Does oil prices impede Islamic stock indices? Fresh insights from wavelet-based quantile-on-quantile approach

Resources Policy 62 (2019) 292–304

Contents lists available at ScienceDirect

Resources Policy journal homepage: www.elsevier.com/locate/resourpol

Does oil prices impede Islamic stock indices? Fresh insights from waveletbased quantile-on-quantile approach

T

Shekhar Mishraa, Arshian Sharifb,∗, Sashikanta Khuntiac, Saeed Aas Meod, Syed Abdul Rehman Khane a

Department of Business Management, C.V. Raman College of Engineering, Bhubaneswar, Odisha, India Othman Yeop Abdullah Graduate School of Business, University Utara Malaysia, Sintok, Kedah, Malaysia c Department of Management Studies, Indian Institute of Technology (Indian School of Mines) Dhanbad, Jharkhand, India d The Superior College, Lahore, Pakistan e School of Economics and Management, Tsinghua University, Beijing, China b

A R T I C LE I N FO

A B S T R A C T

Keywords: Islamic stock index Oil prices Wavelet-based quantile-on-quantile

The present paper analyses the association among the fluctuations in global crude oil prices and the Dow Jones Islamic Stock Index by employing daily data from 1st January 1996 to 13th April 2018. The paper uses a novel approach of Wavelet-based Quantile-on-Quantile Regression Model to examine the influence of different quantiles of the decomposed time series of WTI Brent Crude Oil Prices on quantiles of Islamic stock index. The outcome of the study indicates the heterogeneity in the influence of global crude oil prices on Islamic Stock Index. The original time series data exerts the positive influence across all the quantiles. When we decompose the series, we find the positive influence starts decreasing and with the advent of stability in the time series of global crude oil prices, the negative effect becomes stronger. The analysis of the present study indicates that oil prices fluctuations may have a positive effect on Islamic stock index in short run, but on attaining stability, the oil prices exert a negative influence on the Islamic stock index.

1. Introduction Evidence of dramatic variability in oil price movement and subsequent changes in stock prices has attracted interest of investors, policymakers, and academic researchers in exploring the nexus between oil price shocks and stock returns. The underlying economic theory asserts that oil price changes transmitted to stock returns via production cost can be better explained by the equity valuation theory. The concept of equity valuation assumes that the current stock price is the reflection of the present values of all expected future cash flows. All future cash flows depend on macroeconomic conditions and events with a significant relationship with oil price variability (see Jouini, 2013). In relation to this, several studies have examined the explanatory power of oil price shocks on equity market returns (see Sadorsky, 1999; Ciner, 2001; Basher and Sadorsky, 2006; Elyasiani et al., 2011; Cunado and de Gracia, 2014; Diaz et al., 2016; Wei and Guo, 2017; Tiwari et al., 2018). A particular, strand of studies observed the negative effects of oil prices shocks on stock returns (Sadorsky, 1999; Ciner, 2001; Miller and Ratti, 2009) while another exhibited positive (Zhu et al., 2011; Li et al., 2012)

and no effects (Apergis and Miller, 2009) of oil price variability on stock returns. Moreover, Kilian and Park (2009) observed that despite the consensus on the positive association between oil price variability and stock returns in oil-exporting countries, it remains unclear for oil-importing countries. Most of the previous studies on this topic have examined the nexus of oil price shocks with the stock indices of developed (Sadorsky, 2008; Mollick and Assefa, 2013; Tsai, 2015) and emerging markets (Gupta and Modise, 2013; Cunado and de Gracia, 2014). For instance, Ghouri (2006) found an inverse association between oil prices and United States (U.S.) stock returns. Similarly, Hammoudeh and Choi (2007) and Basher and Sadorsky (2006) revealed the inverse association between oil price shocks and stock returns for emerging markets. Further, it is observed that existing literature in a different note have categorized the stock market on the basis of net oil export and import of that particular nation and subsequently examined impact of oil price shocks on stock returns (see Park and Ratti, 2008; Wang et al., 2013; Salisu and Isah, 2017). Nandha and Faff (2008) revealed the significant positive influence of oil price shocks on stock earnings of oil exporting countries.

∗

Corresponding author. E-mail addresses: [email protected] (S. Mishra), [email protected] (A. Sharif), [email protected] (S. Khuntia), [email protected] (S.A. Meo), [email protected] (S.A. Rehman Khan). https://doi.org/10.1016/j.resourpol.2019.04.005 Received 7 December 2018; Received in revised form 8 March 2019; Accepted 15 April 2019 0301-4207/ © 2019 Elsevier Ltd. All rights reserved.

Resources Policy 62 (2019) 292–304

S. Mishra, et al.

might be the exclusion of information regarding frequency in the existing time series research. Hence, in this backdrop, the present study made an attempt to examine the impact of oil price shocks on Islamic stock and contributed to the literature in the following ways. First, we have implemented wavelet-based quantile-on-quantile regression to estimate the influence of oil price on stock returns. It is observed that there is little evidence about the effect of inter quantile movement of oil price on different quantiles of equity returns (Reboredo and Ugolini, 2016). Thus, in this study to fill the existing vacuum, we have implemented a recently developed model quantileon-quantile regression by Sim and Zhou (2015) to test the explanatory power of oil prices shocks on the Dow Jones Islamic stock index. The present study is the first as per the best knowledge of the authors, which considers nonlinearities issue and time-frequency issue using WaveletBased Quantile-on-Quantile Approach that produce comprehensive outcomes as compared to other time series econometrical studies. The Quantile on Quantile Regression Approach is instrumental in presenting a complete picture of the dependency between the variables by capturing the relationship between them at all the points of their conditional distribution. The same was not possible in the conventional Quantile Regression Approach. Second, it is observed that past studies on a similar line mostly focused on the conventional stock index. Thus, an investigation on Islamic stock indices may exhibit different outcomes as it possesses distinct features in comparison to conventional stock indices. Unlike, Badeeb and Lean (2018) and Narayan et al. (2019) this study has used comparatively robust model and our emphasis on Dow Jones Islamic index can overcome the possible effect of thin trading, which may lead to spurious findings. Moreover, we also utilize a daily dataset on global oil price and DJ Islamic Stock Index for the period of 1996–2018. This type of longitudinal data set can uncover the meaningful economic relationship between the two sets of variables in an empirical investigation. The outcome of the study indicates the heterogenic nature of the relationship between the global crude oil price and the Dow Jones Islamic Stock Index. Although the original time series data of global crude oil price exert a positive influence on Islamic stock index across all the quantiles, but when we decompose the given explanatory variable data, we find a negative impact on the Islamic stock index becoming stronger with the advent of stability in the time series data of global crude oil price. The remaining sections of the study are presented as follows. Section 2 describes the methodology adopted; Section 3 covers the estimated outcomes and discusses the empirical findings and Section 4 of the study delinates the conclusion and policy recommendations.

However, Cunado and de Gracia (2014) documented the significant negative impact of oil price change on stock returns of 12 oil-importing European markets. Aloui et al. (2012) confirmed an insignificant link between association among oil prices and stock returns of 25 emerging markets of oil-importing countries. Ramos and Veiga (2011) observed that the impact of oil price changes on an individual nation's stock returns is dependent on the level of oil dependency of that particular nation. Furthermore, it is observed that existing studies primarily used linear and nonlinear models to explore the effect of oil price shocks on stock returns (see, e.g., Filis et al., 2011; Chang and Yu, 2013; Zhang and Li, 2016; Badeeb and Lean, 2018). Apart from conventional stocks, recently Islamic stocks because of its innovative features have drawn the attention which is evident from the literature. For instance, Hussin et al. (2012) documented that Islamic stocks listed in Malaysia stock markets are not cointegrated with oil price in the long-run. Abdullah et al. (2016) investigated the impact of crude oil price on Islamic indices of South East Asian countries focusing on portfolio diversification and concluded that investors having crude oil in their basket can be benefited by adding Malaysian Islamic stock index in their portfolio. Arshad (2017) analyzing the relationship between oil price and Islamic stock markets observed the concurrent movement of volatility in oil price and Islamic stock markets. The reason for such concurrent movement is that Islamic stocks are generally dependent on the real economy, which is heavily exposed to oil price change. Further, Badeeb and Lean (2018) explored the asymmetric impact of oil price changes on Islamic stocks, particularly on sectoral perspective. The study concluded that in the long-run many Islamic sectoral stocks follow a nonlinear pattern in response to oil price shocks; whereas, in the short-run, these indices show a linear relationship to oil price variation. Recently, Narayan et al. (2019) examining an extensive set of individual Islamic stocks revealed that effect of oil price shocks on Islamic stocks are not homogeneous and trading with oil price sensitivity results annualized return of 5.8–13.6%. Raza et al. (2016) employed NARDL model and confirmed negative impact of oil price on emerging stock markets. Bouri et al. (2017) also confirmed nonlinear relationship between oil price and stock markets. Ftiti and Hadhri (2019) confirmed that lagged oil price could be used to forecast Islamic stock returns. Hassan et al. (2019) found that in BRIC there is not a high correlation among oil price and Islamic stocks. Bahloul et al. (2017) confirmed that oil price granger cause to Islamic stocks. The overall impression from the literature on oil price shocks and stock returns indicates that a relatively larger portion of the studies is dedicated to conventional stocks. However, we consider Islamic stocks because of its distinct features, innovation, and rapid expansion and an alternative for investment diversification. Fundamentally Islamic indices are composed with stocks going through a rigorous screening process. Stocks are screened to exclude any firms involved with nonSharia activities namely gambling and production of alcohol and tobacco. Moreover, Islamic finance has different fundamentals than that of conventional finance. Conventional finance is general interest based and associated with debt whereas Islamic finance is asset-based and asset-driven. Moreover, on methodological perspectives we can witness that majority of the past studies have adopted the simplistic empirical approach and linear models (Arshad, 2017; Badeeb and Lean, 2018). All the standard time series econometric models, present information related to time and ignore the important information pertaining to the frequency domain (Power and Turvey, 2010; Huang et al., 2016). Conversely, hiding the frequency information is a major cause of nonlinearities in the time series examination (Pal, and Mitra, 2017). The association between oil price shocks and Islamic stock is generally expected to be nonlinear. Badeeb and Lean (2018) used nonlinear ARDL cointegration approach and found that oil price shocks and Islamic stock has a nonlinear association. There are various studies which explored the cointegrating relationship between oil price and Islamic index. Unfortunately, findings are often mixed. The possible reason

2. Methodology 2.1. Quantile autoregressive unit root test The present study tested stationary properties of the series on the conditional mean as well as at each quantile of the conditional distribution. To achieve the same, the QAR (Quantile Auto-Regressive) unit root test is being considered advanced by Koenker and Xiao (2004). The QAR model was generalized by Galvao (2009) by incorporating covariates and linear time trend into the model. Let Yt denotes the strict stationarity with a previous information set ItY = (Yt − 1, ….,Yt − s ) ′ε s . We assume FY (.|ItY ) as the conditional distribution function of Yt given ItY . The QAR unit root is performed on the basis of linear QRM (Quantile regression model) represented as follows: p

QτY (Yt |ItY ) = μ1 (τ ) + μ 2 (τ ) t + α (τ ) Yt − 1 +

∑ αj (τ ) ΔYt−j + Fu−1 (τ ) j=1

QτY

(.|ItY )

FX (.|ItY )

(1)

is termed as the ϑ quantile of The drift term is represented by φ1 (ϑ) . t and ω (ϑ) represents the linear trend and

293

Resources Policy 62 (2019) 292–304

S. Mishra, et al.

persistence parameter respectively. z u−1 is termed as the inverse conditional distribution of the errors, for the quantiles τεT ⊂ [0,1]. Thus the numerous tenacity parameters (αˆ ) for each quantile of the conditional distribution of Xt is estimated. Koenker and Xiao (2004) and Galvao (2009) anticipated t-statistics for various quantiles τεT to test the null hypothesis H0: α (τ ) = 1 and QAR model follow this t-statistics.

rise to scaling coefficients, whereas in mother wavelets, differencing coefficients are generated. The father wavelet is represented as follows:

ϕj, k = −2−j /2ϕ ⎛ ⎝ ⎜

ψj, k = −2−j /2ψ ⎛ ⎝ ⎜

The novelty of the present paper lies in its endeavor to identify the systematic influence of different frequencies of global crude oil prices variable on the location, scale, and shape of the DJ-Islamic Stock Index. To achieve the same, we employ Xiao’s (2009) Quantile Cointegration Test. The remove the bottleneck of endogeneity in a standard cointegration model, Xiao (2009) decomposed the errors of the cointegration equation into lead-lag terms and a pure innovation component by following the ideologies of Saikkonen (1991). Furthermore, the QCM (Quantile Cointegration Model) is the extension of cointegration proposed by Engle and Granger (1987), which is considered to be special with β(τ) as a vector of constants. The special case consists of:

∑

(6)

(7)

j = 1... ... ... J

(8)

The maximal scale of the former is 2 , whereas the detailed are derived from the mother wavelets at all scales from 1 to J. The function f (.) from the coefficients as mentioned earlier can be defined as follows:

∑ SJ ,k ϕJ ,k (t ) + ∑ dJ ,k ψJ ,k (t )….+ ∑ dJ ,k ψj,k (t )…+ ∑ d1,k ψ1,k (t ),

f (t ) =

k

k

k

k

ΔZt′− j Пj + Fu−1 (τ )

(9) (3)

j =−k

On simplifying Equation (5), we get

In the given quadratic cointegration model, we also add a quadratic term of the regressor. The same can be represented as follows:

f (t ) = SJ + DJ + DJ − 1 + …+Dj + …+D1

∑

(10)

The orthogonal components are represented as follows:

k

QτY (Yt |ItY . Itz ) = α (τ ) + β (τ )′Zt + γ (τ )′Zt2 +

ΔZt′− j Пj

SJ =

j =−k

∑ SJ ,k ϕJ ,k (t ),

(11)

k

k

∑

∫ ψ (t ) dt = 0

j

k

ΔZt2−′ j Гj Fu−1 (τ )

j =−k

⎟

∫ f (t ) ψJ ,k With

dj, k =

and

+

t − 2 jk ⎞ with 2j ⎠

The detail coefficients derived from the mother wavelet are defined as follows:

(2)

∑

(5)

∫ f (t ) ϕJ ,k

Sj, k =

ΔZt′− j Пj + ut

QτY (Yt |ItY . Itz ) = α (τ ) + β (τ )′Zt +

∫ ϕ (t ) dt = 1

The father and mother wavelets constitute the basic functions which define the sequence of coefficients. The smooth coefficients derived from the father wavelets are shown as follows:

k j = −k

⎟

The mother wavelet is represented as follows:

2.2. Quantile Cointegration Test

Yt = α + β′Zt +

t − 2 jk ⎞ with 2j ⎠

DJ =

(4)

∑ dJ ,k ψJ ,k (t ). k

From the given equation (4), Xiao (2009) deduced the cointegrating coefficients' stability test. Xiao (2009) took the null hypothesis as H0: β (τ ) = β over all the quantiles. Under the null hypothesis, the researcher proposed a supermum rule of the absolute value of the difference Vˆn (τ ) = (βˆ (τ ) − βˆ) as a test statistic. On the basis of this test statistics, the present paper employs test statistic supτ Vˆn (τ ) across all the quantiles’ distribution. On the basis of work of Xiao (2009), the paper does 1000 Monte Carlo simulations to estimate the critical values of supτ Vˆn (τ ) test statistic.

j = 1,&....J (12)

The multi horizon or multi-resolution breakdown of f(t) is represented as {SJ, DJ-1,…, D1}. Dj calculates the jth level wavelet detail which is related with variations in the series at scale λj... Sj is defined as the cumulative sum of alterations at each level. As j increases, Sj becomes smoother and smoother (Gencay et al., 2002). We employ MODWT (Maximal Overlap Discrete Wavelet Transform) for estimating the scaling and wavelet coefficients. There is one of the appealing benefits of MODWT, and that is, it does not suffer from any limitation such as that level of a sample size to an integer multiple of 2 J0 .While its limitation of Discrete Wavelet Transform (DWT) (Percival and Walden, 2000), that is the reason that we prefer MODWT over. The detail and smooth coefficients of a MODWT are related with zero phase filters, which helps in aligning the features of original time series the features of Multi resolution Analysis (MRA). It is considering that DWT-based estimators are asymptomatically less efficient compare to MODWT (Percival, 1995; Percival and Mofjeld, 1997; Gencay et al., 2002). Furthermore, Discrete Wavelet Transform employs weighted differences and makes an average of attached sets of observations, whereas MODWT employs moving difference and average operator, thus keeping the exact number of observations at each wavelet decomposition scale. The present paper employs Daubechies Least Asymmetric (LA) filter of length 8 (LA8). According to Gencay et al. (2002), the LA(8) wavelet is considered to be smoother than HAAR wavelet filters which are widely used in the previous studies. According to Cornish et al. (2006), the LA(8) filter provides better uncorrelatedness across scales as compared to its HAAR filter counterpart. We decompose the series into wavelet coefficients D1 to D9. The

2.3. Wavelet multiscale decomposition The Time and frequency domain of any time series get combined in the Wavelet Analysis. Unlike other econometric methods, the wavelets decompose the time series under study into various wavelet scales. According to Ramsey (1999), the wavelets generate the orthogonal timescale decomposition of the data and give a nonparametric representation to all the individual time series (Ramsey, 1999). They can preserve the time data while performing the frequency decomposition of the series. The Wavelet Transform captures all the information in the time series associated with specific time horizons and time locations (Gencay et al., 2002). This feature of the wavelets makes it possible to deal with the non-stationary property of the time series. The father (ϕ) and mother (ψ) wavelets can represent any function of time (Ramsey, 2002). The father wavelets integrate into one and represent very long scale smooth components of the signal. On the contrary, mother wavelets integrate to zero and are used to represent deviations occurring in the smooth components. Father wavelets give 294

Resources Policy 62 (2019) 292–304

S. Mishra, et al.

detail coefficient Dj gives the resolution of data at scale 2j to 2j+1. The oscillations of periods 2–4, 4–8, 8–16, 16–32, 32–64, 64–128, 128–256, 256–512 and 512–1024 days are represented by λ1, λ2, λ3, λ4, λ5, λ6, λ7, λ8, λ9 respectively. The long-term movements are represented by wavelet smooth S9.

Further the paper analyses the impact various frequencies of oil prices returns, denoted by Oil.d1, Oil.d2, Oil.d3, Oil.d4, Oil.d5, Oil.d6, Oil.d7, Oil.d8, Oil.d9 and Oil.s9 on DJ Islamic Stock Index Returns. In the present paper, we employ Gaussian kernel to weight the observations in the neighborhood of the empirical quantile of uncertainty, based on the specific bandwidth for estimating the effect of τ quantile of uncertainty exerted locally.

2.4. Quantile on quantile approach Majority of papers in a bid to analyze and estimate the relation between any two-time series data, initially employ linear regression and then move on to Koenker and Bassett’s (1978) Quantile Regression framework. The Quantile Regression Analysis (QRA) since inception has become one of the most sought-after econometric tools to model and analyze the time-varying degree and dependence structure between the time series data. The major bottleneck, QRA framework suffers from is that it is unable to capture the entire dependency. Although the QRA framework can analyze and estimate the heterogeneous relationship between the variables at various points of conditional distribution, it doesn't accommodate the fact that nature of uncertainty may also affect the relationship between dependent and independent variable. Owing to this lacuna of QRA, Sim and Zhou (2015) introduced the Quantile on Quantile (QQ) Approach. Under the QQ Approach, the present paper models the quantile of Islamic Stock Index returns as a function of global crude oil prices returns (and its frequencies). The QQ Approach would help in capturing the variations in the relationship between the variables at each point of their conditional distribution, thus providing a clear and complete picture of the dependency relationship. The paper implements the QQ framework by selecting a number of uncertainty quantiles and by analyzing and estimating the local effect of global oil prices uncertainty and its various frequencies on the various quantiles of Islamic Stock returns. In recent years, a remarkable work was done by using Quantileon-Quantile approach (Sharif et al., 2019a, 2019b; Shahbaz et al., 2018). The estimation of Quantile on Quantile (QQ) Regression model involves two approaches to viz. (1) The triangular system of equations proposed by Ma and Koenker (2006) and (2) single equation regression approach proposed by Sim and Zhou (2015). The single equation regression approach is also based upon Ma and Koenker (2006) and is at this moment employed in the paper for analysis and estimation. The approach can be explained as follows. The quantile of the DJ- Islamic Stock Index Returns (ISL) is denoted by θ. Initially, we model the θ-quantile of ISL returns function of its past lag ISLt-1, and global crude oil prices indicated as Oilt. The same can be represented as follows:

ISLt = β θOilt + εtθ

2.5. Ascertaining QQ Method's validity The Quantile on Quantile (QQ) Regression approach, decomposes the estimates obtained from the standard quantile regression model. With the QQ approach, for different quantiles of independent variables, we obtain specific estimates. In the present paper, the quantile regression model regresses the θth quantile of global crude oil prices and its various decomposed series on the DJ-Islamic stock index. The parameters obtained from the quantile regression model are indexed by θ. On the contrary, the QQ analysis regresses the θth quantile of global crude oil prices on τth quantile of DJ Islamic stock market index, and its parameters are indexed by θ and τ. As compared to the quantile regression model, we can obtain more disaggregated information about the relation between oil prices and the Islamic stock market index under QQ analysis framework. The QQ method assumes the relationship between the variables across the different quantiles to be potentially heterogeneous. By this decomposition property inherent in the QQ approach, we can employ the QQ estimates to recover standard quantile regression estimates. The quantile regression parameters generally only indexed by θ can be obtained by averaging the QQ parameters along with τ. The slope coefficient of the quantile regression model, measuring the effect of global crude oil prices and its various decomposed series on DJ Islamic stock market index represented by γ1 (θ) can be deduced as follows:

1 γ1 (θ) ≡ βˆ1 (θ) = s

t)

≈

β θ (Oil τ )

3. Empirical analysis The dataset employed in the present paper comprising of WTI Crude Oil Prices as a proxy for global crude oil prices and Dow Jones (DJ) Islamic Stock Index is sourced from DataStream. The given dataset constitutes 5815 daily observations of both the variables and covers the period from January 01, 1996 to April 13, 2018. The summary statistics and correlation between the global crude oil prices and DJ Islamic Stock Index is presented in Table 1. The significant Jarque-Bera test statistics confirms non-normal distribution nature of WTI Crude Oil

(13)

+β

θ′

(Oil τ )(Oil

t

−

Oil τ ).

Table 1 Descriptive statistics.

(14)

β θ (Oil τ )

θ′

From a study done by Sim and Zhou (2015), and β (Oil τ ) can be redefined as β0 (θ , τ ) and β1 (θ , τ ) respectively. Equation (10) can be rewritten as follows:

β θ (Oilt ) ≈ β0 (θ , τ ) + β1 (θ , τ )(Oilt − Oil τ )

(15)

On substituting Equation (15) into Equation (13), we get,

ISLt = β0 (θ , τ ) + β1 (θ , τ )(Oilt − Oil τ ) + εtθ

(17)

τ

The validity of the QQ approach is ascertained by comparing the estimated quantile regression parameters with the τ averaged QQ parameters.

εtθ is the error term having zero θ-quantile. The relationship function β θ (.) is assumed to be unknown as we do not have prior knowledge about the interlinkage between DJ-Islamic Stock Index returns and changes in Global crude oil prices returns. The paper examines the association between the θ-quantile of DJ Islamic Stock Index Returns and θ-quantile of global crude oil prices, denoted by Oilτ by linearizing the function β θ (.) with first- order Taylor expansion of β θ (.) around Oilτ. The same is represented as follows: β θ (Oil

∑ βˆ1 (θ, τ )

(16)

Parameters

Oil P

ISL

Mean Median Standard Deviation Skewness Kurtosis Minimum Maximum Jarque-Bera Test (Probability) Correlation Matrix OIL ISL

3.822 3.895 0.608 −0.277 1.990 2.381 4.978 321.123 (0.000)

7.727 7.749 0.416 −0.100 2.350 6.739 8.615 112.192 (0.000)

1.000 0.640***

– 1.000

Note: *** indicates that variables are significant at 1% level of significance. 295

Resources Policy 62 (2019) 292–304

S. Mishra, et al.

Fig. 1. Trend plot of oil prices and DJ-Islamic market index.

summary statistics show the evidence of non-normal distribution of the variables and thus suggest for adopting a quantile-based approach for the analysis to deal with the problem of heavy tails.

Prices and DJ Islamic Stock Index at 1% level of significance (Table 1). The presence of non-normality in the dataset also highlights the absence of linear association among the variables of interest, therefore, we opt to analyze the nonlinear association between global oil price and DJ- Islamic stock index by applying quantile approaches (Bekiros et al., 2016; Balcilar et al., 2017; Troster et al., 2018; Sharif et al., 2019a). Finally, the coefficient of correlation between the variables as observed from the table is positive and higher than 0.64. The novel approach adopted in this paper is to analyze the impact of various frequencies of global crude oil prices on DJ Islamic Stock Index. The paper analyzes the association between the time series of the variables under study within the wavelet framework. The raw returns data of the global crude oil prices is decomposed into nine different frequency components by employing the wavelets. The plot of raw returns data of global crude oil prices and DJ Islamic Stock Index as well as decomposed components of global crude oil prices is demonstrated in Fig. 1. Fig. 1 shows the evidence of high frequencies in a short period with fluctuations becoming more stable in the longer periods. The

3.1. Quantile unit root test Table 2 exhibits the results of Quantile unit root test. The Quantile Unit Root Test illustrates the estimates of persistence and the t-statistics for the null hypothesis postulating that H0: a (τ) = 1 for Eq. (1) for the grid of 19 quantiles T = {0.05:0.95}. We employed 10 lags of the endogenous variable to overcome serial correlation. The stationary test documents presence of unit root at level for global crude oil prices and DJ-Islamic Stock Index Prices for different conditional distribution quantiles. The results of the unit root test confirm that all the variables are showing non-stationary behavior at level series (Table 2).

296

Resources Policy 62 (2019) 292–304

S. Mishra, et al.

Index and global crude oil price and its various decomposed series. The QQ analysis presented in Fig. 3 illustrates the slope coefficient β1, (θ , τ ) estimates capturing the effect of τth quantile of oil prices and its decomposed series θth quantile of DJ Islamic Stock Index at dissimilar values of θ and τ. In the original series, the overall positive and but weak effect is observed from OIL to ISL. The OIL and ISL linkages provide significant and positive value for the significant number of groupings of quantiles, suggesting that there is a direct relationship between OIL and ISL. In fact, a comparatively noticeable effect with the positive sign was detected in the region that combines the upper quantiles of oil prices (i.e., 0.80–0.95) with the link to the upper quantiles of Islamic market index (i.e., 0.80–0.95). Also, a noteworthy effect with the positive sign was detected in the region that syndicates the lower quantiles of oil prices (i.e., 0.35–0.55) with the link to the lower quantiles of Islamic market index (i.e., 0.30–0.65). Overall, the influence of OIL on ISL is weak in the middle all the quantiles of oil prices and the Islamic market index. However, the positive impact gets stronger on extreme (i.e., low and high quantiles) tails of both variables. This outcome recommends that sharp boost in the Islamic market index by oil prices which is represented by the lowest and highest quantiles of oil prices. In general, the results reveal a direct association between OIL and ISL. Technically, it may concluded that oil prices and the Islamic market index are complements for long-term sustainable financial development. On analyzing the influence of decomposed series Oil.d1 on Islamic stock market index, we observe the very insignificant influence of Oil prices. d1 on ISL in the middle quantiles of Oil.d1 and Islamic Stock Index. However, at lower quantiles of Islamic Stock Index (0.05–0.35) and Oil.d1 (0.15–0.45), a strong positive effect of Oil.d1 on Islamic Stock Market Index is observed. Similarly, we observe the significant positive effect in the region combining the upper quantiles of Islamic Stock Market Index (i.e., 0.8–0.9) with upper quantiles of Oil.d1 (0.7–0.85). From Fig. 3(8), we observe positive influence gaining strength at lower and higher quantiles of both the variables under study. This indicates the presence of positive influence of decomposed series Oil.d1 on Islamic Stock Market Index. When we analyze the influence of decomposed time series Oil.d2 on Islamic Stock Market Index, we observe a strong positive effect in the region adjoining the lowest quantiles of Islamic stock market index (0.05–0.25) and Oil.d2 (0.1–0.25). In the region adjoining the lower quantiles of Islamic stock (0.3–0.55) and Oil.d2 (0.25–0.5), the strength of the positive effect of Oil.d2 on Islamic Stock Market starts decreasing. At the higher quantiles of Islamic Stock Market (0.70–0.95) and Oil.d2 (0.7–0.9), the positive effect is extremely weak. The overall effect at middle quantiles of both Islamic Stock Market and Oil.d2 is negligible. The similar scenario is also observed in the effect of decomposed time series Oil.d3 on Islamic Stock. We observe a strong effect of Oil.d3 on Islamic Stock in the region adjoining the lowest quantile of Islamic Stock Market (0.05–0.2) and Oil.d3 (0.02–0.25). In the region of lower quantiles of Islamic Stock Market (0.3–0.55) and Oil.d3 (0.35–0.65), the positive influence of oil prices starts becoming weaker. In the region adjoining higher quantiles of Islamic Stock Market (0.75–0.95) and Oil.d3 (0.80–0.95) also, there exists a very weak positive influence of Oil.d3 on Islamic stock market. In the region of middle quantiles, the influence of Oil prices is extremely insignificant. From the analysis of the influence of decomposed series of Oil.d1, Oil.d2 and Oil.d3 on Islamic Stock Market; we find as we decompose the time series of Oil prices, its positive effect on Islamic stock market starts diminishing and the zero effect starts increasing. When we further decompose the time series data of oil prices and analyze the influence of Oil.d4 on Islamic Stock market, we observe the positive effect of Oil.d4 getting diminished, and the negative effect starts increasing. The effect of Oil.d4 on Islamic Stock Market is either zero or negative in the upper quantiles of Islamic Stock Market Index (0.75–0.9) and middle to upper quantiles of Oil.d4 (0.55–0.9). At the lower to middle quantiles of Islamic Stock market and Oil.d4, the

Table 2 Quantile Unit Root test. Quantile

0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.55 0.60 0.65 0.70 0.75 0.80 0.85 0.90 0.95

DJ-Islamic Market Index

Oil Prices

α(τ)

t-stats

C.V

α(τ)

t-stats

C.V

0.989 0.990 0.991 0.997 0.999 0.999 0.999 0.999 0.999 0.999 0.999 0.999 0.999 0.999 0.999 0.998 0.998 0.999 1.003

−2.487 −2.301 −1.981 −0.712 −0.454 −0.475 −0.627 −1.247 −1.846 −2.118 −1.034 −0.686 −0.549 −0.403 −0.447 −0.494 −0.574 −0.393 1.165

−2.542 −2.714 −3.062 −2.936 −2.897 −2.942 −3.012 −3.018 −3.062 −3.046 −2.952 −3.014 −2.975 −2.883 −2.880 −2.905 −2.960 −2.559 −2.315

0.973 0.974 0.984 0.994 0.995 0.998 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.000 1.001 1.003 1.004 1.004 1.019

−1.719 −2.242 −2.383 −1.338 −1.611 −0.963 −0.410 −0.625 −0.348 −0.324 0.066 0.219 0.197 0.197 0.210 0.981 0.519 0.341 1.170

−2.498 −2.792 −2.946 −2.924 −2.793 −2.747 −2.722 −2.733 −2.687 −2.755 −2.755 −2.760 −2.715 −2.796 −2.802 −2.861 −2.846 −2.577 −2.793

Notes: Table 2 presenting points estimated and t-values at 5%. In the table tvalue < C.V, so we reject the null hypothesis of α(τ) = 1. While Quantiles bold values > C.V which means α(τ) ≠ 1 on different quantiles.

3.2. Quantile Cointegration Test Further, we employs the Xiao’s (2009) Quantile Cointegration Test to ascertain whether cointegration exists or not among proposed variables. Furthermore, we examined cointegration among the proposed variables by applying an alike spaced grid of 19 quantiles (0.05–0.95). We used two lags and leads of ( ΔZt , ΔZt2 ) for Eq. (3) of quantile cointegration model. The estimates of the stability test for Quantile Cointegration Model depicted in Eq. (3) is presented in Table 3. We observe the presence of statistically asymmetric long-run relationship (cointegration) between the quantiles of the global crude oil prices and DJ Islamic Stock Index (Table 3). The paper also inspects the association among global crude oil prices and DJ Islamic Stock Index with MODWT-based covariance analysis indicating the covariance between the variables in a specific period. The plot of wavelet covariance between Oil and ISL is presented in Fig. 2. The result of the wavelet covariance indicates the evidence of positive covariance between Oil and ISL in a very short and short period, but not in medium, long and very long run. Fig. 2 presents the wavelet correlation. Fig. 2 suggests the presence of positive and strong correlation among oil prices and Islamic Stock Index, in short, medium, long and very long run. 3.3. Quantile on quantile estimates Fig. 3 plot the estimates of QQ Analysis concerning DJ-Islamic Stock Table 3 Quantile cointegration test results. Model

Coeff.

Supτ | Vn(τ) |

CV1

CV5

CV10

ISLt vs. OILt

β γ

174.247 32.844

37.129 7.351

30.218 4.527

25.031 3.875

Note: Table 3 having quantile cointegration outcomes and ISL and OIL both variables are in logarithm form. The parameters (β and γ) stability of Eq (3) is also being checked. The critical values are produced by using 1000 Monte Carlo simulations and CV1, CV5, and CV10 are the critical values on various levels of significance. Furthermore, we examined cointegration (t-statistic) among the purposed variables by applying an alike spaced grid of 19 quantiles (0.05–0.95). 297

Resources Policy 62 (2019) 292–304

S. Mishra, et al.

Fig. 2. Wavelet Covariance and Correlation Analysis between OIL and ISL.

prices Oil.d3 on Islamic stock market, we can observe the positive effect becoming extremely weak, with zero and negative effect becoming prominent across the quantiles of both dependent and independent variables. The similar scenario is observed when we analyze the effect of decomposed series Oil.d4, Oil.d5, Oil.d6, Oil.d7, Oil.d8, Oil.d9 and the stable component Oil.s9 on Islamic Stock Market. The positive effect becomes extremely weak with negative effect becoming extremely significant. The positive effect although weak is observed only at the higher quantiles. The effect of decomposed series Oil.d6 on the Islamic stock market is negative as observed from both QR and QQ estimates. Fig. 4 reveals the trend of QQ lines being the same as that of QR Lines. However, for the effect of the original time series of oil prices and decomposed time series, i.e., Oil.d1, Oil.d2, Oil.d3, and Oil.d4, the value estimates of quantile on quantile regression and quantile regression model are somewhat different. For the effect of Oil.d5, Oil.d6, Oil.d7, Oil.d8, Oil.d9, and Oil.s9 on Islamic stock market the estimated values of QQ and QR coincide. The analysis point towards the fact that although the fluctuations in oil prices may have the positive effect of Islamic stock index in the short run, with the advent of stability in the oil prices, the detrimental influence of an increase in oil prices on Islamic stock market becomes evident. The outcome of the study could attribute to the fact that the Islamic stock market index comprises of the majority of oil exporters and increase in global oil prices have a short-term positive effect on stock returns owing to increase in revenue. The outcome of the study involving the decomposed series of oil price and Islamic Stock Index is in consonance with the results obtained in the works of Bouri et al. (2017), Badeeb and Lean (2018) and Narayan et al. (2019) who observed heterogeneous or asymmetric or nonlinear relationship between the oil price and Islamic stock index. However if we examine the effect of Oil Price on the Islamic Stock Index, then we observe positive effect of oil price on the stock index, the outcome which is similar to that observed in the work of Arshad (2017) where the researcher found parallel movement between the global crude oil price and Islamic stock market. In case of other conventional stock markets the relationship between the oil price and stock market depended upon the nations' oil dependency (Ramos and Veiga, 2011). The results obtained were also similar to that of Reboredo and Ugolini (2016) work on stock markets of BRICS nations where the researchers also found the presence of asymmetries in oil price spillover to conventional stock returns of BRICS economy.

influence of Oil.d4 is negative. When we analyze the impact of further decomposed series of Oil Prices, i.e., Oil.d5 on Islamic stock market, we find the complete absence of positive effect. The negative effect of Oil.d5 becomes predominant across all the quantiles of Oil.d5 and Islamic Stock Market. Similarly, the effect of Oil.d6 on Islamic Stock Market is negative and predominant across all the quantiles of both dependent and independent variables. On analyzing the impact of Oil.d7 on Islamic stock market, in the region covering the lower to middle quantile of Oil.d7 (0.4–0.5) and lower to higher quantile (0.1–0.9) of ISL, there is a positive influence of decomposed series of oil prices, i.e., Oil.d7 on Islamic Stock Market. In this region, although the effect exists but is very weak. In other regions covering the middle to higher quantile of Oil.d7 (0.51–0.95) and lower to higher quantiles (0.1–0.95), the negative effect is predominant and strong. When we analyze the impact of Oil.d8 and Oil.d9 we find that in the region linking the higher quantiles of oil prices Oil.d8 and Oil.d9 (0.65–0.95) and all the quantiles of Islamic stock market we observe the positive effect of decomposed series of oil prices on Islamic stock market. In other regions comprising the lower and middle quantiles of oil prices, i.e., Oil.d8 and Oil.d9, linking to all the quantiles of the Islamic stock market, the negative effect is strong and predominant. The similar scenario is also observed when we analyze the effect of the most stable component of the decomposed time series of oil prices, i.e., Oil.s9 on Islamic stock market. In the region adjoining the higher quantiles of Oil.s9 (0.7–0.9) and lower to higher quantiles of Islamic stock market the positive effect of Oil prices is observed. In all other regions, we find the negative influence of oil prices on Islamic stock market being predominant. From the analysis, we can infer that in the original series the changes in oil prices have a significant positive effect on the Islamic stock market. The decomposition of the series removes the noise or fluctuation from it and make it stable. On decomposing the original series of oil prices, we find that the positive effect starts decreasing and zero effect starts becoming predominant. With the increase in stability of the series, the positive effect of the oil prices becomes negligible, and the negative effect becomes predominant. The paper checks the validity of the Quantile on Quantile approach by comparing the τ averaged QQ parameters with the estimated quantile regression parameters QR. The plots of quantile regression and the averaged QR parameter estimates of the slope coefficient measures the impact of global crude oil prices and its decomposed series on DJ Islamic Stock Index is illustrated in Fig. 4. When we compare the QR and QQ estimates of the effect of global crude oil prices on Islamic stock index we observe the positive influence across all the quantiles. On decomposing the original series of global crude oil prices, the positive influence of oil prices starts diminishing, and zero effect start becoming prominent. The same is also observed for QR estimates. From the analysis of the effect of decomposed time series of oil

4. Conclusion and policy recommendation The present study endeavors to analyze the influence of global crude oil prices on the DJ-Global Islamic Stock Index by employing the daily data from January 01, 1996 to April 13, 2018. The unique econometric approach adopted in the study is that the paper decomposes the original 298

Resources Policy 62 (2019) 292–304

S. Mishra, et al.

Fig. 3. Quantile-on-Quantile estimates of slope coefficient.

299

Resources Policy 62 (2019) 292–304

S. Mishra, et al.

Fig. 3. (continued)

300

Resources Policy 62 (2019) 292–304

S. Mishra, et al.

Fig. 4. Comparison between QQ and QR estimates.

301

Resources Policy 62 (2019) 292–304

S. Mishra, et al.

Fig. 4. (continued)

302

Resources Policy 62 (2019) 292–304

S. Mishra, et al.

Quantile ARDL) that support to comprehend the nexus over a bigger number of variables as discussed earlier. The present study could be extended to analyze the impact of global crude oil prices on the stock indices before, and post-financial crisis happened during the year 2008. The sectoral analysis of the effect of global crude oil prices could also be carried out which would be instrumental in devising effective hedging strategies. Further, the future study could be oriented towards differentiating the impact of global crude oil prices on stock indices of Islamic oil exporting countries and Islamic oil importing countries.

time series of global crude oil prices to remove the noise, i.e., the fluctuation component from the time series data and analyses its influence on Islamic stock market. Apart from that, the analysis is done under the Quantile on Quantile Regression (QQ) framework recently introduced and developed by Sim and Zhou (2015). The merit of QQ framework over the conventional techniques like OLS and Quantile Regression is that it estimates how the different quantiles of global crude oil prices and its decomposed series influence the different quantiles of the DJ-Islamic stock index. The QQ approach precisely describes the dependence structure between the endogenous and the regressor variable. The outcome of the study indicates the heterogeneity in the influence of global crude oil prices on the DJ-Islamic stock market index. From the study presented under QQ framework, we observe that when we analyze the influence of original time series data of global crude oil prices on Islamic Stock Index, there exists a positive influence of oil prices across all the quantiles. When we start decomposing the original series of global crude oil prices to remove the noise, i.e., fluctuations from the data and analyze its influence, the positive influence starts decreasing, and the zero effect starts becoming prominent. Apart from zero effect, the negative effect also starts coming into existence. With the increase in the stability of time series data of crude oil prices achieved due to its decomposition, its negative effect on Islamic Stock Index becomes stronger and positive effect becomes weaker and insignificant. The crude oil is the major source of energy being used in the manufacturing process across the world and the variation in its prices significantly affect the cost of production and the industries’ profitability (Redin et al., 2018; Troster et al., 2018; Raza et al., 2018). Since the major portion of the GDP of these Islamic nations comprises of oil revenue, the increase in its prices strengthens the GDP thereby causing an increase in aggregate demand level (Badeeb & Lean, 2018; Aloui et al., 2018). However, this increased demand is not appropriately responded by domestic producers owing to inadequate infrastructure thereby leading to excess demand and rising prices causing an increase in inflation level. Thus, the increase in oil prices and increased inflation level causes increased cost, reduced profitability resulting in significant change in the discounted cash flow. Since according to economic theory, the discounted cash flow determines the prices of any asset, the rise in oil prices has a curtailing effect on the stock market. From the perspective of investments, the investors need to take into account oil risk and return for formulating the performance expectations to allocate assets and design the investment portfolio. The investor should pick the stocks which exhibit high positive sensitivity to changes in oil prices. Moreover, the findings of the study indicate the asymmetric nature of the influence of global crude oil price on DJ-Islamic Stock Index. The findings are crucial as an understanding of the interplay between price and volatility behavior can have an important bearing on derivative valuation for hedging. The outcome of the study could help in better forecasting of the ongoing trend in the Islamic stock markets and understanding of possible investment risks. As an investor, the findings would help in evaluating the different portfolio diversification opportunities comprising of different investment horizons. Since the findings are more related to negative or positive oil price shocks, the influence of oil price on stock returns become evident when we observe either better or worse performance in stock markets. This may be attributed to optimistic or pessimistic investors’ expectations. This leads to irrational behavior of the investors which significantly characterizes the stock market and as a policymaker we should avoid uncertain information about oil price which may bring turbulence in the market. The fundamental constraints of the current investigation are the application of bivariate quantile on quantile regression approach. In the present approach we are not having the scope of incorporating control variables which may play a crucial role in the influence of global crude oil price on stock index. We may broaden the econometric model by using, for instance, different multivariate quantile approach (i.e.,

Appendix A. Supplementary data Supplementary data to this article can be found online at https:// doi.org/10.1016/j.resourpol.2019.04.005. References Abdullah, A.M., Saiti, B., Masih, M., 2016. The impact of crude oil price on Islamic stock indices of South East Asian countries: Evidence from MGARCH-DCC and wavelet approaches. Borsa Istanbul Rev. 16 (4), 219–232. Aloui, C., Hkiri, B., Hammoudeh, S., Shahbaz, M., 2018. A multiple and partial wavelet analysis of the oil price, inflation, exchange rate, and economic growth nexus in Saudi Arabia. Emerg. Mark. Finance Trade 54 (4), 935–956. Aloui, C., Nguyen, D.K., Njeh, H., 2012. Assessing the impacts of oil prices fluctuations on stock returns in emerging markets. Econ. Modell. 29 (6), 2686–2695. Apergis, N., Miller, S.M., 2009. Do structural oil-market shocks affect stock prices? Energy Econ. 31 (4), 569–575. Arshad, S., 2017. Analysing the relationship between oil prices and Islamic stock markets. Econ. Pap.: J. Appl. Econom. Pol. 36 (4), 429–443. Badeeb, R.A., Lean, H.H., 2018. Asymmetric impact of oil price on Islamic sectoral stocks. Energy Econ. 71, 128–139. Bahloul, S., Mroua, M., Naifar, N., 2017. The impact of macroeconomic and conventional stock market variables on Islamic index returns under regime switching. Borsa Istanbul Rev. 17 (1), 62–74. Balcilar, M., Bekiros, S., Gupta, R., 2017. The role of news-based uncertainty indices in predicting oil markets: a hybrid nonparametric quantile causality method. Empir. Econ. 53 (3), 879–889. Basher, S.A., Sadorsky, P., 2006. Oil prices risk and emerging stock markets. Glob Financ. J. 17 (2), 224–251. Bekiros, S., Gupta, R., Majumdar, A., 2016. Incorporating economic policy uncertainty in US equity premium models: a nonlinear predictability analysis. Financ. Res. Lett. 18, 291–296. Bouri, E., Jain, A., Biswal, P.C., Roubaud, D., 2017. Cointegration and nonlinear causality amongst gold, oil, and the Indian stock market: evidence from implied volatility indices. Resour. Pol. 52, 201–206. Chang, K.L., Yu, S.T., 2013. Does crude oil price play an important role in explaining stock return behavior? Energy Econ. 39, 159–168. Ciner, C., 2001. Energy shocks and financial markets: nonlinear linkages. Stud. Nonlinear Dynam. Econom. 5 (3), 203–212. Cornish, C.R., Bretherton, C.S., Percival, D.B., 2006. Maximal overlap wavelet statistical analysis with application to atmospheric turbulence. Boundary-Layer Meteorol. 119 (2), 339–374. Cunado, J., de Gracia, F.P., 2014. Oil prices shock and stock market returns: evidence for some European countries. Energy Econ. 42, 365–377. Diaz, E.M., Molero, J.C., de Gracia, F.P., 2016. Oil price volatility and stock returns in the G7 economies. Energy Econ. 54, 417–430. Elyasiani, E., Mansur, I., Odusami, B., 2011. Oil price shocks and industry stock returns. Energy Econ. 33 (5), 966–974. Engle, R.F., Granger, C.W.J., 1987. Cointegration and error correction: representation, estimation and testing. Econometrica 55 (2), 251–276. Filis, G., Degiannakis, S., Floros, C., 2011. Dynamic correlation between stock market and oil prices: the case of oil-importing and oil-exporting countries. Int. Rev. Financ. Anal. 20 (3), 152–164. Ftiti, Z., Hadhri, S., 2019. Can economic policy uncertainty, oil prices, and investor sentiment predict Islamic stock returns? A multi-scale perspective. Pac. Basin Finance J. 53, 40–55. Galvao, A.F., 2009. Unit root quantile autoregression testing using covariates. J. Econom. 152 (2), 165–178. Gencay, R., Selcuk, Whitcher, B., 2002. An Introduction to Wavelets and Other Filtering Methods in Finance and Economics. Academic Press, San Diego, CA. Ghouri, S.S., 2006. Assessment of the relationship between oil prices and US oil stocks. Energy Policy 34 (17), 3327–3333. Gupta, R., Modise, M.P., 2013. Does the source of oil prices shock matter for South African stock returns? A structural VAR approach. Energy Econ. 40, 825–831. Hammoudeh, S., Choi, K., 2007. Characteristics of permanent and transitory returns in oil-sensitive emerging stock markets: the case of GCC countries. J. Int. Financ. Mark. Inst. Money 17 (3), 231–245. Hassan, K., Hoque, A., Gasbarro, D., 2019. Separating BRIC using Islamic stocks and crude oil: dynamic conditional correlation and volatility spillover analysis. Energy Econ.

303

Resources Policy 62 (2019) 292–304

S. Mishra, et al.

Energy 154, 571–580. Raza, N., Shahzad, S.J.H., Tiwari, A.K., Shahbaz, M., 2016. Asymmetric impact of gold, oil prices and their volatilities on stock prices of emerging markets. Resour. Pol. 49, 290–301. Reboredo, J.C., Ugolini, A., 2016. Quantile dependence of oil prices movements and stock returns. Energy Econ. 54, 33–49. Redin, D., Rodriguez, I., Cuñado, J., Perez de Gracia, F., 2018. Oil prices and economic activity: evidence for G-7 economies based on a wavelet approach. Appl. Econ. Lett. 25 (5), 305–308. Sadorsky, P., 1999. Oil price shocks and stock market activity. Energy Econ. 21 (5), 449–469. Sadorsky, P., 2008. Assessing the impact of oil prices on firms of different sizes: its tough being in the middle. Energy Policy 36 (10), 3854–3861. Saikkonen, P., 1991. Asymptotically efficient estimation of cointegration regressions. Econom. Theor. 7 (1), 1–21. Salisu, A.A., Isah, K.O., 2017. Revisiting the oil price and stock market nexus: A nonlinear Panel ARDL approach. Econ. Modell. 66, 258–271. Shahbaz, M., Zakaria, M., Shahzad, S.J.H., Mahalik, M.K., 2018. The energy consumption and economic growth nexus in top ten energy-consuming countries: fresh evidence from using the quantile-on-quantile approach. Energy Econ. 71, 282–301. Sharif, A., Shahbaz, M., Hille, E., 2019a. The Transportation-growth nexus in the USA: fresh insights from pre-post global crisis period. Transport. Res. Pol. Pract. 121, 108–121. Sharif, A., Afshan, S., Qureshi, M.A., 2019b. Idolization and ramification between globalization and ecological footprints: evidence from quantile-on-quantile approach. Environ. Sci. Pollut. Res. 1–21. https://doi.org/10.1007/s11356-019-04351-7. Sim, N., Zhou, H., 2015. Oil prices, US stock return, and the dependence between their quantiles. J. Bank. Financ. 55, 1–8. Tiwari, A.K., Jena, S.K., Mitra, A., Yoon, S.M., 2018. Impact of oil price risk on sectoral equity markets: implications on portfolio management. Energy Econ. 72, 120–134. Troster, V., Shahbaz, M., Uddin, G.S., 2018. Renewable energy, oil prices, and economic activity: a Granger-causality in quantiles analysis. Energy Econ. 70, 440–452. Tsai, C.L., 2015. How do US stock returns respond differently to oil prices shocks precrisis, within the financial crisis, and post-crisis? Energy Econ. 50, 47–62. Wang, Y., Wu, C., Yang, L., 2013. Oil price shocks and stock market activities: Evidence from oil-importing and oil-exporting countries. J. Comp. Econ. 41 (4), 1220–1239. Wei, Y., Guo, X., 2017. Oil price shocks and China's stock market. Energy 140, 185–197. Xiao, Z., 2009. Quantile cointegrating regression. J. Econom. 150 (2), 248–260. Zhang, B., Li, X.M., 2016. Recent hikes in oil-equity market correlations: transitory or permanent? Energy Econ. 53, 305–315. Zhu, H.M., Li, S.F., Yu, K., 2011. Crude oil shocks and stock markets: a panel threshold cointegration approach. Energy Econ. 33 (5), 987–994.

80, 950–969. Huang, S., An, H., Gao, X., Huang, X., 2016. Time-frequency featured co-movement between the stock and prices of crude oil and gold. Phys. Stat. Mech. Appl. 444, 985–995. Hussin, M., Yahya, M., Muhammad, F., Noordin, K., Marwan, N.F., Abdul Razak, A., 2012. The impact of oil price shocks on the Islamic financial market in Malaysia. Labuan eJ. Muamalat Soc. 6, 1–13. Koenker, R., Bassett, G., 1978. Regression quantiles. Econometrica 46 (1), 33–50. Koenker, R., Xiao, Z., 2004. Unit root quantile autoregression inference. J. Am. Stat. Assoc. 99 (467), 775–787. Li, S.F., Zhu, H.M., Yu, K., 2012. Oil prices and the stock market in China: a sector analysis using panel cointegration with multiple breaks. Energy Econ. 34 (6), 1951–1958. Ma, L., Koenker, R., 2006. Quantile regression methods for recursive structural equation models. J. Econom. 134 (2), 471–506. Miller, J.I., Ratti, R.A., 2009. Crude oil and stock markets: stability, instability, and bubbles. Energy Econ. 31 (4), 559–568. Mollick, A.V., Assefa, T.A., 2013. US stock returns and oil prices: the tale from daily data and the 2008–2009 financial crisis. Energy Econ. 36, 1–18. Nandha, M., Faff, R., 2008. Does oil move equity prices? A global view. Energy Econ. 30 (3), 986–997. Narayan, P.K., Phan, D.H.B., Sharma, S.S., 2019. Does Islamic stock sensitivity to oil prices have economic significance? Pac. Basin Finance J. 53, 497–512. Pal, D., Mitra, S.K., 2017. Time-frequency contained co-movement of crude oil and world food prices: a wavelet-based analysis. Energy Econ. 62, 230–239. Park, J., Ratti, R.A., 2008. Oil price shocks and stock markets in the US and 13 European countries. Energy Econ. 30 (5), 2587–2608. Percival, D.B., 1995. On estimation of the wavelet variance. Biometrika 82 (3), 619–631. Percival, D.B., Mofjeld, H.O., 1997. Analysis of subtidal coastal sea level fluctuations using wavelets. J. Am. Stat. Assoc. 92 (439), 868–880. Percival, D.B., Walden, A.T., 2000. Wavelet Methods for Time Series Analysis. Cambridge University Press, Cambridge, UK. Power, G.J., Turvey, C.G., 2010. Long-range dependence in the volatility of commodity futures prices: wavelet-based evidence. Phys. Stat. Mech. Appl. 389 (1), 79–90. Ramos, S.B., Veiga, H., 2011. Risk factors in the oil and gas industry return: international evidence. Energy Econ. 33 (3), 525–542. Ramsey, J.B., 1999. The contribution of wavelets to the analysis of economic and financial data. Phil. Trans. Roy. Soc. Lond. 357 (1760), 2593–2606. Ramsey, J.B., 2002. Wavelets in economics and finance: past and future. Stud. Non-Linear Dynam. Econom. 6 (3), 1–29. Raza, S.A., Shahbaz, M., Amir-ud-Din, R., Sbia, R., Shah, N., 2018. Testing for waveletbased time-frequency relationship between oil prices and US economic activity.

304