky, P., 2013. Do urbanization and industrialization affect energy intensity in developing countries?. Energy Econom. 37, 52–59. Sarwar, S., Shahbaz, M., Anwar, A., Tiwari, A.K., 2019. The importance of oil assets for portfolio optimization: The analysis of firm level stocks. Energy Econom. 78, 217–234. Shahbaz, M., Chaudhary, A., Ozturk, I., 2017. Does urbanization cause increasing energy demand in Pakistan? Empirical evidence from STIRPAT model. Energy 122, 83–93. Smith, S.C., Timmermann, A., Zhu, Y., 2019. Variable selection in panel models with breaks. J. Econometrics. Solangi, Y., Tan, Q., Khan, M., Mirjat, N., Ahmed, I., 2018. The selection of wind power project location in the southeastern corridor of Pakistan: A factor analysis, AHP, and Fuzzy-TOPSIS application. Energies 11 (8), 1940. Syed, N.I., 2010a. Measuring the impact of changing oil prices and other macro economic variables on GDP in the context of Pakistan’s economy. Int. Res. J. Financ. Econom. 52, 40–49. Waheed, R., Wei, C., Sarwar, S., Lv, Y., 2018. Impact of oil prices on firm stock return: industry-wise analysis. Emp. Econom. 55 (2), 765–780. Wakeel, M., Chen, B., Jahangir, S., 2016. Overview of energy portfolio in Pakistan. Energy Procedia 88, 71–75. Wesseh Jr, P.K., Lin, B., 2018. Exchange rate fluctuations, oil price shocks and economic growth in a small net-importing economy. Energy 151, 402–407. Zaman, K., Abdullah, I., Ali, M., 2017. Decomposing the linkages between energy consumption, air pollution, climate change, and natural resource depletion in Pakistan. Environ. Prog. Sustain. Energy 36 (2), 638–648. Zaman, K., Khan, M.M., Ahmad, M., Rustam, R., 2012. Determinants of electricity consumption function in Pakistan: Old wine in a new bottle. Energy Policy 50, 623–634. Zameer, H., Wang, Y., 2018. Energy production system optimization: Evidence from Pakistan. Renew. Sustain. Energy Rev. 82, 886–893.