Effects of Crude Oil and Gold Prices on US Stock Market: for USA from ARDL Bounds Testing

Transcription

1 ORIGINAL ARTICLE Effects of Crude Oil ad Gold Prices o US Stock Market: for USA from ARDL Bouds Testig Evidece Matiur Rahma 1 *, Muhammad Mustafa 2 1 McNeese State Uiversity 2 South Carolia State Uiversity Abstract: This paper explores the effects of chages i crude oil ad gold prices o US Stock market movemet. Daily data are used from the first busiess day of Jauary, 1986 to December 30, Efficiet uit root tests (DF-GLS ad Ng-Perro) are applied to examie the time series property of the variables i terms of statioarity or o-statioarity. ARDL Bouds Testig is applied for co-itegratio. Both DF-GLS ad Ng-Perro tests cofirm o-statioarity of each variable ad depict behavior of all the variables i log-levels, icluded i this study. The ARDL-Bouds testig cofirms co-itegratio amog the variables. There is evidece of log-ru covergece amog all these variables with very tepid adjustmet towards the equilibrium. Short-ru egative effects of chages i gold ad crude oil prices o US stock market returs are observed. The effect is statistically sigificat from gold price chages, but isigificat from crude oil price chages. Keywords: gold price; crude oil price; stock market retur; ARDL; VECM; causality; feedback; log-ru; short-ru 1. Itroductio The causal iteractive relatioship amog US stock market retur ad prices of crude oil ad gold is complex ad ofte ambiguous. The directio of causality is a usettled issue i the empirical literature of macroecoomics ad fiace. The complexity i their relatioship is cofouded by egative political shocks ad global uease i fiacial markets. Crude oil ad gold are strategic commodities. They are traded i US dollar. So, fluctuatios i the exteral value of US dollar sigificatly ifluece crude oil ad gold prices. The US dollar is the corerstoe of the world fiacial system. Obviously, its value matters to ivestors. Quite how it matters is very difficult to uderstad clearly. The explaatios keep chagig. Apparetly, a weak dollar boosts stocks. However, this iferece may occasioally fade away. They are also largely held i istitutioal portfolios i combiatio with global securities for diversificatio of risk. Moreover, cetral baks hold reserves i gold, major key currecies ad highly liquid safe short-term securities. Idividuals ad crude oil-exportig coutries also ivest i gold for hedgig agaist iflatio to preserve asset value. I brief, turmoils i oe market create shockwaves i other markets. The chages i differet markets do ot always move i tadem. The commo causes of chages iclude fluctuatios i iflatio rates, iterest rates ad exchage rates. They, i tur, affect cosumptio ad ivestmet that are coected to commodity, capital, currecy ad job markets. So, deepeig uderstadig of the effects of chages i crude oil ad gold prices o stock prices is of profoud importace [1 12]. The objective of this study is to empirically explore the log-ru ad short-ru iflueces of chages i prices of crude oil ad gold o chages i US S&P 500. To this effect, the ARDL bouds testig approach is applied usig daily data from the first busiess day of Jauary,1986 Copyright 2018 Matiur Rahma et al. doi: /fm.v3i This is a ope-access article distributed uder the terms of the Creative Commos Attributio Uported Licese ( which permits urestricted use, distributio, ad reproductio i ay medium, provided the origial work is properly cited. Fiace ad Market Volume 3 Issue

2 to December 30, The rest of the paper is structured as follows. Sectio 1 reviews the related empirical literature. Sectio 2 outlies the empirical desig. Sectio 3 reports results. Sectio 4 offers coclusios ad implicatios. 2. Review of Related Empirical Literature A extesive ad expadig volume of empirical literature exists o the pairwise ad trilateral dyamic iteractive relatioships betwee/amog prices of crude oil, gold ad equity returs across developed ad developig coutries over varyig sample periods usig data of differig frequecies ad distictly differet ecoometric tools or techiques. Usig data from Jauary, 2008 to Jauary, 2009, Sharma ad Madeep [13] applied regressio equatio model to ivestigate the pairwise relatioship betwee macroecoomic factors like chages i exchage rate, foreig exchage reserves, iflatio rate, gold prices ad stock value. Empirical results of the study revealed a strog relatioship betwee gold price ad the U.S. stock prices. Fidigs of this study revealed cyclic relatioship betwee gold ad stock prices. Historical data from 1930 to 1976 show that gold has egative beta. Whe icluded i ivestmet portfolio, it helps mitigate systematic risk. Fidigs of Dempster ad Artigas [14] proved that ivestmet strategies are highly correlated betwee gold ad stock market alog with profitability i the periods of iflatio ad deflatio. Levi ad Wright [15] examied the relatioship betwee gold prices ad the exteral values of US dollar. They applied co-itegratio techique usig data from Jauary, 1976 to August, 2005, ad foud log-term relatioship betwee these variables. The results revealed further that 1% icrease i the exteral value of US dollar leads to 1% icrease i gold prices. However, the relatioship may be ueve ad weak due to uforesee exteral shocks. Fidigs of their study also suggested positive relatioship betwee gold price movemet ad US iflatio, US iflatio volatility ad credit risk. Baur ad Mc Dermott [16] coducted a descriptive ad ecoometric aalysis of data from 1979 to Results of the study idicated that gold is mostly used as hedge agaist iflatio ad cosidered as safe have for major Europea ad US stock markets except Australia, Caada, Japa ad emergig markets such as BRIC coutries. Gold ivestors use it to protect the wealth i extreme ufavorable market coditios. This pheomeo creates higher demad for gold ad a overall icrease i gold prices, globally. Baur ad Mc Dermott [16] studied the tred of gold prices i egative market coditios ad foud it curviliear. They further suggested that egative market coditios exert sigificat impact o gold ivestors. Further, McCow ad Zimmerma [17] detected gold as Zero-beta asset implyig o market risk. Moore [18] empirically examied the relatioship betwee gold prices ad the value of stock markets empirical for the period of 1970 to They observed a egative relatioship betwee gold prices ad the value of stock markets which demostrated that a icrease i gold prices teds to cause a declie i the value of stock markets. These fidigs were also cofirmed by Buyuksalvarci [19] who ivestigated the effects of seve macroecoomic variables (cosumer price idex, moey market iterest rate, gold price, idustrial productio idex, oil price, foreig exchage rate ad moey supply) o the Turkish capital markets. Fidigs of this study showed that Turkish ivestors used gold as a alterative ivestmet tool for equities. I the cases of risig gold prices, they ivested less i stocks ad more i gold i view of their egative relatioship. Issamad Muride [20] re-explored the relatioship betwee chages i exchage rate ad equity prices i Idia, Pakista, Korea ad the Philippies durig by employig co-itegratio methodology. To add further, Muhammed et al. [21] explored the relatioship betwee exchage rates ad equity prices i Pakista, Idia, Sri Laka ad Bagladesh for the period of Fidigs of the above studies depicted o causal relatioship amog moetary variables ad equity returs. 2 Matiur Rahma et al. Fiace ad Market

3 I geeral, prices of commodities ted to move i uiso because they are iflueced by commo macroecoomic factors such as iterest rate, exchage rate ad iflatio [22]. To explai, oil ad gold, amog others, are the two strategic commodities which have received much research attetio recetly, partly due to surges i their prices ad rises i their ecoomic uses. Crude oil is the world s most commoly traded commodity ad its price is the most volatile i the commodity market. Gold is cosidered the leader i the market of precious metals as icreases i its price seem to lead to parallel movemets i the price of other precious metals[ 23]. Gold is also a ivestmet asset ad commoly kow as a safe have to avoid the icreasig risk i fiacial markets. Usig gold as oe of risk maagemet tools i hedgig ad diversifyig commodity portfolios, ivestors i both advaced ad emergig markets ofte switch betwee oil ad gold or combie them to diversify their portfolios [24]. High crude oil price adversely affects ecoomic growth ad hece pushes dow share prices. Cosequetly, ivestors switch to gold for safety as oe of alterative assets. Such a sceario prevailed durig the 1970s. Whe the oil cartel reduced crude oil output, there was a steep surge i crude oil price. The oil crisis i 1973 set shockwaves through the US ad the rest of the world. They led to a prologed global recessio i the 1970s. The impact of crude oil prices o gold prices could be established through the export reveue chael [15]. I order to disperse market risk ad maitai commodity value, domiat oil-exportig coutries use high reveues from sellig oil to ivest i gold. Sice several coutries icludig crude oil producers keep gold as a asset of their iteratioal reserve portfolios, risig crude oil prices may have implicatios for the surge i gold price. This holds true as log as gold accouts for a sigificat part i the asset portfolio of oil exporters ad oil exporters purchase gold i proportio to their risig oil reveues. Melvi ad Sulta [25] opie that the expasio of oil reveues ehaces the gold market ivestmet ad this causes price volatility of oil ad gold to move i the same directio. I such a sceario, a oil price icrease leads to a rise i demad for gold ad hece its price. Presumably, iflatio chael is the best ad the most commo way to explai the likage betwee oil ad gold markets. A rise i crude oil price leads to a icrease i the geeral price level [26,27]. Whe the geeral price level goes up, the price of gold icreases. Hece, surges i iflatio, caused by oil prices, lead to a icrease i demad for gold to hedge agaist it ad thus push up the gold price. O the other had, whe the gold price fluctuates due to chages i demad for jewelry, beig hoarded as a reserve currecy ad/or beig used as a ivestmet asset, it is ulikely to have aythig related to oil prices [23]. Several other studies support this view. Sari et al. [23] explore directioal relatioships betwee spot prices of four precious metals (gold, silver, platium, ad palladium), oil ad USD/euro exchage rate. They fid a weak ad asymmetric relatioship betwee oil ad gold prices. Specifically, gold price returs do ot explai much of oil price returs while oil price returs accout for 1.7% of gold price returs. O examiig the log-term causal ad lead-ad-lag relatioship betwee oil ad gold markets, Zhag et al. [28] report a sigificat co-itegratig relatioship betwee the prices of the two strategic commodities. They idicate that percetage chages of crude oil price retur sigificatly ad liearly Grager cause the percetage chage of gold price retur. Volatility of the exteral value of US dollar may cause fluctuatios i iteratioal crude oil ad gold prices to move i tadem sice both are traded i US dollar. To add further, durig expected iflatio time, the US dollar weakes agaist other major currecies. Cosequetly, ivestors move from dollar-deomiated soft assets to dollar-deomiated physical assets [23]. Deterioratio of US dollar vis-a'- vis euro may also push up oil price as crude oil trade is deomiated i US dollar. Zhag ad Wei [28] cofirm evidece of high correlatios betwee the US dollar exchage rate ad the prices of crude oil ad gold. Furthermore, Grager causality stemmig from the US dollar idex iduces the price chages of both commodities. Geopolitical evets are also aother factor that may impact the prices of crude oil ad gold simultaeously. I fact, both the commodity markets are very sesitive to the turmoil of iteratioal Fiace ad Market Volume 3 Issue

4 political situatio. Particularly, i times of fiacial crises, ivestors ofte rush to buy gold. Cosequetly, the price of gold escalates. Agai, Soytas et al. [24] show that the world crude oil price has o predictive power of the prices of precious metals icludig that of gold i Turkey. I reality, the situatio ca become eve more complicated, as oe ca observe that the crude oil price ad the gold price relatioship are ot stable over time. Several other studies do ot support ay of the above. I fact, some studies fid two-way feedback relatioships betwee crude oil ad gold prices [29]. Some idicate that the price of gold, amog others, triggers crude oil price. This implies that whe the world ecoomic system is hit by a commo stochastic shock, the gold price moves first ad the oil price follows [22]. This fidig, however, does ot support the commo belief that oil price is the leader of the formatio of overall iflatio. Several other previous studies also have show that oil price fluctuatios have asymmetric effects o macroecoomic variables ad the gold price [23,27,29]. Crude oil ad gold prices have discerible ecoomic impacts o fiacial activities, ad all sectors of the US ecoomy. This impact is directly apparet i cosumptio, idustrial productio ad ivestmet i both real ad fiacial sectors. Volatilities of crude oil ad gold prices ifluece directly stock prices with implicatios for the US capital market. Idirectly, they ifluece iflatio ad uemploymet [1]. Crude oil price is also iflueced by geopolitical ad weather related factors, which may create uexpected shifts i supply ad demad leadig to volatility i crude oil price. Uderstadig the volatility of crude oil price is very critical, because it may create ucertaity i all sectors of the ecoomy. I tur, there would be istability i the ecoomies of both oil-exportig ad oil-importig coutries. The crude oil price volatility exposes idustrial producers ad cosumers to risk. Goods ad services caot be provided i fair price, because of the reliace of these idustries o crude oil ad oil-related products [2]. The crude oil price volatility also affects derivative markets, because the value of a commodity is based o a cotiget claim affected by the volatility [3]. Gold is a precious metal cosidered as a commodity ad a moetary asset. It is viewed as a source of wealth, a uit of value ad medium of exchage [4]. Also, gold is used for ivestmet sice it is highly liquid ad a valuable metal for jewelry [5]. Traditioally, gold has bee a idicator of future iflatio, as a hedge agaist iflatio, a importat asset i portfolio allocatio ad has show its role i crises. Cetral baks ad iteratioal fiacial istitutios retai a large amout of gold for portfolio diversificatio, ad ecoomic security [6]. Volatility of gold price may further lead to egative cosequeces for fiacial markets, because a icrease i the gold price volatility leads to a usafe ivestmet coditio. I cotrast, lower gold price volatility leads to safe ivestmet coditio [7]. Thus, it is essetial to lear about gold price volatility for derivative valuatio, hedgig decisios, fiacial markets ad the overall ecoomy [8]. A icrease i gold volatility is a alert for ivestors ad the gold miig idustry, as it exposes them to risk. So, uderstadig the gold price volatility ehaces the uderstadig of fiacial markets [9]. Oil has a importat place i the US ecoomy, because volatility i oil price leads to chages i stock prices. I efficiet markets, oil ad stock prices are cotemporaeously correlated. If oil price icreases, it would cause declie i the stock price of compaies which cosume oil i their operatio. I efficiet markets, chages i oil price would also adjust with lagged chages i stock price [30]. A study by Joes ad Kaul [31] for the period of foud that oil price has o impact o real stock returs. Huag et al. [32] ivestigated the ifluece of crude oil prices o stock returs of the oil compaies by applyig the VAR model. They observed that oil futures returs ad stock returs move i the same directio. Cier [5] drew attetio to the impact of crude oil price o real stock returs, usig the o-liear coectio. This study foud that variability i the crude oil price affects the stock idex returs. Papapetrou [33] studied the relatioship amog oil price, real stock price, real ecoomic activity, ad iterest rates i Greece, by applyig a 4 Matiur Rahma et al. Fiace ad Market

5 multivariate VAR model. The results show that chages i the crude oil price sigificatly explai chages i the stock returs. Masih et al. [10] reported that volatility of oil price determies real stock returs. Cai et al. [34] examied the relatioship amog GDP, iflatio ad gold price. They argued that GDP ad iflatio have a strog impact o volatility of gold price returs. Capie et al. [35] claim that gold is a hedge agaist foreig exchage volatility. Baur ad McDermott [36] explored the impact of gold price o fiacial market for The fidig illustrated that gold acts as a hedge, ad a safe have, i the U.S. ad the most Europea coutries. Batte et al. [11] foud a sigificat impact of gold price volatility o the fiacial market returs. Mesi et al. [16] studied the correlatio ad volatility trasmissio across commodities such as gold ad crude oil, ad equity market. The results of their study revealed that the chages i S&P500 affect gold ad crude oil price volatility. Bhuia [37] studied the relatioship betwee gold price ad stock returs, usig Grager test. Bidirectioal causality betwee gold price ad stock was observed. Arouri et al. [12] implemeted VAR-GARCH model to explore the effect of gold price volatility o the stock market returs i Chia for Their results demostrated evidece of sigificat impact of gold price volatility o Chia s stock market retur. 3. Empirical Desig For estimatio, the followig basic logarithmic model is specified: LSP t = α 1 + β 1 LGOLD t + β 2 LOIL t + e t (1) Where, LSP = log of S&P500, LGOLD = log of gold price per troy ouce, LOIL = log of Bret Crude oil price per barrel, e = radom error term (white oise) ad t = time subscript. Iitially, o-statioarity/statioarity of each time series variable is ascertaied. A o-statioary time series has a differet mea at differet poits i time, ad its variace icreases with the sample size. The feature of o-statioary time series variables is very crucial. Ay liear combiatios of these time series variables create spurious regressio. Whe this happes, t-values of the coefficiets are highly sigificat, coefficiet of determiatio (R2) is very close to oe ad the Durbi-Watso (DW) statistic value is very low, which ofte lead to a high frequecy of Type-1 errors [38]. Cosequetly, the results of the estimated coefficiets become biased. Hece, it is ecessary to detect the existece of statioarity or o-statioarity i the time series variables i order to avoid the issue of spurious regressio. For this, the uit root tests are coducted usig the efficiet (DF-GLS), ad Ng-Perro tests [39]. If a uit root is detected for more tha oe variable, the test for co-itegratio should be used i terms of urestricted Error Correctio Mechaism (ECM). To aalyze possible log-ru relatioship amog stock market retur, gold price ad crude oil price, traditioal co-itegratio approaches are applied [40,41]. This study ivokes a more recet ad advaced approach to test whether log-ru relatioships amog the variables exist. The autoregressive distributive lag (ARDL) bouds testig approach is applied, as developed i Pesara et al. [42] because of its several advatages. For example, ARDL approach ca be applicable if the variables of iterest have ambiguous order of itegratio i.e. purely I(0), purely I(I)/, I(0) or /I(I) which is ot acceptable i the traditioal approaches. The above approach provides better results for small sample size ad the short-ru ad log-ru parameters simultaeously. Next step is to compare the calculated F-statistic with critical values which are geerated by computer software (EVIEW 9.5). Oe set assumes that all variables i the model are I(0) ad the other set assumes they are all I(1). If the calculated F-statistic exceeds the upper critical boud value, the the Ho of o co-itegratio is rejected. If the F-statistic falls withi the bouds, the the test is icoclusive. Lastly, if the F-statistic falls below the lower critical boud value, there is o co-itegratio. A ARDL represetatio of equatio (1) is specified as equatio (2) below: Fiace ad Market Volume 3 Issue

6 LSP t = λ 0 + i=1 μ t (2) λ 1i LSP t i + i=0 For ull hypothesis (Ho) of o co-itegratio, ψ 1 = ψ 2 = ψ 3 = 0 For alterative hypothesis, (HA) of co-itegratio, ψ 1 ψ 2 ψ 3 0 λ 2i LGOLD t i + i=0 λ 3i LOIL t i + ψ 1 LSP t 1 +ψ 2 LGOLD t 1 + ψ 3 LOIL t 1 + Third, a vector error-correctio model usig the first-differeces of the variables is estimated o the evidece of co-itegratio from equatio (2) for the lagged log-ru solutio, ad to determie the speed of adjustmet toward log-ru equilibrium. A geeral vector error-correctio model relatig to basic model (1) followig Egle ad Grager [40] is specified below: LSP t = θ 0 + e t 1 + θ i=1 1i LSP t i + θ i=0 2i LGOLD t i + i=0 θ 3i LOIL t i + μ t (3) The estimated coefficiet (π) of the error-correctio term (e t 1 ) is expected to be egative for log-ru covergece ad causal flows. If θ 1i s, θ 2i s ad θ 3i s are o-zeros, lagged chages i LSP ad LGOLD ad LOIL lead the chage i curret LSP i the short-ru. Their relative umerical magitudes idicate relative ifluece of the relevat explaatory variable o the depedet variable. The sum of the coefficiets of each lagged idepedet variable shows its et iteractive feedback effect with other variables. The optimum lag-legths are determied by AIC for good fit of the model with miimum loss of iformatio as i Akaike [43]. Fially, followig Pesara ad Pesara [44], CUSUM ad CUSUM-squares tests are applied for parametric stability. Gold price ad stock market retur data are collected from Oil price data are obtaied from Eergy Iformatio Website. 4. Results Table 1: Uit Root Tests (DF- GLS ad Na- Perro)* Level Differece Series DF-GLS NG-PERON DF-GLS NG-PERON LSP LGOLD LOIL *The modified Dicke-Fuller (DF-GLS) critical values are ad at 1% ad 5% levels of sigificace, respectively. The Modified Phillips-Perro (Ng-Perro) critical values are ad at 1% ad 5% levels of sigificace, respectively. Both DF- GLS ad Ng-Perro efficiet uit root tests cofirm o-statioarity of all three time series variables i log-levels at 1 ad 5 percet sigificace levels. At the same time, statioarity is restored o first-differecig of the variables i log-levels depictig I (1) behavior. Next, the ARDL bouds testig procedure is implemeted. The estimated results are reported as follows: Table 2: ARDL Bouds Test Null Hypothesis: No log-ru relatioships exist Test Statistic Value k F-statistic Critical Value Bouds Sigificace I0 Boud I1 Boud 10% % % % The computed F-statistic at exceeds the upper-boud critical F-value providig evidece of co-itegratio amog the variables at 5 percet level of sigificace. The log-ru estimated causal coefficiets are reported as follows: 6 Matiur Rahma et al. Fiace ad Market

7 Table 3: Log-Ru Co-efficiet (LSP5: Depedet Variable) Variable Co-efficiet Std. Error t-statistic Prob. LGOLD LOIL C The log-ru effect of gold o US stock market is positive ad that of the oil price is egative. However, each effect is isigificat i terms of the associated respective t-value. I other words, stock, gold ad oil markets have subdued log-ru causal flows to the US stock market. Fially, the estimates of the VECM are reported as follows: Table 4: ARDL Co-itegratig Ad Log Ru Form Depedet Variable: LSP Selected Model: ARDL(4, 1, 0) Co-itegratig Form Variable Co-efficiet Std. Error t-statistic Prob. (LSP5(-1)) (LSP5(-2)) (LSP5(-3)) (LGOLD) (LOIL) ECT(e t 1 ) Co-itegratig = LSP - (0.6934*LGOLD *LOIL ) The coefficiet of the error-correctio term has expected egative sig. The associated pseudo t-value of this coefficiet is sigificat. The abysmally low umerical value of the coefficiet idicates very tepid adjustmet toward log-ru equilibrium. Short-ru feedback effects of both gold ad oil price movemets to stock market are egative 4. Coclusios ad Implicatios US stock market (proxied by S&P 500) retur, gold price ad oil price i log-levels are ostatioary. They are also foud to coverges toward log-ru equilibrium with almost egligible speed of adjustmet. The short-ru iteractive feedbacks are egative. For chage i gold price, the feedback is statistically sigificat but it is isigificat for crude oil price. The parameters of the model seem somewhat ustable as uveiled i figures 1 ad 2 (Appedix-A) for CUMSUM ad CUMSUM-squares, respectively CUSUM 5% Sigificace CUSUM of Squares 5% Sigificace The markets for crude oil, commo stocks ad gold are itercoected i differet ways. Their relatioships are complex i which exchage rate chages; iflatio rate, uemploymet rate, ecoomic growth ad portfolio ivestmet performace are impacted. So global policy makers, traders ad ivestmet advisors should closely moitor the Fiace ad Market Volume 3 Issue

8 ufoldig developmets i all three markets, although chage i oe market affects the rest. substatial chages are couterproductive to the US ecoomy ad the rest of the world. Ay abrupt ad Refereces 1. CEbrahim, Z., O.R. Iderwidi, D.A. Kig. Macroecoomic impacts of oil price volatility: Mitigatio ad resiliece. Frotiers i Eergy, 2014, 8(1): Pidyck, R. S.. Volatility i atural gas ad oil markets. Cambridge, MA: Massachusetts Istitute of Techology, Naraya, P.K., S. Naraya. Modellig oil price volatility. Eergy Policy, 2007, 35: Goodma, B. The Price of Gold ad Iteratioal Liquidity. Joural of Fiace, 1956, 11(1): Cier, C. Eergy shocks ad fiacial markets: Noliear likages. Studies i Noliear Dyamics & Ecoometrics, 2001, 5: Kaufma, T. D., R. A. Witers. The price of gold: A simple model. Resources Policy, 1989, 15: Baur, D. G. Asymmetric volatility i the gold market. Joural of alterative ivestmets, 2012, 14: Ewig, B. T., F. Malik. Volatility trasmissio betwee gold ad oil futures uder structural breaks. Iteratioal Review of Ecoomics ad Fiace, 2013, 25: Tully, E., B.M. Lucey. A power GARCH examiatio of the gold market. Research i Iteratioal Busiess ad Fiace, 2007, 21: Masih, R., S., Peters, L. D. Mello. Oil price volatility ad stock price fluctuatios i a emergig market: Evidece from South Korea. Eergy Ecoomics, 2011, 33: Batte, J. A., C., Cier, B.M. Lucey. The macroecoomic determiats of volatility i precious metals markets. Resources Policy, 2010, 35: Arouri, M. E., A., Lahiai, D. K. Nguye. World gold prices ad stock returs i chia: Isights for hedgig ad diversificatio strategies. Ecoomic Modellig, 2015, 44: Sharma, G.D., M. Madeep. Impact of macro-ecoomic variables o stock prices i Idia. Global Joural of Maagemet ad Busiess Research, 2010, 10(7): Dempster, N., J.C. Artigas. Gold: Iflatio hedge ad log-ru determiats of the price of gold. World Gold Coucil Report, Research Study No. 32, Busiess School, Lodo, Levi, E.J., A. Motagoli, R.E. Wright. Short-ru ad log-ru determiats of the price of gold. World Gold Coucil Report, Research Study No.32, Busiess School, Lodo, Mesi, W., M., Beljid, A., Boubaker, S. Maagi. Correlatios ad volatility spillovers across commodity ad stock markets: Likig eergies, food, ad gold. Ecoomic Modellig, 2013, 32: McCow, R.J., J.R. Zimmerma. Is gold a zero-beta asset? Aalysis of the ivestmet potetial of precious metals. 18. Moore, G.H. Gold prices ad a leadig idex of iflatio. Challege, 1990, 33(4): Buyuksalvarci, A. The Effects of macroecoomics variables o stock returs: Evidece from Turkey. Europea Joural of Social Scieces, 2010, 14: Issam, S.A.A., V. Muride. Exchae rate ad stock price iteractios i emergig fiacial markets: Evidece o Idia, Korea, Pakista ad the Philippies. Applied Fiacial Ecoomics, 1997, 7(1): Muhammed, N., A. Rasheed, H. Fazal. Stock prices ad exchage rates: Are they related? Evidece from South Asia Coutries, Pakista Developmet Review, 2003, 41(4): Hammoudeh, S., R. Sari, B.T. Ewig. Relatioships amog strategic commodities ad with fiacial variables: A ew look. Cotemporary Ecoomic Policy, 2008, 27(2): Sari, R., S., Hammoudeh, U. Soytas. Dyamics of oil price, precious metal prices, ad exchage rate. Eergy Ecoomics, 2010, 32: Soytas, U., R., Sari, S. Hammoudeh, E. Hacihasaoglu. World oil prices, precious metal prices ad macroecoomy i Turkey. Eergy Policy, 2009, 37: Melvi, M., J. Sulta. South Africa political urest, oil prices, ad the time varyig risk premium i the gold futures market. Joural of Futures Markets, 1990, 10: Hut, B. Oil price shocks ad the U.S. stagflatio of the 1970s: Some isights from GEM. Eergy Joural, 2006, 27: Hooker, M.A. Are oil shocks iflatioary? Asymmetric ad oliear specificatios versus chages i regime. Joural of Moey, Credit ad Bakig, 2002, 34: Zhag, Y.J., Y.M. Wei. The crude oil market ad the gold market: Evidece for coitegratio, causality ad price discovery. Resources Policy, 2010, 35: Matiur Rahma et al. Fiace ad Market

9 29. Wag, K.M., Y.M. Lee. The ye for gold. Resources Policy, 2011, 36(1): Mork, K. A., O. Olse, H.T. Myse. Macroecoomic resposes to oil price icreases ad decreases i seve OECD coutries. Eergy Joural, 1994, 15: Joes, C.M., G. Kaul. Oil ad the stock markets. Joural of Fiace, 1996, 51: Huag, R.D., R.W. Masulis, H.R. Stoll. Eergy shocks ad fiacial markets. Joural of Futures Markets, 1996, 1: Papapetrou, E. Oil price shocks, stock market, ecoomic activity ad employmet i Greece. Eergy Ecoomics, 2001, 23: Cai, J., Y.L., Cheug, M.C., Wog. What moves the gold market? Joural of Future Markets, 2011, 21: Capie F., T. C., Mills, G. Wood. Gold as a hedge agaist the dollar. Joural of Iteratioal Fiacial Markets, Istitutios ad Moey, 2005, 15: Baur, D.G., T.K., McDermott. Is gold a safe have? Iteratioal evidece, Joural of Bakig ad Fiace, 2010, 34: Bhuia, A. Coitegratio ad casual relatioship amog crude price, domestic gold price ad fiacial variables: A evidece of BSE ad NSE. Joural of Cotemporary Issues i Busiess Research, 2013, 2: Grager, C.W.J., P. Newbold. Spurious regressios i ecoometrics. Joural of Ecoometrics, 1974, 2: Elliot, G., T.J. Rotheberg, J.H. Stock. Efficiet tests for a autoregressive uit root. Ecoometrica, 1996, 64(4): Egle, R.F., C.W. J. Grager. Coitegratio ad error correctio represetatio: Estimatio ad testig. Ecoometrica, 1987, 55(2): Johase, S., K. Juselius. Maximum likelihood estimatio ad iferece o coitegratio with applicatio to the demad for moey. Oxford Bulleti of Ecoomics ad Statistics, 1990, 52(2): Pesara, M.H., Y. Shi, R.J. Smith. Bouds testig approaches to the aalysis of level relatioships. Joural of Applied Ecoometrics, 2001, 16(3): Akaike, H. Fittig autoregressio for predictio, aals of the istitute of statistical mathematics, 1969, 21: Pesara, M.H., B. Pesara. Workig with microfit 4.0: Iteractive ecoometric aalysis. Oxford: Oxford Uiversity Press, Fiace ad Market Volume 3 Issue