Oil price shocks and the composition of current account balance

Varlik, Serdar; Berument, Hakan Article Oil price shocks and the composition of current account balance Central Bank Review (CBR) Provided in Cooperation with: Central Bank of The Republic of Turkey, Ankara Suggested Citation: Varlik, Serdar; Berument, Hakan (2020) : Oil price shocks and the composition of current account balance, Central Bank Review (CBR), ISSN 1303-0701, Elsevier, Amsterdam, Vol. 20, Iss. 1, pp. 1-8, https://doi.org/10.1016/j.cbrev.2020.02.002 This Version is available at: https://hdl.handle.net/10419/217342 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-nc-nd/4.0/ Oil price shocks and the composition of current account balance * Serdar Varlik a , M. Hakan Berument b , * a Department of Economics, Hitit University, 19040, Corum, Turkey b Department of Economics, Bilkent University, 06800, Ankara, Turkey article info Article history: Received 10 November 2019 Received in revised form 26 January 2020 Accepted 12 February 2020 Available online 13 March 2020 JEL classification: F14 F17 Q40 Keywords: Oil prices Current account balance FAVAR abstract It is a well-established regularity that permanent oil price shocks do not have a permanent effect on the current account deficit. This requires that sub-components of the current account or trade balance will make the necessary adjustments to accommodate the higher energy bill of a country triggered by permanent crude oil price increases. Empirical evidence gathered from Turkey reveals that, in the long run, balancing the current account is provided by a permanent increase in the net exports of Agricultural Production, Maintenance and Repair Services, Travel, Construction, Financial Services, Compensation of Employees, and Goods under Merchanting (non-tradable components of the current account balance); and a permanent decrease in the net exports of Mining, Fishery, Other Goods for BEC Classification, Investment Income, Manufacturing Services on Physical Inputs Owned by Others, and Transport balances mostly in sectors that use energy heavily in production. All these responses are found to be statistically significant in the more than 24 periods we consider in this study. ©2020 Central Bank of The Republic of Turkey. Production and hosting by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). 1. Introduction The balance of payments of energy-poor countries is vulnerable to sudden oil price shocks, which affects the overall economy through the current account balance (Kaminsky et al., 1998; and Kaminsky and Reinhart, 1999). For oil-importing countries, as the consumption expenditures cannot be reduced immediately after oil price shocks, the initial effect, in the short-term, of oil price increases is a deterioration in the current account balance. Following the initial effect, over time, as consumption expenditures decrease, the current account improves and turns to the pre-shock state or surplus (Agmon and Laffer, 1978). This movement of the current account against oil price shock is similar to the J-curve shape. The initial studies in the literature, which focus on the impact of change in terms of trade induced by oil price shock on the current account balance, reflect the intertemporal aspects. These studies, differentiating temporary and permanent oil price shocks, examine the rebalancing process of the current account balance following the oil price shock. Under the flexible wages and full employment assumptions, Sachs (1981,1982) and Obstfeld (1980,1982) consider that the permanent deterioration of terms of trade induced by the oil price shock triggers to increase savings in order to restore the wealth by accumulating foreign assets and, consequently, the current account balance of the oil-importing country improves. 1 However, Svensson (1984) notes that the response of the current account balance is ambiguous for permanent oil price increases under the assumption of rigid wages. Subsequent studies focusing on intertemporal analyses emphasize that the current account unambiguously improves against temporary deterioration of terms of trade induced by the increase in oil price (see; Svensson and Razin, 1983;Greenwood, 1984;Persson and Svensson, 1985;Bean, 1986;Edwards, 1987; Frenkel and Razin, 1987;Matsuyama, 1988;Ostry, 1988;Sen and Turnovsky, 1989;Turnovsky and Sen, 1991;Otto, 2003;Huang * We would like to thank the anonymous referees, the seminar participants at The Central Bank of the Republic of Turkey, as well as Faruk Aydın, Adnan Eken, Hakan Kara and Tara Sylvestre for their valuable suggestions. *Corresponding author. E-mail addresses: serdarv[email protected] (S. Varlik), [email protected] (M.H. Berument). URL:http://berument.bilkent.edu.tr/ Peer review under responsibility of the Central Bank of the Republic of Turkey. 1 These studies challenge the non-optimizing static model of Harberger (1950) and Laursen and Meltzer (1950) who claim that deterioration of terms of trade decreases real income and savings at given investment expenditure, fiscal policy and nominal income. This process ends with a deterioration of the current account balance. Contents lists available at ScienceDirect Central Bank Review journal homepage: http://www.journals.elsevier.com/central-bank-review/ https://doi.org/10.1016/j.cbrev.2020.02.002 1303-0701/©2020 Central Bank of The Republic of Turkey. Production and hosting by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http:// creativecommons.org/licenses/by-nc-nd/4.0/). Central Bank Review 20 (2020) 1e8 and Meng, 2007; and Cardi, 2007). Moreover, in these studies, as the permanent oil price shock decreases the real income and expenditures by similar amounts, it is concluded that the current account balance will not be affected in the long term. On the other hand, Marion (1984), without making any distinction between temporary and permanent deteriorations, suggests that the current account balance improves overtime after the oil price shock. However, under the assumption that non-tradable goods are added to the analysis, Marion states that the improvement of the current account balance is no longer valid. In this case, production technology in tradable and non-tradable goods sectors determine the movements of the current account balance. In another important study, Van Vijnbergen (1985) shows that the increase in oil prices decreases the current account deficit when the investment expenditures are not taken into consideration. He argues that the increase in oil prices may lead to the current account surplus, even higher prices cause a recession. Another set of studies directly examines the effect of oil price shocks. These studies elaborate on the short-term effects of oil price shocks on the current account balance. Baharumshah, Lau and Fountas (2003),Rebucci and Spatofora (2006),Aristovnik (2007), Gruber and Kamin (2007),Zaouali (2007),Bitzis et al. (2008), Schubert (2014),Kilian et al. (2009),Chuku et al. (2011),Le and Chang (2013),Narayan (2013),Baffes et al. (2015), and Huntington (2015) suggest that oil-importing countries have higher oil import bills due to the relative price inelasticity of oil demand, and so oil price shock increases the current account deficit in the short-term. However, these studies overlook the effect of oil price increases in the long-term. The purpose of this paper is to study the effect of an oil price shock on the current account balance, as well as on the different sub-components of the current account balance over time. We assess this relationship for Turkey. There are several reasons for using Turkish data: (i) Turkey is a small oil-importing open economy. Crude oil import is an important component of the current account and trade (in) balance. Therefore, an increase in crude oil price deteriorates the current account balance. (ii) Most of the sectors in the Turkish economy are oil-dependent. Therefore, oil price movements have a significant effect on economic performance. (iii) Since the demand elasticity of crude oil is low, especially for the industrial sector, an increase in crude oil prices rises the crude oil expenditures of Turkey. Thus, this increases Turkey’s oil bill. There is a set of studies that examines the short-term effects of oil price shock on the current account balance for Turkey. Aytemiz and S¸ eng€ onül (2008),Demirbas¸ et al. (2009),Peker and Hotuno glu (2009),€ Ozlale and Pekkurnaz (2010),Kayıkçı(2012),Bayat et al. (2013), and € Ozata (2014) find that oil price adversely affects the current account balance in the short-term. However, in these studies focusing on Turkey, after the oil price shock, the balancing process of the current account over time is disregarded. Unlike previous studies, the original contribution of this paper is that it is the first study to examine how the increase in oil prices affects the different sub-components of the current account balance over time. Thus, after a permanent increase in oil price, it can be determined which sub-components provide the current account balance. The findings on the effects of crude oil prices on the subcomponents of the current account balance are very important for policymakers because the estimation of when the effects will be observed, which sectors will be affected and how, if these effects will be permanent or not, and when the effect will reach its peak are important considerations when designing policies for those sectors. Overall, this paper is important since we analyze the effects of a crude oil price shock on the sub-components of the current account balance and employ a novel application to analyze the disaggregated data on the sub-components of Turkey’s current account balance. Our econometric method is designed to achieve this goal. The conventional Vector Autoregressive Models (VAR) usually have a limited number of variables in their specifications. However, the sub-components of a current account are numerous. Hence, in order to account for these variables, we will use the Factor Augmented Vector Autoregressive Model (FAVAR) employed by Bernanke et al. (2005). FAVAR includes large data sets that are reduced to a few factors without any significant loss of information and avoids the low degrees of freedom problem. Thus, FAVAR addresses the omitted information problem. On the other hand, Turkey is a small-oil importing open economy. Thus, Turkey is too small to affect the crude oil price in the world but is still affected by world crude oil prices. Therefore, we will use the Block Vector Autoregressive model employed by Cushman and Zha (1997). The mentioned two features will be indebted, for the first time, to an econometric specification that we call Block Exogeneity Factor Augmented VAR (BE-FAVAR). BE-FAVAR allows us to elaborate on the effects of positive innovations in an external variable (such as crude oil prices) on different domestic variables. The empirical evidence gathered from Turkey reveals that a positive permanent innovation in the crude oil real price permanently increases energy import, then the different sub-components of the current account balance should adjust. When a positive permanent shock is given to crude oil real price, the adjustment of the current account balance in the long-term is provided by a permanent increase in the mainly net exports of non-tradable subcomponents, and a permanently decrease in the net exports of the mostly tradable sub-components of the current account balance that heavily use energy as an important component of inputs. Against a crude oil real price shock, the net exports of Agricultural Production, Maintenance and Repair Services, Travel, Construction, Financial Services, Compensation of Employees, and Goods under Merchanting (trade of the imported products for export) permanently increase. However, the net exports of Mining, Fishery, Other Goods for BEC Classification, Investment Income, Manufacturing Services on Physical Inputs Owned by Others, and Transport balance on the current account permanently decrease. These results make sense since the price elasticities of the oil demand of the sectors related to the different sub-components of the current account will differ; thus the responses of these sub-components to the oil price shock differ. Therefore, the effect of a permanent increase in oil prices on the current account is expected to be balanced within some sub-components of the current account balance over time. This paper is organized as follows: In Section 2, we introduce the extension of the FAVAR methodology employed by Bernanke et al. (2005) with block exogeneity specification employed by Cushman and Zha (1997). We provide empirical evidence in Section 3.In Section 4, we conclude the paper. 2. Method Our econometric methodology is an extension of Bernanke et al.’s (2005) FAVAR modeling with the block exogeneity specification of Cushman and Zha’s(1997). FAVAR modeling, employed by Bernanke et al. (2005), without any significant loss of information and the degrees of freedom problem, provides the use of large data sets by reducing them to a few common factors that explain the majority of the data sets. Thus, FAVAR overcomes the omitted information problem which is often found in standard limitedvariable VAR models. The Block Exogeneity Structural Vector Autoregression model employed by Cushman and Zha (1997) is used here to capture for countries that are too small to affect world S. Varlik, M.H. Berument / Central Bank Review 20 (2020) 1e82 oil prices but are still affected by foreign variables. Thus, we call our method Block Exogeneity Factor-Augmented VAR (BE-FAVAR). Factor analyses can be used to capture a zero-mean, stationary time series as linear combinations of common components driven by a small number of factors, and idiosyncratic components. Let X t be N1 time series of 27 sub-components of the current account balance, real effective exchange rate, Euro/Dollar parity, industrial production, consumer price index, and producer price index in their stationary and zero mean form. Nindicates the ‘large’number of informational time series. Y t is a vector of M1 observable variables. In this study, Y t represents crude oil real price. F t is a k1 vector of unobservable factors that have most of the information contained in X t which cannot be estimated in the standard VAR approach. Nis much higher than the number of factors ðN>kþMÞ. Also, F ðLÞ¼I F * ðLÞ¼I F 1 L F 2 L 2 … F d L d ; F ðLÞindicates the appropriate lag of the mxmmatrix polynomial of finite lag order din the lag operator L. F i ðl¼1;…dÞis the coefficient matrix. y t is an mx1 vector of error term with zero mean. 2.1. The joint dynamics of F t and Y t can be written as 2 Ft Yt¼ F *ðLÞFt1 Yt1þ y t4 F ðLÞFt Yt¼ y t(1) Equation (1) is a standard VAR but uses a vector of unobservable factors F t and observable variable Y t .F t Y t is an mx1 vector of variables. However, Equation (1) is not reduced to a standard VAR in Y t . Since Equation (1) has unobservable factors, we cannot directly estimate this standard VAR equation. In this case, Bernanke et al. (2005) estimate Equation (1) by using FAVAR. In order to explain the dynamic factor model, assume that the informational time series X t can be represented as a function of unobservable F t and observable Y t , such that Xt¼LfFtþLyYtþet:(2) here, L f is a Nkand L y is an NMmatrix of the factor loadings, and e t is an N1 vector of error terms. In Equation (2),X t depends only on the current values and not the lagged values of the factors. The block exogeneity issue is captured with Cushman and Zha’s (1997) block exogeneity specification. Our specification is as follows: F ðLÞFt Yt¼ y t:(3) This specification can be presented in matrix form in Equation (4). F ðLÞ¼ F 11ðLÞ F 12ðLÞ 0 F 22ðLÞ; y t¼ y 1ðtÞ y 2ðtÞ:(4) The coefficient matrix of L 0 F 0 , is non-singular and y ðtÞis uncorrelated with past F t Y t s. F 21 ðLÞare all zero, which captures the block exogeneity in the F ðLÞmatrix, as suggested by Cushman and Zha (1997).Y t is our external block, which comprises crude oil real price. F t is our domestic block, which includes the main subcomponents of the current account balance. 3. Empirical evidence Our econometric application comprises two stages. For both of these stages, our data span covers a monthly observation from December 2001 to March 2018 for Turkey. The beginning of our sample period is dictated by data availability. Table 1 reports codes, variable names, and sources of domestic and external blocks’data sets. In the first stage, we estimate how crude oil real price shock affects the current account balance by using the Block Exogeneity VAR method employed by Cushman and Zha (1997). In the second stage, we examine the effects of crude oil real price shock to subcomponents of the current account balance by using the Block Exogeneity Factor-Augmented VAR (BE-FAVAR) method. 3.1. The effects of oil price shock to the current account balance In order to estimate the effects of crude oil real price shock to the current account balance, we use Block Exogeneity VAR. We determine that the lag order of the identified VAR model with block exogeneity is three by using the Schwarz Information Criteria. We also place constant term and 11 seasonal dummy variables to account for seasonality. Our external block comprises the first difference of logarithmic crude oil real price. We calculate the crude oil real price data as the price of Brent-Europe (Dollars per barrel) divided by the US Consumer Price Index for Urban Consumers (All Items). Our domestic block includes the dollar-denominated current account balance for the current month as divided by twelve lag of the dollardenominated interpolated monthly GDP, 3 the twelve-month difference of the logarithmic real effective exchange rate, the twelvemonth difference of the logarithmic industrial production, and the twelve-month difference of the logarithmic producer price index. Furthermore, we include dollar-denominated capital flows over twelve lag of the dollar-denominated interpolated GDP, and world economic growth for domestic block and world economic growth for the external block as the control variables. 4 In order to determine whether these series have a long-run constant mean, we performed a set of unit root tests. The test statistics suggest that all of these series are stationary, and thus we treat them all as stationary. Before reporting the FAVAR estimates, we estimate a model for the effect of crude real oil prices on the Turkish economic performance with a conventional Block Exogeneity VAR model. Here we did not include sub-components of the current account balance but rather the total current account balance. In order to identify the system, we use the Cholesky decomposition. The identification implies that the order of the variables is important. The variables are ordered as the twelve-month difference of the logarithmic producer price index, the twelve-month difference of the logarithmic industrial production, the dollar-denominated current account balance as divided by twelve lag of the dollar-denominated interpolated monthly GDP, the twelve-month difference of the logarithmic real effective exchange rate, and the first difference of the logarithmic crude real oil price. This ordering implies that the 2 For an excellent and easy-to-follow presentation of FAVAR methodology see Bernanke et al. (2005),Stock and Watson (1998, 2005), and Soares (2013). 3 The reason for deflating current account balance sub-components with lagged GDP is to eliminate the simultaneity. Otherwise we could be capturing the effect of real oil prices on GDP. 4 We include the world growth rate and capital flows as control variables rather than endogeneous variables into the system. The main reason for this is that the purpose of this paper is to assess the effects of oil prices shocks on the Tuırkish Current Account balance and its composition, rather than assessing the effects of these two varaibles. Modeling these two varaibles would increase number of parameters to be estiamted conisderably and would lead to less efficienmt estimates. S. Varlik, M.H. Berument / Central Bank Review 20 (2020) 1e83 last variable affects all the previous variables, but is affected by none contemporaneously. Similarly, the first variable is affected by all the following variables. However, all the variables except the crude real oil prices affect each other with lag. The crude real oil prices are not affected by the domestic variables. In order to identify the system, we use the Cholesky decomposition. The identification implies that the order of the variables in the domestic block is important. In the ordering, the preceding variables affect the latter variables, but not vice-versa contemporaneously. However, all the variables in the domestic block affect each other with a lag. Fig. 1 reports the impulse response function when a onestandard-deviation shock is given to the crude oil real price growth rate for the current account balance only. The solid black line is for the impulse responses, and the two dotted lines are for one-standard-deviation confidence intervals for 24 horizons. Since real oil prices are Ið1Þ, giving a one-standard-deviation shock to the growth rate of real oil prices means a permanent increase in real oil prices where oil price growth shock has an exponential decay on itself when we observe impulse responses. Fig. 1 suggests that a one-standard-deviation shock to the crude oil real price growth increases the current account deficit about 20 periods in a statistically significant fashion, but the effects die out in the long run. 5 Table 1 Data sources. Code Variable Name Sources Current Account Balance Detailed Presentation (BPM6) TP.ODAYR6.Q001 I-CURRENT ACCOUNT-Level EDDS TP.ODAYR6.Q007 I-A.1.General merchandise on a balance of payments basis-Level EDDS TP.ODAYR6.Q016 I-A.2.Net exports of goods under merchanting (credit)-Level EDDS TP.ODAYR6.Q017 I-A.3.Nonmonetary gold-Level EDDS TP.ODAYR6.Q023 IeB.1.Manufacturing services on physical inputs owned by others-Level EDDS TP.ODAYR6.Q026 IeB.2.Maintenance and repair services n.i.e.-Level EDDS TP.ODAYR6.Q029 IeB.3.Transport-Level EDDS TP.ODAYR6.Q038 IeB.4.Travel-Level EDDS TP.ODAYR6.Q041 IeB.5.Construction-Level EDDS TP.ODAYR6.Q044 IeB.6.Insurance and pension services-Level EDDS TP.ODAYR6.Q047 IeB.7.Financial services-Level EDDS TP.ODAYR6.Q050 IeB.8.Other business services-Level EDDS TP.ODAYR6.Q053 IeB.9.Government goods and services n.i.e.-Level EDDS TP.ODAYR6.Q056 IeB.10.Other services-Level EDDS TP.ODAYR6.Q062 IeC.1.Compensation of employees-Level EDDS TP.ODAYR6.Q065 IeC.2.Investment income-Level EDDS TP.ODAYR6.Q084 I-D.1.General Government-Level EDDS TP.ODAYR6.Q085 I-D.2.Other Sectors-Level EDDS Foreign Trade Broad Economic Categorization (BEC)(TURKSTAT) TP.DT.ARA.IH.B Intermediate Goods (Exports)-Level EDDS TP.DT.ARA.IT.B Intermediate Goods (Imports)-Level EDDS TP.DT.DIG.IH.B Other (Exports)-Level EDDS TP.DT.DIG.IT.B Other (Imports)-Level EDDS TP.DT.GEN.IH.B Total (Exports)-Level EDDS TP.DT.GEN.IT.B Total (Imports)-Level EDDS TP.DT.SER.IH.B Capital Goods (Exports)-Level EDDS TP.DT.SER.IT.B Capital Goods (Imports)-Level EDDS TP.DT.TUK.IH.B Consumption Goods (Exports)-Level EDDS TP.DT.TUK.IT.B Consumption Goods (Imports)-Level EDDS Foreign Trade International Standard Industry Categorization(ISIC REVIZE 3)(TURKSTAT) TP.DT.BAL.IH.I Fishing (Exports)-Level EDDS TP.DT.BAL.IT.I Fishing (Imports)-Level EDDS TP.DT.DIG.IH.I Others (Exports)-Level EDDS TP.DT.DIG.IT.I Others (Imports)-Level EDDS TP.DT.GEN.IH.I Total (Exports)-Level EDDS TP.DT.GEN.IT.I Total (Imports)-Level EDDS TP.DT.IMA.IH.I Manufacturing (Exports)-Level EDDS TP.DT.IMA.IT.I Manufacturing (Imports)-Level EDDS TP.DT.MAD.IH.I Mining and Quarrying (Exports)-Level EDDS TP.DT.MAD.IT.I Mining and Quarrying (Imports)-Level EDDS TP.DT.TAR.IH.I Agriculture and Forestry (Exports)-Level EDDS TP.DT.TAR.IT.I Agriculture and Forestry (Imports)-Level EDDS Other Macroeconomic Variable and Oil Price TP RK T1 Y Real Effective Exchange Rate EDDS EXUSEU Euro Dollar Parity FRED Data Mineral Fuels, Mineral Oils and Product of Their Distillation Turkish Statistical Institute TP ODAYR6 Q090 TP ODAYR6 Q092 Capital and Financial Account EDDS TP GSYIH26 HY CF Gross Domestic Product EDDS NYGNPMKTPCDWLD Growth of World Output FRED Data TP SANAYREV4 Y1 Industrial Production EDDS TP FG JO Consumer Price Index EDDS MCOILBRENTEU Crude Oil Prices: Brent-Europe, Dollars per barrel FRED Data 5 The results of the impulse responses under the Cholesky decomposition may be sensitive to the ordering. The VAR models with alternative ordering are estimated. The results are mostly are robust. These estimates are not reported here to save space but are available from the authors upon a request. S. Varlik, M.H. Berument / Central Bank Review 20 (2020) 1e84 3.2. The effects of oil price shock to sub-components of the current account balance Next, using the BE-FAVAR method, we estimate the effects of crude oil real price growth shock to sub-components of the current account balance for the same period. Our external block consists of the first difference of logarithmic crude oil real price, which is calculated as the price of Brent-Europe (Dollars per barrel) divided by the US Consumer Price Index for Urban Consumers (All Items). In the domestic block, we use the dollar-denominated sub-compo- nents of the current account balance for the current month as divided by twelve lag of the dollar-denominated interpolated monthly GDP, the twelve-month difference of the logarithmic real effective exchange rate, the twelve-month difference of the logarithmic Euro/Dollar parity, the twelve-month difference of the logarithmic industrial production, the twelve-month difference of the logarithmic producer price index, and the twelve-month difference of the logarithmic consumer price index. Also, we use the dollar-denominated capital flows over twelve lag of the dollardenominated GDP and the world economic growth as control variables for the domestic block. Our control variable for the external block is the world economic growth. All of these series are stationary, and thus we treat them all as stationary. In order to determine the number of factors for the domestic block variables, we use Bai and Ng’s (2002) Factor Determination Test. The test results are reported in Table 2. Two of the test statistics suggest the number of common factors to be four, and one test statistics suggests five. Five factors explain 58% of the variation for the domestic block that we consider. Thus, we took the number of factors to be five. Similar to the VAR specification in Section 3.1, the BE-FAVAR specification includes the constant term and 11 seasonal dummy variables to account for seasonality. Using the Schwarz Information Criteria, we determine that the lag order of the identified VAR model with block exogeneity is two. Fig. 2 reports the impulse response functions of the subcomponents of the current account balance when a onestandard-deviation shock is given to the crude oil real price growth rate. The solid black line is for the impulse responses, and the two dotted lines are for the one-standard-deviation confidence intervals for 24 horizons. 6 Fig. 2 suggests that a one-standard- deviation shock to the crude oil real price growth permanently increases the energy import that we consider. For the effects of crude real oil price shocks on the other sub-components of the current account deficit, the balancing of the current account, in the long run, is provided by a permanent increase (higher level of surplus or lower level of deficit) in Goods under Merchanting (that captures the trade of the imported products for export), Agricultural Production, Maintenance and Repair Services, Travel, Construction, Financial Services, and Compensation of Employees. All these responses are found to be statistically significant in the more than 24 periods that we consider. When a positive permanent shock is given to crude oil real price, Mining, Fishery, Other Goods for BEC Classification, Investment Income, Manufacturing Services on Physical Inputs Owned by Others, and Transport balance decrease in a statistically significant fashion in more than 24 periods. These are usually the sectors that energy is one of the most important inputs. Thus, our findings suggest that the long-term balancing process in the current account is mainly provided by services trade. On the other hand, Consumption Goods, Other Sectors, General Government, and Other Services increase statistically significantly against a positive permanent innovation in the crude oil real price, in the short run. However, when a positive permanent shock is given to crude oil real price, Manufacturing Industry, Capital Goods, Intermediate Goods and Government Goods and Services decrease in a statistically significant fashion in the short run. Overall, we may claim that the correction comes in service or non-tradable industries’current account surpluses in the longterm. However, for the sectors that use energy as input for a significant part of production processes, then permanent deteriorations are observed. There might be various reasons for this. For example, first, when oil prices rise, then domestic currency depreciates as a shock absorber, and the service or non-tradable sectors may benefit from the competitiveness gain through relative prices more than tradable sectors because of their higher Fig. 1. Responses of Current Account Balance to Oil Price Shock Table 2 Bai-Ng’s factor determination test and variance shares for net exports of main subcomponents of current account balance. # Factors PCP1 PCP2 PCP3 Cumulated Variance Share 1 0.7962 0.7909 0.7833 0.2081 2 0.7125 0.7019 0.6866 0.3620 3 0.6643 0.6483 0.6253 0.4595 4 0.6607 a 0.6394 a 0.6088 0.5298 5 0.6720 0.6454 0.6071 a 0.5805 6 0.6864 0.65451 0.6085 0.6279 7 0.7089 0.6717 0.6180 0.6725 8 0.7332 0.6906 0.6293 0.7141 9 0.7601 0.7122 0.6432 0.7524 10 0.7902 0.7369 0.6603 0.7867 Note. a is for the number of optimum factors. 6 Following Bernanke et al. (2005), we use the two-stage principal component method. Thus, we use the Kilian (1998) bootstrap methodology in order adjust the confidence interval with respect to error band for impulse response analysis due to uncertainty in factor estimates. S. Varlik, M.H. Berument / Central Bank Review 20 (2020) 1e85 Fig. 2. Responses of Net Exports of Main Sub-components of Current Account Balance over Twelve Monthly Lag of GDP to Oil Price Shock Once Capital Flows and World Economic Growth for Domestic Block and World Economic Growth for Oil Price are Controlled. S. Varlik, M.H. Berument / Central Bank Review 20 (2020) 1e86 domestic value-addition. Second, effects of higher oil price shocks can also be observed on the real wages of the service and nontradable sectors; since wages in the latter sectors are more flexible than the tradable sectors, then an increase in oil price may cause a higher decrease in real wages in tradeable sectors and make them more competitive. Last, higher oil prices mean income transfer from oil-importing countries to oil-exporting countries, higher demand from the oil-exporting countries for the service or non-tradable sectors in the oil-importing countries may stimulate these sectors permanently as well (see Guidi, 2009). As robustness tests, for the first alternative, different from the benchmark model, we re-performed the analyses under the same specification by using the first lag rather than the twelfth lag. In the second alternative, different from the benchmark model and the first alternative, we re-analyzed by changing our control variables. We used the dollar-denominated capital flows over the first lag of the dollar-denominated interpolated GDP as control variables for the domestic block, and the world economic growth as control variables for the external block. For the third alternative, we reperformed the analyses differently from the benchmark model in terms of control variables. We used the dollar-denominated capital flows over twelve lag of the dollar-denominated GDP as control variables for the domestic block, and world economic growth as control variables for the external block. All of these three robustness analyses reveal that our results are robust. 7 4. Conclusion Following the sudden oil price shocks, energy-importing- countries, which are vulnerable to increases in oil prices, cannot reduce the consumption expenditures immediately. Thus, increases in oil price deteriorate the current account balance of these countries. However, the current account balance improves just as consumption expenditures decrease over time. This response of the current account balance to oil price shocks being similar to the J- curve shape requires the adjustment of sub-components of the current account balance over time. Therefore, which subcomponents of current account balance will be affected by an oil price shock, whether these impacts will be permanent and when these impacts will peak, are important for energy-importing countries. The aim of this study is that the effects of a permanent increase in crude oil real prices on the current account balance and on different sub-components of the current account balance are examined for Turkey which is a small oil-importing open economy. Since we use a large data set and willing to capture block exogeneity assumption, our econometric method designing to examine the aim of this study requires to extend the Factor Augmented Vector Autoregressive Model (FAVAR) employed by Bernanke et al. (2005) with the Block Vector Autoregressive Model employed by Cushman and Zha (1997). We call this method Block Exogeneity Factor Augmented VAR (BE-FAVAR). Thus, BE-FAVAR, which is reduced the large data sets to a few common factors and captured the exogeneity assumption, allows us to examine the effects of positive innovations in an external variable (such as crude oil real prices) on different domestic variables (here on sub-components of current account balance) by overcoming the omitted information problem seen in standard limited-variable VAR models. The studies in the literature either focus on the impact of change in terms of trade induced by the oil price shock on the current account balance or focus on the short-term effects of oil price shock on the current account balance. However, these studies overlook the effect of oil price shocks in the long-term. Unlike the previous study in the literature, this study provides an original contribution to the literature in terms of examining how the increase in oil prices affects the different sub-components of the current account balance over time. Thus, this study provides further empirical evidence on the implication of the neutrality effect of real oil price shocks on the current account balance in the long-term. Empirical evidence gathered from Turkey reveals that, in the long run, balancing the current account is provided by a permanent increase in the net exports of Agricultural Production, Maintenance and Repair Services, Travel, Construction, Financial Services, Compensation of Employees, and Goods under Merchanting (nontradable components of the current account balance); and a permanent decrease in the net exports of Mining, Fishery, Other Goods for BEC Classification, Investment Income, Manufacturing Services on Physical Inputs Owned by Others, and Transport balances mostly in sectors that use energy heavily in production. Overall, we may claim that the correction comes in service or non-tradable industries with current account surpluses in the long-term. However, for the sectors that use energy as input for a significant part of production processes, then a permanent deterioration is observed. These information about the effects of crude oil prices on the sub-components of current account balance might be useful for policymakers as they design their economic policies across sectors; it could help speed up the adjustment in order to minimize the social loss, as well as prioritize the sub-sectors, which would support them in an adverse oil shock. Moreover, the government collects sizable tax revenues from oil products. Thus, a lower net export may mean higher tax revenue due to higher domestic consumption associated with a higher import bill. Thus, the government may adjust its tax policies on different oil products for the sub-components of different sub-sectors to stimulate the output of those sectors. References Agmon, T., Laffer, A.B., 1978. Trade, payments and adjustment: the case of the oil price rise. Kyklos 31, 68e85. Aristovnik, A., 2007. Short- and Medium-Term Determinants of Current Account Balances in Middle East and North Africa Countries. William Davidson Institute Working Paper, No, p. 862. Aytemiz, T., S¸ eng€ onül, A., 2008. Regression tree analysis of effects of energy prices on Turkish current account deficit. _ Iktisat, _ Is¸ letme ve Finans 23 (269), 94e109. Baffes, J., Kose, M.A., Ohnsorge, F., Stocker, M., 2015. The Great Plunge in Oil Prices: Causes, Consequences, and Policy Responses. https://doi.org/10.2139/ ssrn.2624398. Baharumshah, A.Z., Lau, E., Fountas, S., 2003. On the sustainability of current account deficits: evidence from four ASEAN countries. J. Asian Econ. 14, 465e487. Bai, J., Ng, S., 2002. Determining the number of factors in approximate factor models. Econometrica 70 (1), 191e221. Bayat, T., S¸ ahbaz, A., Akçacı, T., 2013. Petrol fiyatlarınındıs¸ ticaret açı  gıüzerindeki etkisi: Türkiye € orne gi. Erciyes Üniversitesi _ Iktisadi ve _ Idari Bilimler Fakültesi Dergisi 42, 67e90. Bean, C.R., 1986. The terms of trade, labor supply and the current account. Econ. J. 96, 38e46. Bernanke, B.S., Boivin, J., Eliasz, P., 2005. Measuring the effects of monetary policy: a factor-augmented vector autoregressive (FAVAR) approach. Q. J. Econ. 120 (1), 387e422. Bitzis, G., Paleologos, J.M., Papazoglou, C., 2008. The determinants of the Greek current account deficit: the EMU experience. Journal of International and Global Economic Studies 1 (1), 105e122. Cardi, O., 2007. Another view of the j-curve. Macroecon. Dyn. 11, 153e174. Chuku, C.A., Akpan, U.F., Sam, N.R., Effiong, E.L., 2011. Oil price shocks and the dynamics of current account balances in Nigeria. OPEC Energy Review 35 (2), 119e139. Cushman, D.O., Zha, T., 1997. Identifying monetary policy in a small open economy under flexible exchange rates. J. Monetary Econ. 39 (3), 433e448. Demirbas¸ , M., Türkay, H., Türko glu, M., 2009. Petrol fiyatlarındaki gelis¸ melerin Türkiyenin cari açı  gıüzerine etkisinin analizi. Süleyman Demirel Üniversitesi _ Iktisadi ve _ Idari Bilimler Fakültesi Dergisi 14 (3), 289e299. Edwards, S.,1987. Tariffs, Terms of Trade, and the Real Exchange Rate in an Intertemporal Optimizing Model of the Current Account. NBER Working Paper, No. 2214. 7 The results of these analyses are not presented here to save space but are available from the authors upon request. S. Varlik, M.H. Berument / Central Bank Review 20 (2020) 1e87 Frenkel, J.A., Assaf, R., 1987. Fiscal Policies and the World Economy: an Intertemporal Approach. MIT Press, Cambridge, Massachusetts, 1987. Greenwood, J., 1984. Non-traded goods, the trade balance, and the balance of payments. Can. J. Econ. 1984, 806e823. Gruber, J.W., Kamin, S.B., 2007. Explaining the global pattern of current account imbalances. J. Int. Money Finance 26, 500e522. Guidi, F., 2009. The Economic Effects of Oil Prices Shockson the UK Manufacturing and Servicessector Munich Personal RePEc Archive. Paper No. 16171. Harberger, A.C., 1950. Currency depreciation, income, and the balance of trade. J. Polit. Econ. 58, 47e60. Huang, K.X.D., Meng, Q., 2007. The Harberger-Laursen-Metzler effect under capital market imperfections. J. Int. Money Finance 26, 1001e1015. Huntington, H.G., 2015. Crude oil trade and current account deficits. Energy Econ. 50, 70e79. Kaminsky, G., Lizonda, S., Reinhart, C.M., 1998. Leading indicators of currency crises. Int. Monetary Fund Staff Pap. 45 (1), 1e48. Kaminsky, G., Reinhart, C.M., 1999. The twin crises: the causes of banking and balance-of-payments problems. Am. Econ. Rev. 89 (3), 473e500. Kayıkçı, F., 2012. Determinants of current account balance in Turkey: vector autoregression (VAR) approach. Afr. J. Bus. Manag. 6 (17), 5725e5736. Kilian, L., 1998. Small-sample confidence intervals for impulse response functions. Rev. Econ. Stat. 80 (2), 218e230. Kilian, L., Rebucci, A., Spatafora, N., 2009. Oil shocks and external balances. J. Int. Econ. 77 (2), 181e194. Laursen, S., Metzler, L.A., 1950. Flexible exchange rates and the theory of employment. Rev. Econ. Stat. 32, 281e299. Le, T.H., Chang, Y., 2013. Oil price shocks and trade imbalances. Energy Econ. 36, 78e96. Marion, N.P., 1984. Nontraded goods, oil price increases and the current account. J. Int. Econ. 16, 29e44. Matsuyama, K., 1988. Terms of trade, factor intensities and the current account in a life-cycle model. Rev. Econ. Stud. 55 (2), 247e262. Narayan, S., 2013. A structural VAR model of the Fiji islands. Econ. Modell. 31, 238e244. Obstfeld, M., 1980. Intermediate imports, the term of trade, and the dynamics of exchange rate, and current account. J. Int. Econ. 10, 461e480. Obstfeld, M., 1982. Aggregate spending and the terms of trade: is there a Laursen- Metzler effect? Q. J. Econ. 97, 251e270. Ostry, J.D., 1988. The balance of trade, terms of trade, and real exchange rate: an intertemporal optimizing framework. Int. Monetary Fund Staff Pap. 35 (4), 541e573. Otto, G., 2003. Terms of trade shocks and the balance of trade: there is a Harberger- Laursen- Metzler effect. J. Int. Money Finance 22, 155e184. € Ozata, E., 2014. Sustainability of current account deficit with high oil prices: evidence from Turkey. International Journal of Economic Sciences 3 (2), 71e88. € Ozlale, Ü., Pekkurnaz, D., 2010. Oil prices and current account: a structural analysis for the Turkish economy. Energy Pol. 38 (8), 4489e4496. Peker, O., Hotunluo glu, H., 2009. Türkiye’de cari açı  gın nedenlerinin ekonometrik analizi. Atatürk Üniversitesi _ Iktisadi ve _ Idari Bilimler Dergisi 23 (3). Persson, T., Svensson, L.E.O., 1985. Current account dynamics and the terms of trade: Harberger-laursen-metzler two generations later. J. Polit. Econ. 93, 43e65. Rebucci, A., Spatafora, N., 2006. Oil prices and global imbalances. IMF World Economic Outlook 4, 71e96, 2006. Sachs, J., 1981. The current account and macroeconomic adjustment in the 1970s. Brookings Pap. Econ. Activ. 1, 201e282. Sachs, J.D., 1982. In: Bhandari, J., Putnam, B. (Eds.), Energy and Growth under Flexible Exchange Rates: A Simulation Study, the International Transmission of Economic Disturbances under Flexible Ex-Change Rates. M.I.T. Press, Cambridge, MA. Schubert, S.F., 2014. Dynamic effects of oil price shocks and their impact on the current account. Macroecon. Dyn. 18 (2), 316e337. Sen, P., Turnovsky, S.J., 1989. Deterioration of the terms of trade and capital accumulation: are-examination of the Laursen-Metzler effect. J. Int. Econ. 26, 227e250. Soares, R., 2013. Assessing monetary policy in the Euro area: a factor-augmented VAR approach. Appl. Econ. 45 (19), 2724e2744. Stock, J.H., Watson, M.W., 2005. Implications of dynamic factor models for VAR analysis. NBER Work. Pap. 11467, 1e65. Svensson, L.E.O., Razin, A., 1983. The terms of trade and the current account: the Harberger-Laursen-Metzler effect. J. Polit. Econ. 91, 97e125. Svensson, L.E.O., 1984. Oil prices, welfare, and the trade balance. Q. J. Econ. 99 (4), 649e672. Turnovsky, S.J., Sen, P., 1991. Fiscal policy, capital accumulation, and debt in an open economy. Oxf. Econ. Pap. 43, 1e24. Van Wijnbergen, S., 1985. Oil price shocks, unemployment, investment and the current account: an intertemporal disequilibrium analysis. Rev. Econ. Stud. 52 (4), 627e 645. Zaouali, S., 2007. Impact of higher oil prices on the Chinese economy. OPEC Energy Review 31 (3), 191e214. S. Varlik, M.H. Berument / Central Bank Review 20 (2020) 1e88