A comparative analysis of the dynamic relationship between oil prices and exchange rates

Int. Fin. Markets, Inst. and Money 32 (2014) 397–414

Contents lists available at ScienceDirect

Journal of International Financial Markets, Institutions & Money j o ur na l ho me pa ge : w w w . e l s e v i e r . c o m / l o c a t e / i n t f i n

A comparative analysis of the dynamic relationship between oil prices and exchange rates夽,夽夽 M. Ibrahim Turhan, Ahmet Sensoy ∗, Erk Hacihasanoglu Borsa Istanbul, Resitpasa mah., Tuncay Artun cad., Emirgan, Istanbul 34467, Turkey

a r t i c l e

i n f o

Article history: Received 20 May 2014 Accepted 13 July 2014 Available online 19 July 2014 JEL classification: C58 C61 E44 F31 G01 Keywords: Exchange rate Crude oil G20 Dynamic conditional correlation Penalized contrast function

a b s t r a c t This paper applies cDCC model to compare the dynamic correlations between oil prices and exchange rates of G20 members. The significant shifts in the correlations are then endogenously detected. For each pair of oil price-exchange rate, empirical evidence confirms of a strengthening negative correlation in the last decade. Methodology suggests only two events; US’ invasion of Iraq in 2003 and the 2008 global financial crisis, associating shifts of correlations to stronger negative level. While the first event has a shifting effect on mainly developed members, the latter affects them all. The new relationship provides benefits in risk diversification and inflation targeting. © 2014 Elsevier B.V. All rights reserved.

夽 The views expressed in this work are those of the authors and do not necessarily reflect those of the Borsa Istanbul or their members. 夽夽 An earlier version of this paper was presented at the “8th Conference on Risk, Banking and Financial Stability” co-organized by the Journal of Financial Stability, the Central Bank of Indonesia, the Central Bank of Finland, the Central Bank of Brazil, the Central Bank of Turkey and Fordham University, in Indonesia, September 24–27, 2013. We thank to conference participants for helpful discussions and insightful suggestions. ∗ Corresponding author. Tel.: +90 212 298 27 39; fax: +90 212 298 25 00.

E-mail addresses: [email protected], [email protected] (A. Sensoy). http://dx.doi.org/10.1016/j.intfin.2014.07.003 1042-4431/© 2014 Elsevier B.V. All rights reserved.

398

M.I. Turhan et al. / Int. Fin. Markets, Inst. and Money 32 (2014) 397–414

1. Introduction Oil is one of the most important forms of energy. It is a natural and non-renewable resource that goes into making virtually everything including steel, aluminum, plastics, rubber, fabrics and fertilizers. Moreover, it is commonly regarded as a comparative advantage and a key strategic resource. As widely accepted, oil is a significant determinant of global economic performance and its price dynamics can affect the world economy in many different ways: An increase in the oil price will raise the cost of production and services, so will lead to an increase in price levels. Concerns about likely increases in price levels in the near future will produce uncertainty and negative sentiment in the financial markets, and the expected inflation will lower the equity values. In addition, the price of oil can set economic trends by dominating growth in the gross domestic product (GDP). An oil price increase will also have an effect on a country’s wealth by a transfer of income from oil importing to oil exporting countries through a shift in the terms of trade thus, inevitably, exchange rates are also expected to change (Turhan et al., 2013). These facts make it essential to understand the crude oil price dynamics and its impacts on the world economy, however it is a challenging question as the crude oil prices experienced very large fluctuations over the last three decades and recently became more volatile than ever. Fig. 1 shows the history of the oil price from Jan 2000 to April 2013. Oil prices have increased very sharply after the US’ invasion of Iraq, rising from about $30 per barrel at the beginning of 2003 to their highest level of $147 in July 2008 but by the end of December 2008, had fallen to $40 due to the global financial crisis. However, the prices again reached above $110 in 2012. The recently observed frequent and uncertain changes in oil prices are transmitted to the real economy and financial markets primarily through exchange rates vis-à-vis United States (US) dollar. In case of a strong US dollar, the major invoicing and settlement currency in international oil markets, it is probable that crude oil importing countries (except US) will be adversely affected (and vice versa), or an increase in the volatility of the US dollar exchange rate will create an environment of uncertainty for the crude oil exporting countries in terms of international purchasing power (Zhang et al., 2008). Financial globalization presents new challenges in understanding the effects of the oil prices to local economies and financial markets: Prior to financial globalization (first episode), oil price was affecting the economies mainly through the mechanisms related to fundamental or “more tangible” real factors such as the standard supply side effect and the current account balances. Even when the financial sector was concerned, the impact was limited to inflationary consequences. However since the late 1990s (second episode) with financial markets started to function in an integrated pattern at global scale, that is cross-border capital flows liberalized and became a major determinant on local economies, the relationship between the oil prices and the macroeconomic performance has changed in several ways.

Crude Oil 150

Price (USD)

120

90

60

30

Jan.00

Jul.02

Dec.04

Jun.07

Nov.09

May.12

Fig. 1. Crude oil price per barrel between Jan 2000 and April 2013.

M.I. Turhan et al. / Int. Fin. Markets, Inst. and Money 32 (2014) 397–414

399

First change is in terms of global income redistributions: as oil prices increase, oil producers get richer. However, unlike the first episode, this does not cause a decline in consumption in oil consumer economies thank to re-channeling the capital by the financial investments. Second change is related to expectations as they have become to be the most important factor for financial sector. While during the first episode, a real increase in the oil prices had a negative impact at overall, the financial globalization has just reversed or at least partially offset this mechanism. Now, better are the expectations concerning economic activity, higher is the demand for oil thus higher the oil prices would be. Moreover, the financialization of the crude oil with the introduction of innovative financial instruments and the practice of financial engineering increased the weight of futures markets, enabling expectation to play a much more determining role. Besides, this situation also brought out investment and speculation opportunities rising from the joint movement between the oil and the US dollar. Therefore, analysis of the interaction between oil prices and the exchange rates is not easy but crucial, not only for investment and risk management issues but also for the economic and financial stability. Majority of the previous research have examined this interaction using the methods of cointegration theory, causality tests, vector autoregressive models (VAR) and the vector error correction models (VECM). However, (surprisingly) less attention has been paid to oil price and exchange rate co-movements (Reboredo, 2012; Aloui et al., 2013). Our research aims to fill this gap. In that manner, we estimate a consistent dynamic conditional correlation (cDCC) model (Aielli, 2013) to observe the degree of co-movements between crude oil price (in US dollar) and exchange rates (US dollar/local currency) of G20 members in the last thirteen years.1 Then we endogenously detect the significant shifts in these dynamic correlations by a relatively new penalized contrast methodology (Lavielle, 2005).2 Finally, we try to answer the following two questions: (i) How and why did the correlation levels significantly change on the given dates? (ii) What happens next? 2. Literature review Since the seminal work of Hamilton (1983), the linkages between oil prices and macroeconomic fundamentals have been widely investigated in literature, in particular the information transmission between oil prices and exchange rates has been a topic of great interest. Theoretically, the link between the two is well established by the early studies of Golub (1983) and Krugman (1983). According to authors, an oil exporting (oil importing) country may experience exchange rate appreciation (depreciation) when oil prices rise, and depreciation (appreciation) when oil prices fall. Similarly, Blomberg and Harris (1995) explain the potential impact of exchange rates on oil prices by the law of one price for tradable goods: since oil is homogeneous, internationally traded and USD priced commodity, a depreciation in USD increases the purchasing power of foreigners that also increases the oil demand, in turn, pushing up the crude oil price in USD.3 Amano and Van Norden (1998) empirically study the relationship between US dollar real effective exchange rate and oil price by co-integration analysis and error correction model using a monthly data set from 1972 to 1995, and conclude that oil prices cause persistent US dollar real exchange rate shocks. Benassy-Quere et al. (2007) find similar causality from oil to real effective exchange rate of

1 G20 represents approximately 2/3 of the world population and has considerable weight in the global economy as producing about 90% of the global economic output and having a share of around 80% of world trade. Its importance to the advancement of international relations and the world economy gets significant each day as it is becoming the world’s macroeconomic coordination mechanism. Moreover, the 2008 global financial crisis showed that the current international system, led by the advanced economies, has limitations and is inadequate in reflecting the demands of emerging economies. In that case, G20 would have a balancing role on the over-centralized world economic power structure (Canrong and Shiqiang, 2010). 2 Determining breakdowns in co-movements is highly controversial. Many previous studies have used exogenously identified breaks, however such a choice is usually subject to criticism. Since we endogenously detect the shift points, periods of relatively high and low correlations are defined regardless of whether a financial crisis is the true cause. Furthermore, one of the advantages of using a penalized contrast methodology is that the variables are not necessarily normally distributed or independent. For a recent application, see Sensoy (2013). 3 See also Rogoff (1991).

400

M.I. Turhan et al. / Int. Fin. Markets, Inst. and Money 32 (2014) 397–414

the dollar between 1974-2004 and show that a 10% increase in the oil price coincides with a 4.3% appreciation of the dollar in the long run. Raurava (2004) studies the role of oil prices and the real exchange rate in Russia’s economy using VAR and co-integration techniques on the quarterly data from the first quarter of 1995 to the last quarter of 2002. An indirect conclusion of his study is that an increase in the oil price is associated with the real depreciation of the ruble in the long run. Chen and Chen (2007) use panel co-integration to test the relationship between real oil price and real exchange rate of G7 countries with a monthly data from 1972 to 2005. Their study suggests the existence of a significant long-term equilibrium relationship where a rise in oil prices depreciates the domestic currency against the dollar. Similar conclusions also come from Ghosh (2011) where author examines the relationship between crude oil price and exchange rate for India from July 2007 to November 2008 using the GARCH and EGARCH models, and reveals that an increase in the oil price return leads to the depreciation of Indian currency against US dollar. Cavalcanti and Jalles (2013) support these findings for Brazil by using a dataset from 1975 to 2008 and showing that negative oil shocks (where the price decreases) appreciate the local currency of Brazil, therefore reduce competitiveness by means or worsening terms of trade. Conducting a four-dimensional structural VAR model, Huang and Guo (2007) show that real oil price shocks imply a minor appreciation in the real exchange rate for China. Using the general method of moments, Yousefi and Wirjanto (2004) examine the impact of fluctuations in the US dollar value on the formation of OPEC oil prices. The study reports evidence of a negative correlation between fluctuations in the US dollar rate and the formation of OPEC oil prices. Narayan et al. (2008) use GARCH and EGARCH models to investigate the relationship between oil prices and the exchange rate between the Fijian dollar and the US dollar using daily data for the period 2000 to 2006. Their main result is that a rise in oil prices leads to an appreciation of the Fijian dollar. Using a structural VAR model on a quarterly data from 1990 to 2007, Akram (2009) presents evidence of a weaker US dollar leading to higher oil prices and that US dollar shocks significantly account for oil price fluctuations. Askari and Krichene (2010) examine the empirical relationship between monetary policy and oil prices using quarterly data for the sample period 1970–2009. Authors focus on the relationship between oil prices and the US dollar’s nominal effective exchange rate. Accordingly, a depreciation of the US dollar would lead to higher oil prices. Ji and Fan (2012), using a dataset from 2006 to 2010, argue that an increase in oil price is associated with the depreciation of the US dollar, however this relationship weakens after the global financial crisis due to loose monetary policy of the US government. Using daily data from 2000 to 2005, Zhang et al. (2008) report evidence regarding the long-term influence and a short-term limited effect of the US dollar exchange rate on oil prices, with an insignificant risk spillover effect from the currency market to the oil market. Concerning volatility, Cifarelli and Paladino (2010) conduct a trivariate GARCH-M model including oil prices, stock prices, and dollar exchange rates. Their findings provide strong evidence for the negative relation between oil price shifts and exchange rate changes. Buetzer et al. (2012) identify various shocks to real oil prices in a structural VAR and find no evidence leading to a systematic appreciation of exchange rates of oil exporters against those of oil importers. Basher et al. (2012) also examine the relationship between oil prices, exchange rates and stock prices in emerging markets, using structural VAR models for the period 1988–2008. The authors offer limited support for the relationship between oil prices and exchange rates.4 A recent study was provided by Turhan et al. (2013). Applying a VAR framework and generalized impulse response analysis on daily exchange rates of a large collection of emerging countries, authors show that a rise in oil prices leads to significant appreciation of domestic currencies against the US dollar between 2003 and 2010. On the other hand, as suggested by the theoretical models, the level of oil dependence of a country may be the reason why the rise in oil prices is associated with depreciation or appreciation of the US dollar. For example, Lizardo and Mollick (2010) show that an increase in the real price of oil leads to a significant depreciation of the US dollar relative to the currency of oil exporting countries such as

4 Besides, authors emphasize that the behavior of stock markets may be treated as a leading economic indicator since it signals expectations of higher economic growth. This is especially the case after the financial crisis, as emerging economies are leading the growth pattern of the global economy.

M.I. Turhan et al. / Int. Fin. Markets, Inst. and Money 32 (2014) 397–414

401

Canada, Mexico and Russia, however, the currency of oil importing countries, such as Japan, depreciated relative to the US dollar in the same scenario. They also find that currencies of countries that are neither oil exporters nor importers (UK and the European Union) appreciate relative to the US dollar when oil prices rise. Similarly, Krichene (2005, 2006) conducts a vector error correction model analysis and concludes that an appreciation of the nominal effective dollar exchange rate may lead to both an increase and a decrease in oil prices.5 Finally, in one of the very few studies on oil price and exchange rate co-movement, Reboredo (2012) finds that although the dependence between the two rose substantially in the aftermath of the global financial crisis, it is, in general, weak and there is no extreme market dependence between oil prices and exchange rates. In a recent study, Aloui et al. (2013) find evidence of significant and symmetric dependence for almost all oil-exchange rate pairs considered over the period 2001–2011. Moreover, the rise in the price of oil is found to be associated with the depreciation of the US dollar.6 From the recent literature it is apparent that the oil price-exchange rate relationship is crucial for several reasons, yet there is no clear-cut conclusion. Although empirical studies generally state that exchange rates and oil prices are co-integrated (where causality may run in both directions); they are divided into two groups concluding that an increase in the oil price is associated with either a US dollar appreciation (Dibooglu, 1996; Amano and Van Norden, 1998; Raurava, 2004; Benassy-Quere et al., 2007; Chen and Chen, 2007; Coudert et al., 2008; Ghosh, 2011; Basher et al., 2012; Cavalcanti and Jalles, 2013) or depreciation (Narayan et al., 2008; Zhang et al., 2008; Akram, 2009; Askari and Krichene, 2010; Wu et al., 2012; Ji and Fan, 2012; Aloui et al., 2013; Turhan et al., 2013) against other currencies.7 Furthermore, although alternative energy sources are fiercely promoted, the importance of oil prices and exchange rates in explaining the changes in each other has not declined and is not expected to do so in the near future. 3. Methodology 3.1. Consistent dynamic conditional correlation The dynamic correlations between oil prices and the exchange rates will be obtained by the cDCC model of Aielli (2013). First, we start by reviewing the DCC modeling (Engle, 2002) approach. Denote by yt = [y1,t , . . ., yM,t ] the M × 1 vector of the asset returns at time t, and assume that Et−1 [yt ] = 0 and Et−1 [yt y t ] = Ht , where Et [·] is the conditional expectation on yt , yt−1 , . . .. The asset conditional covariance matrix can be written as 1/2

Ht = Dt

1/2

Rt Dt

(1)

where Rt = [ij,t ] is the asset conditional correlation matrix and the diagonal matrix of the asset conditional variances is given by Dt = diag(h1,t , . . ., hM,t ). By construction, Rt is the conditional covariance matrix of the asset standardized returns, that is Et−1 [εt εt  ] = Rt , where εt = [ε1,t , . . ., εM,t ], and  εi,t = yi,t / hi,t . Engle (2002) models the right hand side of Eq.(1) rather than Ht directly Rt = {Qt∗ }−1/2 Qt {Qt∗ }−1/2 , Qt = (1 − ˛ − ˇ)S + ˛εt−1 εt−1 + ˇQt−1 ,

5

(2)

See also Rafiq (2011) for a discussion on what exchange rate policy should a country follow if its economy is dependent on

oil. 6 The literature on this bilateral relationship extends even further. For other recent studies, see Sari et al. (2010), Wang and Wu (2012), Benhmad (2012), Jain and Ghosh (2013), Reboredo and Rivera-Castro (2013), Reboredo et al. (2014), Ahmad and Hernandez (2013), Beckmann and Czudaj (2013a,b). 7 Where the majority of the works that associate an increase in the oil price with depreciation of the US dollar use various data sets anterior to the global financial crisis.

402

M.I. Turhan et al. / Int. Fin. Markets, Inst. and Money 32 (2014) 397–414

where Qt ≡ [qij,t ], S ≡ [sij ], Qt∗ = diag{Qt } and ˛, ˇ are scalars. The resulting model is called DCC. The cDCC model assumes that the correlation driving process is ∗1/2

∗1/2

Qt = (1 − ˛ − ˇ)S + ˛{Qt−1 εt−1 εt−1 Qt−1 } + ˇQt−1

(3)

Explicitly, in the bivariate case the correlation is defined as ij,t =



ωij,t−1 + ˛εi,t−1 εj,t−1 + ˇij,t−1

(4)

{ωii,t−1 + ˛ε2i,t−1 + ˇii,t−1 }{ωjj,t−1 + ˛ε2j,t−1 + ˇjj,t−1 }



where ωij,t ≡ (1 − ˛ − ˇ)sij /

qii,t qjj,t . Since Et−1 [εi,t εj,t ] = ij,t , the formula for ij,t combines a sort of

GARCH devices for the relevant past values and innovations into a correlation-like ratio. The parameters ˛ and ˇ are the dynamic parameters of the correlation GARCH devices. The time-varying intercept ωij,t can be seen as an ad hoc correction required for purposes of tractability (Aielli, 2013). 3.2. Detection of the shifts in dynamic correlations We will use the method of Lavielle (2005) to detect mean shifts in the dynamic correlations:8 Consider a sequence of random variables Y1 , . . ., Yn that take values in Rp . Assume that  ∈  is a parameter denoting the characteristics of the Yi ’s that changes abruptly and remains constant between two changes. The change occur at some instants 1 < 2 < · · · < K  −1 . Here K − 1 is the number of change points hence we have K number of segments where  is used to denote the true value. Now, let K be some integer and  = ( 1 ,  2 , . . .,  K−1 ) be a sequence of integers satisfying 0 <  1 <  2 < · · · <  K−1 < n. For any 1 ≤ k ≤ K, let U(Yk−1 +1 , . . ., Yk ; ) be a contrast function useful for estimating the unknown ˆ  +1 , . . ., Y ), true value of the parameter in the segment k; i.e. the minimum contrast estimate (Y k−1

computed on segment k of , is defined as a solution of the following minimization problem: ˆ  +1 , . . ., Y )) ≤ U(Y +1 , . . ., Y ; ), U(Yk−1 +1 , . . ., Yk ; (Y k k k−1 k−1

∀ ∈ ,

k

(5)

For any 1 ≤ k ≤ K, let G be ˆ  +1 , . . ., Y )) G(Yk−1 +1 , . . ., Yk ) = U(Yk−1 +1 , . . ., Yk ; (Y k k−1

(6)

then define the contrast function J(, y) as 1 G(Yk−1 +1 , . . ., Yk ) n K

J(, y) =

(7)

k=1

where  0 = 0 and  k = n. When true number K segments is known, for any 1 ≤ k ≤ K , the sequence ˆ n of change point instants that minimizes this kind of contrast has the property that Pr(|ˆn,k − k | > ı) → 0, when ı → ∞ and n → ∞

(8)

In particular, this result holds for weak or strong dependent processes. We consider the model Yi = i + i εi , 1 ≤ i ≤ n where (εi ) is a sequence zero-mean random variables with unit variance. We assume that ( i ) is a piecewise constant sequence and ( i ) is a constant sequence. Now, even if (εi ) is not normally distributed, a Gaussian log-likelihood can be used to define the contrast function. Let U(Yk−1 +1 , . . ., Yk ; ) =

k 

(Yi − )2

(9)

i=k−1 +1

8 To prevent misunderstandings, the reader is asked to consider the mathematical notations in this sub-section independent from the other parts of the manuscript.

M.I. Turhan et al. / Int. Fin. Markets, Inst. and Money 32 (2014) 397–414

403

then G(Yk−1 +1 , . . ., Yk ) =

k 

2 (Yi − Y¯ k−1 +1:k )

(10)

i=k−1 +1

where Y¯ k−1 +1:k is the empirical mean of (Yk−1 +1 , . . ., Yk ). When the number of shift points is unknown, it is estimated by minimizing a penalized version of J(, y). For any sequence of change point instants , let pen() be a function of  that increases with the number K() of segments of . Then, let ˆ n be the sequence of change point instants that minimizes F() = J(, y) + ϕpen()

(11)

where ϕ is a function of n that goes to zero at an appropriate rate as n goes to infinity. The estimated number of segments K(ˆ n ) converges in probability to K . The proper pen() and the penalization parameter ϕ are chosen according to Lavielle (2005).9 4. Data and results We consider the daily price of the world’s most prominent benchmark crude oil, Americas Dated Brent – EUCRBRDT (in US dollar), and exchange rates (US dollar/local currency) of G20 members from 02/01/2000 to 17/04/2013.10 Our research excludes Argentina, China and Saudi Arabia due to their controlled currency regime. Among the rest, Russia, India and Indonesia have managed float currencies and the currencies of the remaining float freely. Before applying the methodologies, raw returns are filtered with an ARMA process to obtain the yi,t series with zero means where the optimal lag selections are based on Bayesian Information Criterion. In the cDCC estimation, we use a GJR-GARCH(1,1) process for an additional weight to negative returns.11 Fig. 2 presents a visual representation of the shifts in the dynamic correlation levels and the exact shift dates are given in Table 1.12,13 Two specific years, 2003 and 2008, come forward in terms of correlation shifts. In particular, first shift starts with Canadian dollar (CAD) on 21/03/2003. Not so later, shifts for the Russian ruble (RBL), South African rand (ZAR), US dollar (USD), euro (EUR), British pound (GBP) and Australian dollar (AUD) follow in late 2003 and early 2004. The date 21/03/2003 is not surprising as it is one day after the US’ invasion of Iraq that hit the global oil market hard. The invasion created political instability within various oil producing nations. Moreover, the time line coincided with an increased global demand for oil. Since Iraq contains large amount of global oil reserves, a reduce in Iraq’s oil production (due to the invasion) increased oil prices significantly. However, it should be noted that after this event, shifts in the oil price-exchange rate correlations mainly occur for the exchange rates of developed economies.

9

For further details, refer to Lavielle (2005) and Lavielle and Teyssiere (2007). G20 includes United States (US), United Kingdom (UK), Japan, Canada, Russia, Australia, Brazil, Argentina, India, China, Indonesia, Mexico, Saudi Arabia, South Africa, South Korea, Turkey and the European Union (EU). 11 GJR-GARCH and cDCC parameters can be found in Table A.1 and the US dollar/local currencies are displayed in Fig. A.2 in Appendix A. On the other hand, for space saving purposes and because of their irrelevance to our analysis, we did not present the descriptive statistics of the raw returns. However, they can be obtained upon request. 12 US dollar index is a measure of the US dollar value relative to a basket of six major foreign currencies. It is calculated as 50.14348112 × (USD/EUR)0.576 × (USD/JPY)0.136 × (USD/GBP)0.119 × (USD/CAD)0.091 × (USD/SEK)0.042 × (USD/CHF)0.036 . An increase in the US dollar index indicates US dollar appreciation against other currencies in the basket. 13 For comparison purposes and to contribute further to the paper, we also perform our analysis by using the original DCC model of Engle (2002) (The estimated parameters of this model are not presented in the text, however they can be obtained upon request). The correlations obtained by this way and the corresponding mean shifts are visually displayed in Fig. A.1, and the exact shift dates are presented in Table A.2 in Appendix A. Accordingly, although we observe some slight changes in the shift dates compared to cDCC model, these are usually limited to a few days thus the qualitative results inferred from both approaches are indistinguishable. An exception exists in the case of India where the difference in the shift dates between cDCC and DCC model is around four months. The shift date arising with cDCC is right after the Lehman’s collapse. This situation might suggest that, besides the consistency factor, using cDCC might be a better choice and be more beneficial in this particular case from market participants’ and policymakers’ points of view. 10

404

M.I. Turhan et al. / Int. Fin. Markets, Inst. and Money 32 (2014) 397–414 Crude Oil−Euro 1

0.75

0.75

0.5

0.5

0.25

0.25 cDCC

cDCC

Crude Oil−Dollar index 1

0

0

−0.25

−0.25

−0.5

−0.5

−0.75

−0.75

Jan.00

Jul.02

Dec.04

Jun.07

Nov.09

May.12

Jan.00

Jul.02

1

1

0.75

0.75

0.5

0.5

0.25

0.25

0

−0.25

−0.5

−0.5

−0.75

−0.75 Jul.02

Dec.04

Jun.07

Nov.09

May.12

Jan.00

Jul.02

Nov.09

May.12

Dec.04

Jun.07

Nov.09

May.12

Crude Oil−South Africa 1

0.75

0.75

0.5

0.5

0.25

0.25 cDCC

cDCC

Crude Oil−Canada 1

0 −0.25

0 −0.25

−0.5

−0.5

−0.75

−0.75

Jan.00

Jun.07

0

−0.25

Jan.00

Dec.04

Crude Oil−Australia

cDCC

cDCC

Crude Oil−UK

Jul.02

Dec.04

Jun.07

Nov.09

May.12

Jan.00

Jul.02

Dec.04

Jun.07

Nov.09

May.12

Fig. 2. Mean level shifts in the dynamic correlations (estimated by cDCC) between crude oil price (in US dollar) and exchange rates (US dollar/local currency) of G20 members. Red and blue lines denote downwards and upwards shifts in the correlation levels respectively. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of the article.)

The dates of the second shifts in the correlations cluster in the interval of June–October 2008. After reaching an all time high of $147 per barrel in July 2008 on the back of a six-year commodity boom cycle, as of August 2008, oil prices plunged rapidly to $40 as demand from the OECD countries came to a sudden halt and recession loomed as the 2008 financial crisis, considered to be the worst financial crisis since the Great Depression of the 1930s, severely impacted the global economy. However, this time the shifts in exchange rate-oil price correlations present itself for both developed and emerging economies.

M.I. Turhan et al. / Int. Fin. Markets, Inst. and Money 32 (2014) 397–414 Crude Oil−Japan

1

1

0.75

0.75

0.5

0.5

0.25

0.25 cDCC

cDCC

Crude Oil−Russia

0

0

−0.25

−0.25

−0.5

−0.5

−0.75

−0.75

Jan.00

Jul.02

Dec.04

Jun.07

Nov.09

May.12

Jan.00

Jul.02

1

1

0.75

0.75

0.5

0.5

0.25

0.25

0 −0.25

−0.5 −0.75 Jul.02

Dec.04

Jun.07

Nov.09

May.12

Jan.00

Jul.02

May.12

Dec.04

Jun.07

Nov.09

May.12

Crude Oil−Indonesia

1

1

0.75

0.75

0.5

0.5

0.25

0.25 cDCC

cDCC

Crude Oil−Turkey

0 −0.25

0 −0.25

−0.5

−0.5

−0.75

−0.75 Jul.02

Dec.04

Jun.07

Nov.09

May.12

Jan.00

Jul.02

Crude Oil−Mexico 1

1

0.75

0.75

0.5

0.5

0.25

0.25

0

−0.25

−0.5

−0.5

−0.75

−0.75 Dec.04

Jun.07

Jun.07

Nov.09

May.12

0

−0.25

Jul.02

Dec.04

Crude Oil−South Korea

cDCC

cDCC

Nov.09

0

−0.5

Jan.00

Jun.07

−0.25

−0.75

Jan.00

Dec.04

Crude Oil−India

cDCC

cDCC

Crude Oil−Brazil

Jan.00

405

Nov.09

May.12

Jan.00

Fig. 2. (Continued )

Jul.02

Dec.04

Jun.07

Nov.09

May.12

406

M.I. Turhan et al. / Int. Fin. Markets, Inst. and Money 32 (2014) 397–414

Table 1 Shift dates of the dynamic correlation levels (estimated by cDCC) between crude oil prices and exchange rates of G20 members. Crude oil-Dollar index Date

ij

28/11/2003 05/06/2008

Down Down

Crude oil-Euro Date

ij

08/12/2003 05/06/2008

Down Down

Crude oil-UK Date

ij

06/01/2004 18/09/2008

Down Down

Crude oil-Australia Date

ij

30/01/2004 23/09/2008

Down Down

Crude oil-Canada Date

ij

21/03/2003 25/08/2008

Down Down

Crude oil-South Africa Date

ij

17/10/2003 02/01/2006 02/03/2009

Down Down Down

Crude oil-Russia Date

ij

09/10/2003 23/01/2008 02/03/2010

Down Down Down

Crude oil-Japan Date

ij

26/09/2008 01/10/2010

Up Down

Crude oil-Brazil Date

ij

31/07/2007

Down

Crude oil-India Date

ij

14/10/2008

Down

M.I. Turhan et al. / Int. Fin. Markets, Inst. and Money 32 (2014) 397–414

407

Table 1 (Continued ) Crude oil-Turkey Date

ij

26/09/2008

Down

Crude oil-Indonesia Date

ij

18/09/2008

Down

Crude oil-Mexico Date

ij

26/09/2008

Down

Crude oil-South Korea Date

ij

31/03/2009

Down

One thing to notice is that when a shift occurs in the correlation, it is in the negative direction in a strengthening way. And the shifts are irremeable, suggesting an increased permanent dependence between oil prices and exchange rates in the last decade. Moreover the correlation is symmetric in the sense that the negative co-movement degrees are at least preserved in both increasing and decreasing trends in the oil prices. On the other hand, an exceptional case exists for Japan where the correlation increases a little from its level around zero in the 2008 crisis and returns back to its previous level after a while. This is parallel to the findings of Lizardo and Mollick (2010), Reboredo (2012) and Aloui et al. (2013) where authors depend this situation on Japan’s being net oil importer and mostly imports from OPEC countries. Another observation is about the highly controversial Eurozone sovereign debt crisis. It started when five of the region’s countries; Portugal, Ireland, Italy, Greece and Spain, started to be questioned about their capability of fully paying back to their bondholders. Although these five are seen as the most problematic, their possible default is claimed to have far-reaching consequences beyond their borders. For a considerable amount of economists, this crisis has been argued to produce significant adverse effects on both financial markets and the real economies of several countries. In fact, the head of the Bank of England referred to it as “the most serious financial crisis at least since the 1930s, if not ever,” in October 2011. However, as our analysis shows, Eurozone debt crisis does not have a compelling effect on crude oil-exchange rate interaction, at least in the sense of a significant change in the correlations. 5. Conclusion and policy implications This study aims to analyze the co-movements of oil price (in US dollar) and exchange rates (US dollar/local currency) of G20 members from 2000 to 2013 using dynamic correlation modeling and penalized contrast function methodology. The study reveals that the link between oil prices and exchange rates has intensified in the last decade; they became strongly negatively correlated (which also associates an increase in the oil prices with the US dollar depreciation against other currencies). This result has twofold implications. The first implication is that such strengthened negative correlations provides risk diversification opportunities for the investors worldwide.14 Other implication concerns policymakers, in particular central

14 Indeed, Cifarelli and Paladino (2012) show that from the year 2000, oil returns tend to become more reactive to the bonds and stocks, more associated with real exchange rates and feedback trading is more pervasive. Accordingly, the trade-off between risk and returns indicates that in the last decade, oil diversifies away the empirical risk of portfolios.

408

M.I. Turhan et al. / Int. Fin. Markets, Inst. and Money 32 (2014) 397–414

banks: When oil prices increase, the cost of production and services are likely to increase hence the inflation tends to go up. However, the negative correlation creates a buffer in this case as the local currency will appreciate against the US dollar implying a decrease in the prices (in local currency) of the imported goods. Thus, a rise in the price of the weighted basket (to measure inflation) is expected to be limited. We found two events related with the significant shifts in the correlation levels. First event is the US’ invasion of Iraq in 2003. In this case, correlation levels shift downwards significantly mainly for the pairs oil price-exchange rates of developed economies. The possible reason beyond this fact is the following: In the 1990s, emerging economies of the G20 were dealing with high inflation, high debt/GDP ratio and economic–political instability leading to severe recessions. They were not major players of the global economic and financial system as they are now. Their integration to the global system was achieved in early 2000s after successful monetary and fiscal policy reforms. Thus, an insensitivity in the correlations is expected. The second event is the 2008 global financial crisis. Compared to 2003, it is more puzzling that why a financial crisis led to a strengthened link between oil prices and exchange rates. Iraq is a major oil supplier to the world, a political instability or a war in the country can produce an abrupt change in the correlation dynamics however this is not the case in 2008 financial crisis. We believe a possible reason is the increased use of oil as a financial asset. In particular, the financialization of the crude oil and its usage in other innovative financial instruments causes the oil prices to react to other asset prices i.e. they respond immediately to information captured in other assets. Moreover, this time the significant shift in the correlation levels are found to occur for all members of the G20 (except Japan). As explained above, reason lies within the fact that in 2008, compared to even five years ago, world financial and product markets get significantly more integrated: Since the emerging markets join the game, the share of goods and services that constitute the tradable sector and the cross-border flows of real and financial capital have increased dramatically. In both cases, the shifts in the correlation levels seem to be irremediable, suggesting a systemic change that creates permanent increase in the degree of negative co-movement. Furthermore, the correlation is symmetric in the sense that it does not respond differently in bear or bull global oil market periods. In contrast to the general belief, we observed no effect of the ongoing Eurozone debt crisis on the oil price-exchange rate correlation dynamics in the sense of a shift. World is focused on this crisis so much that the increasing political heat and the ongoing threats to political–economical stability in Middle East and North Africa (MENA) countries are ignored. MENA region, unlike Europe, currently contains most of the global oil reserves. Invasion of Iraq shows that a problem in the region can easily contribute to run up in oil prices, which may severely slow down the global economic growth and significantly alter the exchange rate-oil price correlation. Moreover, another important topic to be stressed is the effects of Fed’s possible tapering on the relationship between oil prices and exchange rates. Since the early 2013, it has been rumored that the excess liquidity environment will come to an end which was eventually validated by the speech of Bernanke on May 22, 2013. Without easy access to liquidity, the “flight to quality” effect may kick in which would possibly alter the correlation dynamics in favor of the pre-2008 status. It is too early to make such strong comments on these situations, however possible scenarios have to be taken into account by policymakers, risk managers and investors. Acknowledgements The authors thank the editor Geoffrey Booth and the anonymous referee for their valuable and insightful comments and suggestions that significantly improved this paper. Appendix A. See Tables A.1 and A.2 ; Figs. A.1 and A.2.

M.I. Turhan et al. / Int. Fin. Markets, Inst. and Money 32 (2014) 397–414 Crude Oil − Euro 1

0.75

0.75

0.75

0.5

0.5

0.5

0.25

0.25

0.25

0

cDCC

1

−0.25

0

−0.5

−0.5

−0.5

−0.75

−0.75

Jul.02

Dec.04

Jun.07

Nov.09

May.12

Jan.00

Jul.02

Dec.04

Jun.07

Nov.09

May.12

0.75

0.75

0.75

0.5

0.5

0.5

0.25

0.25

0.25

0

cDCC

1

−0.25

0 −0.25 −0.5

−0.5

−0.75

−0.75

Jun.07

Nov.09

May.12

Jan.00

Jul.02

Crude Oil − Brazil

Dec.04

Jun.07

Nov.09

May.12

0.75

0.75

0.5

0.5

0.5

0.25

0.25

0.25 cDCC

0.75

cDCC

1

0

0 −0.25 −0.5

−0.5

−0.75

−0.75

Nov.09

Jun.07

May.12

Jan.00

Jul.02

Crude Oil − Mexico

Dec.04

Nov.09

Jun.07

May.12

Jan.00

1

1

1

0.75

0.75 0.5 0.25 cDCC

0.5 0.25 cDCC

0.5 0.25 0

0

−0.5

−0.5

−0.5

−0.75

−0.75

Jun.07

Nov.09

May.12

Jan.00

Jul.02

Dec.04

Jun.07

Nov.09

May.12

Jun.07

Nov.09

May.12

Jan.00

Jul.02

Dec.04

Jun.07

Nov.09

May.12

Crude Oil − Indonesia

1

1

0.75

0.75

0.5

0.5

0.25

0.25 cDCC

cDCC

Crude Oil − Japan

0 −0.25

0 −0.25

−0.5

−0.5

−0.75

−0.75

Jan.00

Dec.04

0

−0.75 Dec.04

May.12

−0.25

−0.25

Jul.02

Nov.09

Crude Oil − Australia

0.75

Jan.00

Jul.02

Crude Oil − South Korea

−0.25

Jun.07

0

−0.5

Dec.04

Dec.04

−0.25

−0.75 Jul.02

May.12

Crude Oil − Turkey

1

Jan.00

Jul.02

Jan.00

Crude Oil − India

1

−0.25

Nov.09

0

−0.5

Dec.04

Jun.07

−0.25

−0.75 Jul.02

Dec.04

Crude Oil − Russia

1

Jan.00

Jul.02

Jan.00

Crude Oil − South Africa

1

cDCC

cDCC

Crude Oil − Canada

cDCC

0 −0.25

−0.25

−0.75 Jan.00

cDCC

Crude Oil − UK

1

cDCC

cDCC

Crude Oil − Dollar index

409

Jul.02

Dec.04

Jun.07

Nov.09

May.12

Jan.00

Jul.02

Dec.04

Jun.07

Nov.09

May.12

Fig. A.1. Mean level shifts in the dynamic correlations (estimated by DCC) between crude oil price (in US dollar) and exchange rates (US dollar/local currency) of G20 members. Vertical lines denote the shifts in the correlation levels.

410

M.I. Turhan et al. / Int. Fin. Markets, Inst. and Money 32 (2014) 397–414

100 90 80

1.1 1 0.9 0.8 0.7 0.6 Jan.00

Jul.02 Dec.04 Jun.07 Nov.09 May.12 Canada

0.7 0.65 0.6 0.55 0.5

Jul.02 Dec.04 Jun.07 Nov.09 May.12

Jan.00

Russia 36

USD / Local Currency

1.4 1.3 1.2 1.1

USD / Local Currency

13

1.5

12 11 10 9 8 7

1 Jul.02 Dec.04 Jun.07 Nov.09 May.12

Jan.00

34 32 30 28 26 24

6 0.9 Jan.00

Jul.02 Dec.04 Jun.07 Nov.09 May.12

Jan.00

Turkey 2

3.5 3 2.5 2

50

45

Jul.02 Dec.04 Jun.07 Nov.09 May.12

Jan.00

1.5

1

0.5

40

1.5

Jul.02 Dec.04 Jun.07 Nov.09 May.12

Jan.00

Jul.02 Dec.04 Jun.07 Nov.09 May.12

South Korea

Mexico

Australia

1600

14 13 12 11 10 9

2

1500 USD / Local Currency

USD / Local Currency

15

1400 1300 1200 1100 1000

1.8 1.6 1.4 1.2 1

900 Jul.02 Dec.04 Jun.07 Nov.09 May.12

Jan.00

Jul.02 Dec.04 Jun.07 Nov.09 May.12

USD / Local Currency

130 120 110 100 90

Jan.00

Jul.02 Dec.04 Jun.07 Nov.09 May.12

Indonesia

Japan

USD / Local Currency

USD / Local Currency

55

USD / Local Currency

USD / Local Currency

USD / Local Currency

4

Jan.00

Jul.02 Dec.04 Jun.07 Nov.09 May.12

India

Brazil

Jan.00

Jul.02 Dec.04 Jun.07 Nov.09 May.12

South Africa

1.6 USD / Local Currency

USD / Local Currency

USD / Local Currency

USD / Local Currency Basket

1.2

110

70 Jan.00

UK

Eurozone

Dollar index 120

12000 11000 10000 9000 8000

80 7000 Jan.00

Jul.02 Dec.04 Jun.07 Nov.09 May.12

Jan.00 Jul.02 Dec.04 Jun.07 Nov.09 May.12

Fig. A.2. Exchange rates from Jan 2000 to April 2013.

M.I. Turhan et al. / Int. Fin. Markets, Inst. and Money 32 (2014) 397–414

411

Table A.1 GJR-GARCH parameters for the filtered returns and the driving parameters of cDCC (between FX and crude oil prices) GJR-GARCH

Crude oil Dollar index Euro UK Australia Canada South Africa Russia Japan Brazil India Turkey Indonesia Mexico South Korea

cDCC

c × 104

a

g

b

˛

ˇ

0.033026 (0.0610) 0.00095124 (0.0491) 0.00162363 (0.0239) 0.00282407 (0.0026) 0.00830382 (0.0010) 0.00122943 (0.0254) 0.017699 (0.0187) 0.00037554 (0.0454) 0.00768078 (0.0806) 0.011682 (0.0014) 0.00065763 (0.0851) 0.114252 (0.0374) 0.00127044 (0.2222) 0.00560380 (0.0029) 0.00406782 (0.0043)

0.028179 (0.0058) 0.028965 (0.0000) 0.037034 (0.0000) 0.048780 (0.0000) 0.082013 (0.0000) 0.054625 (0.0000) 0.098098 (0.0000) 0.104305 (0.0000) 0.020900 (0.0133) 0.196417 (0.0000) 0.168019 (0.0057) 1.946204 (0.2266) 0.064864 (0.0006) 0.118645 (0.0000) 0.115222 (0.0000)

0.032219 (0.0171) 0.000386 (0.9580) −0.011532 (0.1317) −0.021001 (0.0235) −0.055437 (0.0004) −0.012750 (0.3098) −0.047113 (0.0136) −0.010199 (0.6251) 0.024447 (0.2841) −0.110992 (0.0000) −0.043734 (0.1855) −1.766365 (0.2530) −0.030819 (0.0371) −0.061266 (0.0014) −0.045517 (0.0149)

0.949573 (0.0000) 0.967471 (0.0000) 0.965145 (0.000) 0.952462 (0.000) 0.931019 (0.0000) 0.948479 (0.0000) 0.910842 (0.0000) 0.905972 (0.0000) 0.948120 (0.0000) 0.854532 (0.0000) 0.865794 (0.0000) 0.462347 (0.0056) 0.951312 (0.0000) 0.894862 (0.0000) 0.897504 (0.0000)

– – 0.004494 (0.5135) 0.004771 (0.0702) 0.004568 (0.2093) 0.008606 (0.0987) 0.010201 (0.0251) 0.008564 (0.4151) 0.005494 (0.0000) 0.017726 (0.0267) 0.012821 (0.0614) 0.004689 (0.0689) 0.018282 (0.0049) 0.002635 (0.8529) 0.014438 (0.0016) 0.008611 (0.0039)

– – 0.995496 (0.0000) 0.995219 (0.0000) 0.995422 (0.0000) 0.991253 (0.0000) 0.989594 (0.0000) 0.991426 (0.0000) 0.994496 (0.0000) 0.966376 (0.0000) 0.986038 (0.0000) 0.995301 (0.0000) 0.981028 (0.0000) 0.997355 (0.0000) 0.982929 (0.0000) 0.989663 (0.0000)

2 1. GJR-GARCH is estimated by hi,t = c + (a + gI yi,t−1 <0 )yi,t−1 + bhi,t−1 . ∗1/2

∗1/2

2. cDCC process is driven by Qt = (1 − ˛ − ˇ)S + ˛{Qt−1 εt−1 εt−1 Qt−1 } + ˇQt−1 . For a more precise estimation, each pairwise dynamic correlation is calculated separately thus, we have different driving parameters ˛ and ˇ for each pair of oil priceexchange rate. 3. The values in the parentheses are the p-values obtained from robust standard errors. Table A.2 Shift dates of the dynamic correlation levels (estimated by DCC) between crude oil prices and exchange rates of G20 members. Crude oil-Dollar index Date

ij

28/11/2003 23/05/2008a

Down Down

Crude oil-Euro Date

ij

08/12/2003 05/06/2008

Down Down

Crude oil-UK Date

ij

06/01/2004 18/09/2008

Down Down

412

M.I. Turhan et al. / Int. Fin. Markets, Inst. and Money 32 (2014) 397–414

Table A.2 (Continued ) Crude oil-Australia Date

ij

09/01/2004a 23/09/2008

Down Down

Crude oil-Canada Date

ij

21/03/2003 20/08/2008a

Down Down

Crude oil-South Africa Date

ij

17/10/2003 02/01/2006 02/03/2009

Down Down Down

Crude oil-Russia Date

ij

09/10/2003 23/01/2008 02/03/2010

Down Down Down

Crude oil-Japan Date

ij

26/09/2008 01/10/2010

Up Down

Crude oil-Brazil Date

ij

16/05/2007a

Down

Crude oil-India Date

ij

17/02/2009a

Down

Crude oil-Turkey Date

ij a

01/10/2008

Down

Crude oil-Indonesia Date

ij

18/09/2008

Down

Crude oil-Mexico Date

ij

26/09/2008

Down

Crude oil-South Korea Date

ij

31/03/2009

Down

a

The dates different than the cDCC case.

M.I. Turhan et al. / Int. Fin. Markets, Inst. and Money 32 (2014) 397–414

413

References Ahmad, A.H., Hernandez, R.M., 2013. Asymmetric adjustment between oil prices and exchange rates: empirical evidence from major oil producers and consumers. Journal of International Financial Markets Institutions & Money 27, 306–317. Aielli, G.P., 2013. Dynamic conditional correlation: on properties and estimation. Journal of Business & Economic Statistics 31, 282–299. Akram, Q.F., 2009. Commodity prices, interest rates and the dollar. Energy Economics 31, 838–851. Aloui, R., Ben Aissa, M.S., Nguyen, D.K., 2013. Conditional dependence structure between oil prices and exchange rates: a copula-GARCH approach. Journal of International Money and Finance 32, 719–738. Amano, R.A., Van Norden, S., 1998. Oil prices and the rise and fall of the US real exchange rate. Journal of International Money and Finance 17, 299–316. Askari, H., Krichene, N., 2010. An oil demand and supply model incorporating monetary policy. Energy 35, 2013–2021. Basher, S.A., Haug, A.A., Sadorsky, P., 2012. Oil prices, exchange rates and emerging stock markets. Energy Economics 34, 227–240. Beckmann, J., Czudaj, R., 2013a. Is there a homogeneous causality pattern between oil prices and currencies of oil importers and exporters? Energy Economics 40, 665–678. Beckmann, J., Czudaj, R., 2013b. Oil prices and effective dollar exchange rates. International Review of Economics & Finance 27, 621–636. Benassy-Quere, A., Mignon, V., Penot, A., 2007. China and the relationship between the oil price and the dollar. Energy Policy 35, 5795–5805. Benhmad, F., 2012. Modeling nonlinear Granger causality between the oil price and U.S. dollar: a wavelet based approach. Economic Modelling 29, 1505–1514. Blomberg, S.B., Harris, E.S., 1995. The commodity-consumer price connection: fact or fable? Federal Reserve Board of New York Economic Policy Review 1, 21–38. Buetzer, S., Habib, M.M., Stracca, L., 2012. Global Exchange Rate Configurations: Do Oil Shocks Matter? European Central Bank Working Paper Series 1442. Canrong, J., Shiqiang, L., 2010. The G20’s role is irreplaceable, November. China Daily. Cavalcanti, T., Jalles, J.T., 2013. Macroeconomic effects of oil price shocks in Brazil and in the United States. Applied Energy 104, 475–486. Chen, S.S., Chen, H.C., 2007. Oil prices and real exchange rates. Energy Economics 29, 390–404. Cifarelli, G., Paladino, G., 2010. Oil price dynamics and speculation: a multivariate financial approach. Energy Economics 32, 363–372. Cifarelli, G., Paladino, G., 2012. Can oil diversify away the unpriced risk of a portfolio? International Journal of Finance & Economics 17, 73–88. Coudert, V., Mignon, V., Penot, A., 2008. Oil price and the dollar. Energy Studies Review 15, 45–58. Dibooglu, S., 1996. Real disturbances, relative prices, and purchasing power parity. Journal of Macroeconomics 18, 69–87. Engle, R.F., 2002. Dynamic conditional correlation: a simple class of multivariate generalized autoregressive conditional heteroskedasticity models. Journal of Business & Economic Statistics 20, 339–350. Ghosh, S., 2011. Examining crude oil price-exchange rate nexus for India during the period of extreme oil price volatility. Applied Energy 88, 1886–1889. Golub, S., 1983. Oil prices and exchange rates. Economics Journal 93, 576–593. Hamilton, J.D., 1983. Oil and the macroeconomy since World War II. Journal of Political Economy 99, 228–248. Huang, Y., Guo, F., 2007. The role of oil price shocks on China’s real exchange rate. China Economic Review 18, 403–416. Jain, A., Ghosh, S., 2013. Dynamics of global oil prices, exchange rate and precious metal prices in India. Resources Policy 38, 88–93. Ji, Q., Fan, Y., 2012. How does oil price volatility affect non-energy commodity markets? Applied Energy 89, 273–280. Krichene, N., 2005. A Simultaneous Equations Model for World Crude Oil and Natural Gas Markets. IMF Working Paper Series 32. Krichene, N., 2006. World Crude Oil Markets: Monetary Policy and the Recent Oil Shock. IMF Working Paper Series 62. Krugman, P., 1983. Oil and the dollar. In: Bhandari, J.S., Putnam, B.H. (Eds.), Economic Interdependence and Flexible Exchange Rates. Cambridge University Press, Cambridge. Lavielle, M., 2005. Using penalized contrasts for the change-point problem. Signal Processing 85, 1501–1510. Lavielle, M., Teyssiere, G., 2007. Adaptive detection of multiple change-points in asset price volatility. In: Teyssiere, G., Kirman, A.P. (Eds.), Long Memory in Economics. Springer Berlin, Heidelberg. Lizardo, R.A., Mollick, A.V., 2010. Oil price fluctuations and US dollar exchange rates. Energy Economics 32, 399–408. Narayan, P.K., Narayan, S., Prasad, A., 2008. Understanding the oil price-exchange rate nexus for the Fiji islands. Energy Economics 30, 2686–2696. Rafiq, M.S., 2011. Sources of economic fluctuations in oil-exporting economies: implications for choice of exchange rate regimes. International Journal of Finance & Economics 16, 70–91. Raurava, J., 2004. The role of oil prices and the real exchange rate in Russia’s economy – a cointegration approach. Journal of Comparative Economics 32, 315–327. Reboredo, J.C., 2012. Modelling oil price and exchange rate co-movements. Journal of Policy Modeling 34, 419–440. Reboredo, J.C., Rivera-Castro, M.A., 2013. A wavelet decomposition approach to crude oil price and exchange rate dependence. Economic Modelling 32, 42–57. Reboredo, J.C., Rivera-Castro, M.A., Zebende, G.F., 2014. Oil and US dollar exchange rate dependence: a detrended crosscorrelation approach. Energy Economics 42, 132–139. Rogoff, K., 1991. Oil, Productivity, Government Spending and the Real Yen-dollar Exchange Rate. Federal Reserve Bank of San Francisco Working Paper Series 91 (6).

414

M.I. Turhan et al. / Int. Fin. Markets, Inst. and Money 32 (2014) 397–414

Sari, R., Hammoudeh, S., Soytas, U., 2010. Dynamics of oil price, precious metal prices, and exchange rate. Energy Economics 32, 351–362. Sensoy, A., 2013. Dynamic relationship between precious metals. Resources Policy 38, 504–511. Turhan, M.I., Hacihasanoglu, E., Soytas, U., 2013. Oil prices and emerging market exchange rates. Emerging Markets Finance and Trade 49, 21–36. Wang, Y., Wu, C., 2012. Energy prices and exchange rates of the U.S. dollar: further evidence from linear and nonlinear causality analysis. Economic Modelling 29, 2289–2297. Wu, C.C., Chung, H., Chang, Y.H., 2012. The economic value of co-movement between oil price and exchange rate using copula based GARCH models. Energy Economics 34, 270–282. Yousefi, A., Wirjanto, T.S., 2004. The empirical role of the exchange rate on the crude-oil price formation. Energy Economics 26, 783–799. Zhang, Y.J., Fan, Y., Tsai, H.T., Wei, Y.M., 2008. Spillover effect of US dollar exchange rate on oil prices. Journal of Policy Modeling 30, 973–991.