Do Global Financial,Oil and Gold Volatility Shocks Affect the GCC Stock Markets?

Abstract

This study employed the non-structural VAR econometrics approach to examine the impact of Global Oil (OVX), Financial (VIX), and Gold (GVZ) volatility indices on GCC stock markets using a daily data set spanning from January 5, 2009 to August 16, 2018. From the VAR result obtained, disequilibrium in the global financial volatility (VIX) was able to significantly transmit negative shock to Bahrain and Kuwait stock markets and positive shock on GVZ. While the global Gold volatility was capable of transmitting fairly positive shock to the UAE and VIX market. The OLS also revealed more to the result obtained from VAR as it shows that OVX and VIX can have impact on the GCC stock markets. The causality test revealed that there is a unidirectional causality running from Qatar and UAE to OVX; none of the variables was able to granger cause VIX, while unidirectional causality exist from VIX and UAE to GVZ and VIX and Qatar to Bahrain. VIX and Qatar can granger cause Kuwait stock market, and only Saudi Arabia and Oman have bidirectional causality. Unidirectional causality exists from Saudi Arabia to Qatar, and Qatar is the only stock market capable of causing UAE unidirectionally. Hence, the study concludes that VIX and GVZ are capable of transmitting shocks to three of the six GCC stock markets—(Bahrain, Kuwait and The UAE) negatively (Bahrain and Kuwait) and positively (The UAE). And on this note, the study recommends that appropriate financial and gold transaction policies should be institutionalized so as to mitigate the transmission of shocks into the markets. Also, financial and gold experts who regulate the stock and gold markets especially in Bahrain and Kuwait should watch for any abnormality changes in the volatility movement of the financial and gold markets.

Keywords

VAR GCC stock market international investment

Introduction

The study of markets volatilities is a unique area which has attracted much interest and gained more platforms in the global market scenes. This is because studies on volatilities are good for the proper functioning of the global economy since good number of social economic factors can trigger it. Furthermore, its destructive power can halt most social and economic global activities. Global volatilities have their source(s), namely from the global oil sphere (OVX), from the financial circle (VIX) or even from the gold selling markets (GVZ). All these sources of volatilities that can affect the global economy positively or negatively deepening on the type of market it comes from. Furthermore, the need to study volatilities becomes imperative because there has been a large influx of investors into any of these three markets on a daily basis. Hence, investors need to measure and know the level of confidence on the structure and mode of returns of any of the markets before they invest their resources into it. For instance, in the United States as well as other global financial markets, financial volatility is measured by CBOE volatility index or VIX. This is because the VIX functions measure the average market confidence and uses other pricing options on stock index to measure the certainty of financial market for a 30 days future. Therefore, the more the value of VIX, the more the level of risk in the market and vice versa (Connolly, Stivers, & Sun, 2007).

In the other market, crude oil is the most volatile and most traded commodity in the world, as its volatile shocks do spread affecting the other two markets—stock and gold markets. These among other reasons make volatility very important in determining the activities of other related markets globally. With regards to the GCC economies, oil is the mainstay of all the six regional countries; it accounted to more than 40 percent of the regional economies’ gross domestic product. Though the most vulnerable economies are the Bahraini and Oman economies, as they are most susceptible to external shocks with regards to financial and gold volatility (Arouri, Lahiani, & Nguyen, 2011; Ludvigson & Ng, 2009; Raza, Shahzad, Tiwarib, & Shahbaz, 2016).

Unlike the two volatilities, the gold market volatility (GVZ) has had less attention with respect to pricing and returns behaviors. This is because of its random uniform and global standardized means of measuring its value, as it possesses dual exchange characteristics as both commodity and financial assets leading to its stochastic data problem (Akgiray, Booth, John, & Chowdhury, 1991; Ball, Torous, & Tschogel, 1982). Also, as identified by Raza, Shahzad, Tiwarib, and Shahbaz (2016), gold volatility can spring from bad news and other socioeconomic events within the economy or even exogenously. Though, despite the problems associated with gold, investors prefer to stay away from the risks associated with financial and oil markets by investing more in precious metals like gold because of its non-rapid volatile characteristics and good returns compare to stocks and oil. For instance, Daskalaki and Skiadopoulos (2011) reported that due to increasing instability and uncertainty in stock markets, investors prefer to invest in the metal markets taking them to be a safer haven. Also, Baur and Lucey (2010) found that gold serves as a better option against stocks in the UK, Germany and the US, particularly following severe adverse shocks which is an inherent feature of oil and stock markets which can easily create more vacuum for global uncertainties. The great depression of the 1930s, the 2008 global financial crisis and the 2006 Qatar oil crisis are evidences to this fact (Angellini, Bacchiocchi, Caggiano, & Fanelli, 2017; Arouri et al., 2011; Ludvigson & Ng, 2009; Ludvigson, Ma, & Ng, 2017).

There are two ways to measure global volatility. The first is the traditional approach, which involves the use of mean, standard deviation and variance. The second approach is the simplified approach which uses diagrams like histograms in explaining the concept. Most investors prefer the latter, which captures the problem of skewness and kurtosis and does not require investment performance to be normally distributed; it also provides investors with vital information which can help mitigate unexpected volatility; and lastly, it measures the degree of returns experienced (Kuang, 2018).

Now, to the regional GCC countries, these volatilities call for great concern due to the influx of investors within and outside the region. Therefore, having better understanding on the risk attached to each of these markets will help boost the degree of certainty when investing in any of these regional markets. This and other objectives form the motivation for this study—to investigate the impacts of these volatilities on the GCC stock markets using daily data from 5/1/2009 to 16/8/2018, to determine the degree of these three variables (Oil, Financial and Gold) transmission shocks on the GCC stock markets in a period of non-major global financial crisis employing the non-structural vector auto-regression (VAR) and to test for any causal link among the variables using the block exogeneity Granger Causality Wald test. To achieve this, the study is sectionalized into five sections—the first part takes account of the general introduction, followed by literature review, then methodology is explained in the third section, after that comes the estimation and discussion of results in the fourth section and, lastly, the fifth section states the findings and conclusion and proffer suitable policy recommendations.

Literature Review

Oil Volatility is defined as follows in Raza et al. (2016): Oil volatility is the unstable trend in the prices and volume of supply and demand that is characterized within the global oil market. VIX is the instrument that is used the measure the performance of the US stock market. It’s a number derived from the prices of options premium in the S & P 500 index (Index comprising of large capital Stocks). The index is sometimes referred to as the ‘Fear Index’ because sometimes its construes fear and sometimes reflect complacency. The index is mostly used by the Chicago Board Options Exchange (CBOE), one of the world’s largest exchange holding companies (Adam, Marcet, & Pablo, 2015; Williams, 2013). Gold Volatility as discussed by Tang (2017) is the unstable movement associated with the precious metal called ‘Gold’; this volatility is not as rapid as the other two (oil and Stocks) variables (Raza et al., 2016).

On the other hand, stock exchange market is a place where the sellers and buyers meet each other to trade stocks and bonds. Investors come to stock market and buy, hold the stocks for some time and later sell them at the higher prices (Tang, 2017). All the three markets volatilities are mostly related with each other in the business cycle (Caballero, Emmanuel, & Gourinchas, 2008).

GCC as explained by Maktoun in 2014 is an integrated group of six Arab countries located in the Persian Gulf (except Iraq) of the Middle East. The cooperation was formed with the main aim of achieving unity and economic progress, strengthen relations among member countries and promote military cooperation. The cooperation was formed in Riyadh in May 1981; the cooperation is regarded as the most renowned and strongest in the Middle East. The cooperation was able to achieve this feat because of the countries’ close regional location and similar political and cultural identities. In addition, these countries have a common trade product—crude oil production, stocks and even gold (Article 4, GCC Charter).

Empirical

From the survey of previous literature which we have reviewed, we discovered that some scholars considered to investigate the relationship between oil and stock market, some considered oil and the US dollar while others considered gold and oil prices. For oil and stock markets, studies like Alkhatlan (2010), Gomes and Chaibi (2014), and Masih, De Mello, and Peters (2011) all investigated the relationship between these two variables. And their findings that there is a significant interaction (positively and negatively) depend on the context of which the studies were carried out. For gold and oil prices, Narayan, Narayan, and Zheng (2010) found the evidence of the long-run relationship between these two variables. His result showed that when oil price increases, it creates pressures of inflation after that investors prefer to invest in some financial assets such as gold to protect their investment portfolios. Similarly, Reboredo (2013) concerned the hedging role of gold against oil price changes. He showed that there is tail independence between the oil and gold markets, meaning that gold can work as an effective hedge against the fluctuation of oil price. Other studies that investigated on this subject matter were Akgül, Bildirici, and Özdemir (2015), Basher and Sadorsky (2006), Chkili (2016), Hassan and Shabi (2015), Kumar (2017), Le and Chang (2012), and Mollick and Assefa (2012).

Identified Gap

The following are the gaps that this study will help address with regards to previous literature:

Data: Like Sum (2013) where monthly and annual data were used, this study uses daily data from 2009 to 2018 so as to capture for a quick transfer of shocks. The reason for this period is to examine how these selected variables OVX, VIX and GVZ impact on the six GCC stock markets. Also, the period is a stable period void of global financial crises and less of global oil price volatilities which can easily be transpired to the GCC region.

On Variables: This study combines three global volatility variables to determine their impact on the GCC stock market. Furthermore, the study is a global studies, which involve not only one global variable in relation to GCC region, but consist of three major global variables (OVX, VIX, and GVZ) to determine their influence in the region. The justification for considering these three global variables is that the GCC regional countries engage in three services (Oil, forex trading and gold) with oil as its highest outputs. And only Saudi Arabia producing about 12.7 percent of daily global oil production in 2017 as compared to 15.3 percent produced by the US (OPEC annual bulletin 2017). Furthermore, the study does not concentrate on one country but group of countries.

Scope: The scope of the study is adequate and sufficient enough to give a clear account of the influence of the three global variables on GCC stock markets taking into account periods of less global and regional financial crisis and oil price disequilibrium.

Methodology

This section will help in achieving the objectives of this study by unraveling the effective and most suitable technique that fits this subject matter. In doing this, this section shall capture the study area, the source of data, empirical or model specification, the estimation technique(s) and the anticipated results.

Study Area

This study is a global study which incorporates the global oil and gold volatility indices (OVX and GVZ, respectively). But in specific, the study area with regards to this work is majorly the GCC region. The VIX which is a stock market volatility index of the US is used here, hence United States is also a study area. The GCC region occupies six Middle East countries (Bahrain, Kuwait, Oman, Qatar, Saudi Arabia, and the United Arab Emirates) with 1032,093-square-mile (2,673,110 km²) of land area (Charter of the Gulf Cooperation Council, 2017). While the United States is a Federal Republic made up of 50 states with Washington DC as its national capital. The states occupied a total area of 9,525,067³ km² and has an estimated population of about 308,745,538 people (2017 est.). The region uses dollar as her official currency (Britanica.com).

Data

OVX, VIX, and GVZ are used to measure the GCC stock market strength and resistance with respective to three different global volatilities and finally The US dollar Index is used as a control variable factoring in its enormous influence on the global economy. All data are in daily ranged from 5/1/2009 to 16/8/2018. All the data were sourced from ‘Bloomberg’ and were compared to the official website of each variable. The rationale for the use of daily data for all the variables is because weekly or monthly data are weakened when dealing with holidays and their lead/lag relationships; furthermore, daily data reveal clear and quick reaction to level shifts and changes in trends because they are modeled daily without waiting till the end of the week or month (see Reilly, 2018); also, the choice for the period was considered by the availability of data which gives the researcher the mandate to investigate the strength of the GCC stock market with respect to global volatilities after the 2006 Qatar stock market nose-dived and with respect to the 2008 global financial crisis. Hence, 2009–2018 is suitable for this study (Table 1).

Estimation Technique and Empirical Model

The technique used in this study is a combined technique of ordinary least squares (OLS) and non-structural VAR. The OLS technique takes into cognizance the three main variables (OVX, VIX and GVZ) as dependent variables with each variables having an OLS equation of its own.

Table 1.

The Sources of Data

S/N	Symbol	Description	Source
1	OVX	Global Oil Volatility Index	http://www.cboe.com/products/vix-index-volatility/volatility-on-etfs/cboe-crude-oil-etf-volatility-index-ovx
2	VIX	Chicago Board Options Exchange Stock Market Volatility	http://www.cboe.com/vix
3	GVZ	Global Gold Volatility Index	http://www.cboe.com/products/vix-index-volatility/volatility-on-etfs/cboe-gold-etf-volatility-index-gvz
4	USD Index	United States Dollar Index	http://www.ub.edu/rfa/uncertainty-index/#Reference
5	BRSI	Bahrain Stock Market Index	Bloomberg Terminal
6	KWTSI	Kuwait Stock Market Index
7	OMNSI	Oman Stock Market Index
8	QTSI	Qatar Stock Market Index
9	SARSI	Saudi Arabia Stock Market Index
10	UAESI	United Arab Emirates Stock Market Index

Source: Author’s compilation from various variable sites.

The equations are as follows:

LNBRSI = (OVX + VIX + GVZ) + u

(1.1)

{LNBRSI}_{t} = {F(β}_{1} (OVX + β_{2} (VIX) + β_{3} (GVZ)) + µ_{t}

(1.2)

LNKWTSI = F (OVX + VIX + GVZ) + u

(2.1)

{LNKWTSI}_{t} = {F(β}_{1} (OVX + β_{2} (VIX) + β_{3} (GVZ)) + µ_{t}

(2.2)

LNOMNSI = F (OVX + VIX + GVZ) + u

(3.1)

{LNOMNSI}_{t} = {F(β}_{1} (OVX + β_{2} (VIX) + β_{3} (GVZ)) + µ_{t}

(3.2)

LNQTSI = F (OVX + VIX + VZ) + u

(4.1)

{LNQTSI}_{t} = {F(β}_{1} (OVX + β_{2} (VIX) + β_{3} (GVZ)) + µ_{t}

(4.2)

LNSARSI = F (OVX + VIX + GVZ) + u

(5.1)

{LNSARSI}_{t} = {F(β}_{1} (OVX + β_{2} (VIX) + β_{3} (GVZ)) + µ_{t}

(5.2)

LNUAESI = F (OVX + VIX + GVZ) + u

(6.1)

{LNUAESI}_{t} = {F(β}_{1} (OVX + β_{2} (VIX) + β_{3} (GVZ)) + µ_{t}

(6.2)

In order to empirically estimate the impact of global volatilities and stock exchange index shocks on GCC stock markets indices, we specify two models, namely the non-structural VAR by Chris Sims (1980) and the Granger Block Exogeneity Wald Test. This is because the VAR model does not require much knowledge about the endogenous forces that influence the variables that can be hypothesized to affect each other inter-temporally. Furthermore, the VAR technique is appropriate to determine the dynamic shocks within the model by endogenizing each variable, consequently allowing for the ability to trace the expected response of current and last time impact on each of the markets. The VAR model is expressed as follows:

Y_{t} = C + A_{1} y_{t - 1} + A_{2} y_{t - 2} + \dots + A_{p} y_{t}_{- p} + e_{t}

(7)

Where the L - periods back observations y_t-1 is called the L-th lag of y, C is a k x 1 vector constants (intercepts) A_i is the time invariant k x 1 matrix and e_t is the k x 1 vector of the residuals (Hamilton, 2004).

Variables

The three Global Volatilities (Oil, Financial and Gold) are symbolically written as OVX, VIX, and GVZ. While the GCC stock markets indices are represented as Bahrain Index = (BRSI), Kuwait Index = (KWTSI), Oman Index (OMNSI), Qatar Index = (QTSI), Saudi Arabia Index = (SARSI) and the UAE index = (UAESI) and the Global exchange rate coded USD Index. All the GCC stock market variables are logged in order to strengthen out exponential growth, reduce heteroscedasticity and to have a robust and efficient result because a log-linear specification gives a better estimates than simple specifications (Feng, Wang, Lu, & Tu, 2012; Kowalski & Tu, 2007). Also, all variables are integrated to order one using the ADF stationarity test.

Stationary Test

The Augmented Dickey–Fuller (ADF) stationary test is employed to avoid spurious estimation result, it is necessary because most time series data may not be stationary at level and using non-stationary time series give a stochastic regression (Granger Newbold, 1974). All the variables are subjected to the ADF test procedure as follows:

Δ y_{t} = a + β t + γ y_{t - 1} + δ_{1} Δ y_{t - 1} + ... + δ_{p - 1} Δ y_{t - p + 1} + ε_{t}

(8)

Where α is a constant, β is the coefficient of the time trend and p is the lag order of the auto-regressive process. The unit root test is then applied under the null hypothesis that $γ = 0$ against the alternative hypothesis that $γ = 0$ using

Model 2

To examine the existence of causal relationship between the ten variables, we employed the Granger Block exogeneity Wald test model, and it is thus expressed as:

Y_{t} = γ_{0} + Σ_{t - 1}^{q} γ_{1} Y_{t - i} + Σ_{j = 1}^{q} δ_{j} X_{t - j} + μ_{1 t}

(9)

X_{t} = a_{0} + Σ_{t - 1}^{q} a_{1} X_{t - i} + Σ_{j = 1}^{q} a_{j} Y_{t - j} + μ_{t}

(10)

It is assumed that disturbances µ_1t and µ_2t are uncorrelated. Thus, there is unidirectional causality from X to Y if α_i = 0 and δ_i ≠ 0. Similarly, there is unidirectional causality from Y to X if δ_i = 0 and α_i ≠ 0. The causality is considered as mutual if δ_i ≠ 0 and α_i ≠ 0. Finally, there is no link between X and Y if δ_i = 0 and α_i = 0. To determine the causal link between global volatilities and GCC Stock Markets’ variables, the above equation is modeled as follows:

\begin{array}{l} O V X_{t} = γ_{0} + Σ_{t - 1}^{q} γ_{1} O V X_{t - i} + Σ_{j = 1}^{q} δ B R S I_{t - j} + μ_{1 t} \\ B R S I_{t} = a_{0} + Σ_{t - 1}^{q} a_{1} B R S I_{t - i} + Σ_{j = 1}^{q} a O V X_{t - j} + μ_{t} \\ O V X_{t} = γ_{0} + Σ_{t - 1}^{q} γ_{1} O V X_{t - i} + Σ_{j = 1}^{q} δ_{j} K W T S I_{t - j} + μ_{1 t} \\ K W T S I_{t} = a_{0} + Σ_{t - 1}^{q} a_{1} K W T S I_{t - i} + Σ_{j = 1}^{q} a_{j} O V X_{t - j} + μ_{t} \\ O V X_{t} = γ_{0} + Σ_{t - 1}^{q} γ_{1} O V X_{t - i} + Σ_{j = 1}^{q} δ_{j} U A E S I_{t - j} + μ_{1 t} \\ U A E S I_{t} = a_{0} + Σ_{t - 1}^{q} a_{1} U A E S I_{t - i} + Σ_{j = 1}^{q} a_{j} O V X_{t - j} + μ_{t} \end{array}

These equations are also repeated in the same manner for the other global and GCC variables.

A Priori Expectation: Based on theories we expect to have the signs as shown in Table 2.

Estimation and Discussion Results

Table 2.

A Priori Expectation

S / N	Monetary Policy Uncertainty	Expected Impacts on All the GCC Countries
1	Oil Volatility Index (OVX)	Negative
2	Global Stock Market Volatility Index (VIX)	Negative
3	Gold Volatility Index (GVZ)	Positive

Source: Author calculation.

Table 3.

Descriptive Statistics

Volatility/ Country	Stock Market	Mean	SD	No. of Obs.
OVX	Global Oil Price Volatility	34.91087	12.29791	1581
VIX	Global Financial Index Volatility	18.03621	7.484002	1581
GVZ	Global Gold Price Volatility	18.34176	5.576031	1581
Bahrain	Bahraini SE Market	1313.838	148.6146	1581
Kuwait	Kuwait SE Market	6600.942	766.2038	1581
Oman	Oman SE Market	5947.764	703.5606	1581
Qatar	Qatar SE Market	9472.546	1936.549	1581
Saudi Arabia	Saudi Arabia SE Market	7301.695	1301.452	1581
United Arab Emirates	UAE SE Market.	3772.524	938.7465	1581

Source: Author’s computation using E-Views 10.

Table 4.

Correlation Matrix

	OVX	VIX	GVZ	Bahrain	Kuwait	Oman	Qatar	SaudiArabia	UAE
OVX	1.0000
VIX	0.64503	1.0000
GVZ	0.547932	0.782340	1.0000
Bahrain	0.187779	0.368488	0.214447	1.0000
Kuwait	-0.408815	0.011258	0.141035	0.553247	1.0000
Oman	-0.307514	-0.165919	0.119570	0.243946	0.442811	1.0000
Qatar	-0.285108	-0.573907	-0.422298	-0.129141	-0.052871	0.434101	1.0000
Saudi Arabia	-0.510600	-0.531044	-0.417901	0.076451	0.297344	0.480134	0.820763	1.0000
UAE	-0.262861	-0.598410	-0.570401	-0.044272	-0.009504	0.0384410	0.788406	0.660550	1.0000

Source: Author’s computation using E-Views 10.

From the correlation Table 4, from OVX column, it’s only Bahrain that is positively related to OVX, VIX, and GVZ. This depict that Bahrain stock market is susceptible to changes caused by OVX, VIX and GVZ. While Kuwait is positively related to VIX and GVZ, and Oman is positively correlated to GVZ.

More factual details are evidenced from the bar graphs in Figure 1. The bar graph shows over the years under review (5/1/2009 to 16/8/2018) that within the six GCC stock markets, the highest stock trading comes from Qatar with 14,976,094.93 (28%) million Volume of QE Index; after Qatar is Saudi Arabia with a total of 11,543,979.84 (21%) million of TASI Index; then Kuwait with total Volume of KWSE Index of 10,436,088.55(19%) million. Markets have traded a total of 54,401,117.38 million worth of stocks in different GCC currencies.

Figure 1.

Source: Author’s computation with Excel 2016, using data from Bloomberg.

Figure 2.

Source: Author’s compilation with E-Views 10, using data from Bloomberg.

While the least trading country is Bahrain with a total volume of 2,077,177.72 (4%) million worth of Fils. More details to this information is represented in the graph in Figure 2 which shows the volume of individual stock market that are traded in all the countries of the GCC.

Again from the histogram graphs in Figure 2, it is easier to depict the variable with less skewed and one with most skewed. From the graphs in Figure 1, the highly skewed markets among the GCC region are the Qatar stock market (−0.456240), followed by the UAE stock market (−0.39150) while the less skewed markets are the Saudi Arabia (0.266771) and Bahrain (0.021458). All the highly skewed markets have small kurtosis while the less skewed have high kurtosis. For the OVX, VIX, GVZ and USD indexes, the highly skewed variable is the USD Index (0.202003) with less Kurtosis of 1.573698, while the less skewed variable is VIX (1.856185) with the highest kurtosis of 6.892281.

VAR Estimation

This study employed the VAR estimation to test for the magnitude of impact and transmitted shocks. This is done so as to achieve one of the objectives of this study. All GCC stock market indices data were logged and differenced while the three global volatilities and US dollar index data were not logged neither differenced. This is because they are integrated of order zero, unlike the GCC stock markets indices which are integrated of order one. The estimation of the variables are ordered in the following manner: OVX, VIX, GVZ, D(LNBRSI), D(LNKWTSI), D(LNOMNSI), D(LNQTRSI), D(LNSARSI), D(LNUAESI), (USD INDEX).

Lag Length Criteria

The lag length criteria test was carried out to determine a suitable number of lags which is used for the VAR estimation and which is capable of capturing the dynamic shocks. In regards to this, the test was carried out with lag 20. This is because it is a daily data with 1,581 data points, which is sufficient enough for 20 lags. The lag length result suggested Hannan–Quinn lag length criteria of First, we chose to consider the Hannan–Quinn Criteria¹

https://support.numxl.com/hc/en-us/articles/215531183-Apendix-E-Hannan-Quinn-Information-criterion-HQC HQC = n .ln. (RSS/n) + 2 .K .ln (ln.n)

because the criteria takes into account the log likelihood and the sum of squares residual (RSS) unlike the AIC and the BSIC which do not. The lag length result is shown in Appendix II (Table A2) in the Appendices’ section.

VAR Result

From the VAR estimation in the Appendix II, it is obvious that all the three volatility variables are integrated of order 0. The VAR result suggests that the OVX is not capable of influencing any shock on any of the GCC stock markets, rather Qatar and UAE can affect it negatively (−4.5814) and positive (2.09288), respectively. While VIX can transmit shock on Bahrain (−2.92812) and Kuwait stock markets negatively (−2.06919); on the other hand, none of the six GCC stock markets can transmit shock to VIX. For GVZ, it can only transmit shock fairly on VIX and UAE. This is due to the relation which Gold has with global stock market and The United Arab Emirates.

For the six GCC markets, only Qatar and UAE can shock OVX negatively and positively, respectively. Among the six GCC stock markets, it is only UAE that is capable to transmit a positive shock to Gold Volatility market (GVZ). For more details of the result, see Appendix III (Table A3) in the Appendices’ section.

The AR Root Test: To determine the stability (stationarity) condition of the VAR system model, the AR root table and circle was carried out. And the result suggests the absolute value of all the modules were less than one which means all the roots lie within the unit circle. Hence, the system is stationary and satisfies the VAR stability condition (see Appendices IV and V [Figure A1]).

The OLS Results

From the OLS result given in Table 5, Bahrain, Kuwait, Oman, and UAE were all affected significantly by OVX, VIX, and GVZ while Qatar and Saudi Arabia were affected by OVX and VIX only. This result reveals the link with which global OVX and VIX can affect the GCC stock markets. Also, it shows how related the OVX and VIX are to each other globally. Gold on its own doesn’t have close relation with VIX compared to OVX. The OLS table for this analysis is as follows (Table 5).

Causality Test

The VAR Granger Causality/Block exogeneity Wald test revealed that there is a unidirectional causality running from Qatar and UAE to OVX; none of the variables was able to granger cause VIX, while unidirectional causality exist from VIX and UAE to GVZ and VIX and Qatar to Bahrain. The figurative details of this analysis is shown on Table 6.

Table 5.

OLS Regression Result

Country/Variables	Coefficient	SE	t	Sig	R²
Bahrain
OVX	-0.000805	0.000285	-2.824159	0.0048	13.805
VIX	0.008394	0.000629	13.35018	0.0000
GVZ	-0.009932	0.009932	-5.330566	0.0000
Kuwait
OVX	-0.007282	0.000252	-28.86930	0.0000	36.9505
VIX	0.003375	0.000557	6.062701	0.0000
GVZ	8.836747	0.000675	11.68879	0.0000
Oman
OVX	-0.003547	0.000279	-12.70357	0.0000	27.0600
VIX	-0.007528	0.000616	-12.21648	0.0000
GVZ	0.015013	0.000748	20.08243	0.0000
Qatar
OVX	0.002538	0.000437	5.803860	0.0000	42.0468
VIX	-0.021627	0.000965	-22.40773	0.0000
GVZ	0.001489	0.001171	1.272078	0.2035
Saudi Arabia
OVX	-0.004173	0.000366	-11.41176	0.0000	39.4518
VIX	-0.009334	0.000807	-11.56465	0.0000
GVZ	0.000434	0.000979	0.443593	0.6574
UAE
OVX	0.005556	0.000543	10.22633	0.0000	43.6333
VIX	-0.020403	0.001199	-17.01685	0.0000
GVZ	-0.012705	0.001454	-8.735400	0.0000

Source: Author’s computation.

Table 6.

The Var Granger Causality/Block Exogeneity Wald Tests

Equation	Excluded	X²	df	P > X²
OVX
	VIX	1.654222	1	0.1984
	GVZ	1.439128	1	0.2303
	D(LNBRSI)	0.140757	1	0.7075
	D(LNKWTSI)	1.154608	1	0.2826
	D(LNOMNSI)	3.409268	1	0.0648
	D(LNQTSI)	20.98929	1	0.0000
	D(LNSARSI)	0.681747	1	0.4090
	D(LNUAESI)	4.380146	1	0.0364
	ALL	27.71027	8	0.0005
VIX
	OVX	0.956984	1	0.3279
	GVZ	3.510004	1	0.0610
	D(LNBRSI)	0.050671	1	0.8219
	D(LNKWTSI)	0.088314	1	0.7663
	D(LNOMNSI)	0.256142	1	0.6128
	D(LNQTSI)	0.514621	1	0.4731
	D(LNSARSI)	0.632886	1	0.4263
	D(LNUAESI)	1.889201	1	0.1693
	ALL	8.100542	8	0.4237
GVZ
	OVX	0.061933	1	0.8035
	VIX	11.93859	1	0.0005
	D(LNBRSI)	0.086136	1	0.7691
	D(LNKWTSI)	0.884746	1	0.3469
	D(LNOMNSI)	0.032339	1	0.8573
	D(LNQTSI)	0.033131	1	0.8556
	D(LNSARSI)	0.375694	1	0.5399
	D(LNUAESI)	3.822425	1	0.0506
	ALL	18.49792	8	0.0178
D(LNBRSI)
	OVX	2.283589	1	0.1307
	VIX	8.573892	1	0.0034
	GVZ	1.545894	1	0.2137
	D(LNKWTSI)	2.720346	1	0.0991
	D(LNOMNSI)	0.683551	1	0.4084
	D(LNQTSI)	4.200757	1	0.0404
	D(LNSARSI)	0.007217	1	0.9323
	D(LNUAESI)	0.039360	1	0.8427
	ALL	48.85256	8	0.0000
D(LNKWTSI)
	OVX	0.000970	1	0.9752
	VIX	4.281543	1	0.0385
	GVZ	2.180222	1	0.1398
	D(LNBRSI)	0.223863	1	0.6361
	D(LNOMNSI)	0.070635	1	0.7904
	D(LNQTSI)	8.539386	1	0.0035
	D(LNSARSI)	2.794607	1	0.0946
	D(LNUAESI)	2.509892	1	0.1131
	ALL	21.40760	8	0.0061
D(LNOMNSI)
	OVX	0.174032	1	0.6766
	VIX	1.060809	1	0.3030
	GVZ	1.731637	1	0.1882
	D(LNBRSI)	0.331774	1	0.5646
	D(LNKWTSI)	0.158318	1	0.6907
	D(LNQTSI)	1.286755	1	0.2566
	D(LNSARSI)	9.212295	1	0.0024
	D(LNUAESI)	0.153406	1	0.6953
	ALL	16.56189	8	0.0350
D(LNQTSI)
	OVX	1.873421	1	0.1711
	VIX	0.000149	1	0.9903
	GVZ	0.579251	1	0.4466
	D(LNBRSI)	0.055308	1	0.8141
	D(LNKWTSI)	0.613279	1	0.4336
	D(LNOMNSI)	0.003180	1	0.9550
	D(LNSARSI)	28.23610	1	0.0000
	D(LNUAESI)	1.956300	1	0.1619
	ALL	35.63566	8	0.0000
D(LNSARSI)
	OVX	0.543369	1	0.4610
	VIX	0.118545	1	0.7306
	GVZ	1.186151	1	0.2761
	D(LNBRSI)	0.907703	1	0.3407
	D(LNKWTSI)	0.083514	1	0.7726
	D(LNOMNSI)	4.659536	1	0.0309
	D(LNQTSI)	0.139678	1	0.7086
	D(LNUAESI)	0.403825	1	0.5251
	ALL	8.854477	8	0.3900
D(LNUAESI)
	OVX	0.274097	1	0.6006
	VIX	2.343512	1	0.1258
	GVZ	3.084640	1	0.0790
	D(LNBRSI)	0.093509	1	0.7598
	D(LNKWTSI)	0.077376	1	0.7809
	D(LNOMNSI)	0.253103	1	0.6149
	D(LNQTSI)	8.096441	1	0.0044
	D(LNSARSI)	0.480020	1	0.4884
	ALL	16.03934	8	0.0418

Source: Author’s computation using E-Views 10.

VIX and Qatar can granger cause Kuwait stock market, and only Saudi Arabia and Oman have bidirectional causality. Unidirectional causality exist from Saudi Arabia to Qatar, and Qatar is the only stock market capable of causing UAE unidirectionally. These are evidenced from our VAR result in Appendix II.

Findings, Conclusion, and Policy Recommendation

The VAR result reveals that a shock in global oil volatility (OVX) cannot transmit significant shock to VIX and GVZ. But it can transmit some significant shocks on the six GCC stock markets. For instance, the VIX can transmit shocks on GVZ, Bahrain and Kuwait stock markets. This finding concords with the finding of Raza et al. (2016) that states that Bahrain is the most susceptible stock market in the GCC region. The global gold volatility (GVZ) can fairly affect the VIX and the UAE markets. From the whole result, the Bahraini and Kuwaiti markets are the two stock markets that are most susceptible to these global volatility variables. The finding further reveals the dominance of the GCC market in the Oil market as the OVX cannot transmit shock to any of the markets in the region but Qatar and UAE can cause OVX. This is true because these countries are the major contributors of global oil trades.

Conclusion

From the estimated results, the study empirically conclude that VIX and GVZ are capable of transmitting shocks to some of the GCC stock markets—(Bahrain and Kuwait) negatively and positively, respectively. While the OLS result revealed that OVX and VIX have impact on all the GCC stock markets. Hence, the study recommends that proper and strict financial and gold transaction policies should be institutionalized so as to mitigate the transmission of shocks into the markets. Also, financial and gold experts who regulate the stock and gold markets especially in Bahrain and Kuwait should attentively watch for any abnormality changes in the volatility movement of the financial and gold markets. Other GCC markets that are not affected should watch the same on a regular basis because a shock in any of the markets is capable of affecting other ally markets possibly in the long run.

Appendix

Appendix I: LA Length with only GCC Difference- 20 LAGs

Table A1. VAR Lag Order Selection Criteria. Endogenous variables: OVX VIX GVZ D(LNBRSI) D(LNKWTSI) D(LNOMNSI) D(LNQTSI) D(LNSARSI) D(LNUAESI). Exogenous Variables: C. Date: 11/30/18 Time: 12:41; Sample: 1/05/2009 8/16/2018. Included Observations: 1560

Lag	LogL	LR	FPE	AIC	SC	HQ
0	15587.56	NA	1.71e-20	-19.97251	-19.94163	-19.96103
1	21900.67	12545.29	5.81e-24	-27.96240	-27.65361*	-27.84759*
2	22033.38	262.1857	5.43e-24	-28.02870	-27.44199	-27.81055
3	22152.66	234.2805	5.17e-24*	-28.07777*	-27.21315	-27.75630
4	22231.29	153.5308	5.19e-24	-28.07474	-26.93220	-27.64993
5	22292.60	118.9853	5.32e-24	-28.04948	-26.62903	-27.52135
6	22352.29	115.1842	5.47e-24	-28.02217	-26.32380	-27.39071
7	22401.90	95.15488	5.70e-24	-27.98193	-26.00564	-27.24714
8	22449.29	90.33964	5.95e-24	-27.93884	-25.68463	-27.10071
9	22513.03	120.7788	6.08e-24	-27.91671	-25.38459	-26.97525
10	22575.89	118.3826	6.23e-24	-27.89345	-25.08342	-26.84866
11	22628.38	98.24812	6.46e-24	-27.85690	-24.76895	-26.70878
12	22679.19	94.53083	6.72e-24	-27.81820	-24.45234	-26.56675
13	22748.73	128.5470	6.82e-24	-27.80350	-24.15972	-26.44872
14	22803.54	100.7042	7.06e-24	-27.76993	-23.84823	-26.31182
15	22866.44	114.8246	7.23e-24	-27.74672	-23.54711	-26.18528
16	22925.26	106.7171	7.45e-24	-27.71829	-23.24076	-26.05352
17	22987.47	112.1338	7.64e-24	-27.69419	-22.93876	-25.92610
18	23057.72	125.8072	7.75e-24	-27.68040	-22.64705	-25.80898
19	23128.80	126.4955*	7.86e-24	-27.66769	-22.35642	-25.69294
20	23173.86	79.66475	8.24e-24	-27.62162	-22.03243	-25.54353

Source: Author calculation.

Notes: * indicates lag order selected by the criterion; LR: Sequential modified LR test statistic (each test at 5% level); FPE: Final Prediction Error; AIC: Akaike Information Criterion; SC: Schwarz Information Criterion; HQ: Hannan–Quinn Information Criterion.

Appendix II: The VAR Result with LAG 1

Table A2. Vector Autoregression Estimates. Date: 11/30/18 Time: 12:56. Sample (adjusted): 1/13/2009 8/16/2018. Included Observations: 1579 After Adjustments. Standard Errors in ( ) & t-statistics in [ ]

	OVX	VIX	GVZ	D(LNBRSI)	D(LNKWTSI)	D(LNOMNSI)	D(LNQTSI)	D(LNSARSI)	D(LNUAESI)
OVX(-1)	0.977682	0.005178	-0.000927	-2.58E-05	7.19E-07	-1.21E-05	-5.42E-05	-2.90E-05	-1.72E-05
	(0.00570)	(0.00529)	(0.00373)	(1.7E-05)	(2.3E-05)	(2.9E-05)	(4.0E-05)	(3.9E-05)	(3.3E-05)
	[171.492]	[0.97826]	[-0.24886]	[-1.51115]	[0.03115]	[-0.41717]	[-1.36873]	[-0.73714]	[-0.52354]

VIX(-1)	0.016272	0.940655	0.028568	-0.000111	-0.000106	-6.61E-05	-1.07E-06	-3.00E-05	-0.000112
	(0.01265)	(0.01175)	(0.00827)	(3.8E-05)	(5.1E-05)	(6.4E-05)	(8.8E-05)	(8.7E-05)	(7.3E-05)
	[ 1.28617]	[ 80.0757]	[ 3.45523]	[-2.92812]	[-2.06919]	[-1.02996]	[-0.01219]	[-0.34430]	[-1.53085]
GVZ(-1)	-0.018357	0.026618	0.936722	5.71E-05	9.14E-05	0.000102	8.09E-05	0.000115	0.000155
	(0.01530)	(0.01421)	(0.01000)	(4.6E-05)	(6.2E-05)	(7.8E-05)	(0.00011)	(0.00011)	(8.8E-05)
	[-1.19964]	[ 1.87350]	[ 93.6725]	[ 1.24334]	[ 1.47656]	[ 1.31592]	[ 0.76109]	[ 1.08911]	[ 1.75631]
D(LNBRSI(-1))	-3.385812	1.886182	1.730894	0.012830	-0.017278	0.026365	0.014735	0.059251	-0.015904
	(9.02461)	(8.37926)	(5.89762)	(0.02708)	(0.03652)	(0.04577)	(0.06266)	(0.06219)	(0.05201)
	[-0.37518]	[ 0.22510]	[ 0.29349]	[ 0.47384]	[-0.47314]	[ 0.57600]	[ 0.23518]	[ 0.95273]	[-0.30579]
D(LNKWTSI(-1))	-7.599693	-1.951514	-4.347470	0.034999	0.108645	-0.014273	0.038455	-0.014085	0.011338
	(7.07260)	(6.56683)	(4.62197)	(0.02122)	(0.02862)	(0.03587)	(0.04910)	(0.04874)	(0.04076)
	[-1.07453]	[-0.29718]	[-0.94061]	[ 1.64935]	[ 3.79619]	[-0.39789]	[ 0.78312]	[-0.28899]	[ 0.27817]
D(LNOMNSI(-1))	11.19083	-2.848061	0.712264	0.015034	-0.006518	0.039706	0.002373	-0.090157	-0.017572
	(6.06083)	(5.62742)	(3.96078)	(0.01818)	(0.02453)	(0.03074)	(0.04208)	(0.04177)	(0.03493)
	[ 1.84642]	[-0.50610]	[ 0.17983]	[ 0.82677]	[-0.26577]	[ 1.29166]	[ 0.05639]	[-2.15860]	[-0.50309]
D(LNQTSI(-1))	-20.29815	-2.951063	-0.527019	0.027245	0.052391	0.025491	0.083149	0.011411	0.072653
	(4.43055)	(4.11372)	(2.89538)	(0.01329)	(0.01793)	(0.02247)	(0.03076)	(0.03053)	(0.02553)
	[-4.58141]	[-0.71737]	[-0.18202]	[ 2.04957]	[ 2.92222]	[ 1.13435]	[ 2.70307]	[ 0.37374]	[ 2.84542]
D(LNSARSI(-1))	3.581699	3.204182	-1.737573	0.001106	0.029344	0.066779	0.160037	0.075253	0.017320
	(4.33788)	(4.02768)	(2.83482)	(0.01301)	(0.01755)	(0.02200)	(0.03012)	(0.02989)	(0.02500)
	[ 0.82568]	[ 0.79554]	[-0.61294]	[ 0.08495]	[ 1.67171]	[ 3.03518]	[ 5.31377]	[ 2.51739]	[ 0.69283]
D(LNUAESI(-1))	11.13220	6.788172	6.796018	0.003166	-0.034100	-0.010567	-0.051653	-0.023293	0.037418
	(5.31908)	(4.93871)	(3.47604)	(0.01596)	(0.02152)	(0.02698)	(0.03693)	(0.03666)	(0.03065)
	[ 2.09288]	[ 1.37448]	[ 1.95510]	[ 0.19839]	[-1.58426]	[-0.39167]	[-1.39868]	[-0.63547]	[ 1.22066]
C	0.787448	0.377511	0.660511	0.001691	0.000197	-0.000397	0.000644	-0.000303	0.000154
	(0.20020)	(0.18589)	(0.13083)	(0.00060)	(0.00081)	(0.00102)	(0.00139)	(0.00138)	(0.00115)
	[ 3.93321]	[ 2.03085]	[ 5.04844]	[ 2.81434]	[ 0.24277]	[-0.39067]	[ 0.46340]	[-0.21954]	[ 0.13370]
R-squared	0.970688	0.932068	0.939184	0.035418	0.032949	0.018704	0.042397	0.007599	0.017418
Adj.R-squared	0.970520	0.931678	0.938836	0.029885	0.027402	0.013075	0.036904	0.001907	0.011782
Sum sq. resids	6877.528	5929.070	2937.171	0.061911	0.112617	0.176924	0.331524	0.326610	0.228418
S.E. equation	2.093653	1.943935	1.368211	0.006282	0.008472	0.010619	0.014536	0.014428	0.012066
F-statistic	5773.150	2391.940	2692.258	6.401298	5.939815	3.322866	7.718458	1.334985	3.090410
Log likelihood	-3402.228	-3285.072	-2730.511	5770.245	5297.889	4941.250	4445.457	4457.247	4739.565
Akaike AIC	4.322011	4.173619	3.471198	-7.296068	-6.697770	-6.246041	-5.618058	-5.632992	-5.990583
Schwarz SC	4.355985	4.207593	3.505173	-7.262093	-6.663795	-6.212067	-5.584084	-5.599018	-5.956608
Mean dependent	34.85200	18.00528	18.31607	-0.000166	-5.12E-06	-0.000128	0.000253	0.000268	0.000423
S.D. dependent	12.19380	7.437055	5.532275	0.006378	0.008591	0.010689	0.014812	0.014442	0.012137
Determinant resid covariance (dof adj.)		9.70E-24
Determinant resid covariance		9.16E-24
Log likelihood		21716.39
Akaike information criterion		-27.39252
Schwarz criterion		-27.08675
Number of coefficients		90

Source: Author calculation.

Appendix III: The AR Root Table

Table A3. Roots of Characteristic Polynomial. Endogenous variables: OVX VIX GVZ. D(LNBRSI), D(LNKWTSI), D(LNOMNSI), D(LNQTSI) D(LNSARSI) D(LNUAESI), USD_INDEX. Exogenous variables: C; Lag specification: 1 1; Date: 11/06/18 Time: 13:05

Root	Modulus
0.998263	0.998263
0.978344	0.978344
0.933263	0.933263
0.907282	0.907282
0.104661	0.104661
0.070296 – 0.076708i	0.104047
0.070296 + 0.076708i	0.104047
0.052265 – 0.024229i	0.057608
0.052265 + 0.024229i	0.057608
0.001078	0.001078

Source: Author calculation.

Note: No root lies outside the unit circle. VAR satisfies the stability condition.

Appendix V

Figure A1. The AR Root Circle

Source: Author calculation.

Footnotes

Declaration of Conflicting Interests

The author declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.

Funding

The author received no financial support for the research, authorship and/or publication of this article.

References

Adam

Marcet

Pablo

J. N.

(2015). Stock market volatility and learning. Working Paper, 720, Federal Reserve Bank of Minneapolis Research Department.

Akgiray

V. G.

Booth

G. H.

John

Chowdhury

(1991). Conditional dependence in precious metal prices. The Financial Review , 26(3), 367–386.

Akgül

Bildirici

Özdemir

(2015). Evaluating the nonlinear linkage between gold prices and stock market index using Markov-switching Bayesian VAR models. Procedia - Social and Behavioral Sciences , 210, 408–415. https://doi.org/10.1016/j.sbspro.2015.11.388

Angellini

Bacchiocchi

Caggiano

Fanelli

(2017). Uncertainty across volatility regimes . CESifo Working Papers, CESifo Group Munich.

Arouri

M. E. H.

Lahiani

Nguyen

D. K.

(2011). Return and volatility transmission between world oil prices and stock markets of the GCC countries. Economic Modelling , 28(4), 1815–1825.

Ball

Torous

Tschogel

(1982). Gold and the weekend effect. Journal of Futures Markets , 2(2), 175–182.

Basher

Sadorsky

(2006). Oil price risk and emerging stock markets. Global Finance Journal , 17(2), 224–251.

Baur

D. G.

Lucey

B. M.

(2010). Is gold a hedge or a safe haven? An investigation of the informational efficiency of the GCC equity markets: Evidence from bootstrap simulation. International Review of Financial Analysis , 19(1), 47–54.

Caballero

Emmanuel

Gourinchas

(2008). An equilibrium model of global imbalances and low interest rates. American Economic Review , 98(1), 358–393.

10.

Chkili

(2016). Dynamic correlations and hedging effectiveness between gold and stock markets: Evidence for BRICS countries. Research in International Business and Finance , 38, 22–34. https://doi.org/10.1016/j.ribaf.2016.03.005

11.

Connolly

Stivers

Sun

(2007). Commonality in the time variation of stock-bond and stock-stock return comovements. Journal of Financial Markets , 10(2), 192–218.

12.

Daskalaki

Skiadopoulos

G. S.

(2011). Should investors include commodities in their portfolios after all? New evidence. Journal of Banking and Finance , 35(10), 2606–2626.

13.

Feng

Wang

X. M.

(2012). Log-transformation: Applications and interpretation in biomedical research Statistics in Medicine. Statistics in Medicine , 32(2), 230–239.

14.

Gomes

Chaibi

(2014). Volatility spillovers between oil prices and stock returns: A focus on frontier markets. The Journal of Applied Business Research , 3(2), 5027–5067.

15.

Kowalski

X. M.

(2007). Modern applied U statistics . New York, NY: Wiley.

16.

Kuang

(2018). Calculating volatility: A simplified approach . Investopedia.

17.

Kumar

(2017). On the nonlinear relation between crude oil and gold. Resources Policy , 51, 219–224. https://doi.org/10.1016/j.resourpol.2017.01.003

18.

Chang

(2012). Oil price shocks and gold returns. International Economics , 131, 71–103. https://doi.org/10.1016/S2110-7017(13)60055-4

19.

Ludvigson

S. C.

(2009). Macro factors in bond risk premia. Review of Financial Studies , 22(12), 5027–5067.

20.

Ludvigson

S. C.

(2017). Uncertainty and business cycles: Exogenous impulse or endogenous response? Tech. rep. National Bureau of Economic Research.

21.

Masih

De Mello

Peters

(2011). Oil price volatility and stock price fluctuations in an emerging market: Evidence from South Korea. Energy Economics , 33(5), 975–986.

22.

Mollick

A. V.

Assefa

T. A.

(2012). U.S. stock returns and oil prices: The tale from daily data and the 2008–2009 financial crisis. Energy Economics , 36, 1–18. https://doi.org/10.1016/j.eneco.2012.11.021

23.

Narayan

Zheng

(2010). Gold and oil futures markets: Are markets efficient? Applied Energy , 87(10), 3299–3303.

24.

OPEC annual bulletin. (2017). Organisation of Petroleum Exporting Countries (OPEC) annual bulletin (2017) .

25.

Raza

Shahzad

S. J. H.

Tiwarib

A. K.

Shahbaz

(2016). Asymmetric impact of gold, oil prices and their volatilities on stock prices of emerging markets. Resources Policy , 49, 290–301. https://doi.org/10.1016/j.resourpol.2016.06.011

26.

Reboredo

(2013). Is gold a hedge or safe haven against oil price movements? Resources Policy , 38(2), 130–137.

27.

Reilly

(2013). Advantages and disadvantages of using monthly, weekly and daily data . Retrieved from https://autobox.com/cms/index.php/blog/entry/advantages-and-disadvantages-of-using-monthly-weekly-and-daily-data

28.

Sum

(2013). The ASEAN stock market performance and economic policy uncertainty in the United States. Economic Society of Australia Papers , 32(4), 512–521.

29.

Tang

(2017). The relationship between gold price, euro, US dollar, oil price and stock market . Thesis, Centria University of Applied Sciences Business Management.

30.

Williams

(2013). VIX: What is it, what does it mean, and how to use it . Retrieved from Seekingalpha.com