This action might not be possible to undo. Are you sure you want to continue?
Gold Price Volatility and Stock Market Returns in India
P K Mishra Faculty in Economics, Siksha O Anusandhan University, Orissa, India E-mail: firstname.lastname@example.org J R Das Faculty in Management, Siksha O Anusandhan University, Orissa, India E-mail: email@example.com S K Mishra Faculty in Economics, TITE, Orissa, India E-mail: firstname.lastname@example.org Abstract The study of the capital market of a country in terms of a wide range of macroeconomic and financial variables has been the subject matter of many researches since last few decades. Recently one such variable, that is, gold price volatility has attracted the attention of many researchers, academicians and analysts. Thus, this paper is an attempt to analyse the causality relation that may run between domestic gold prices and stock market returns in India. The study by taking into consideration the domestic gold prices and stock market returns based on BSE 100 index, investigates the Granger causality in the Vector Error Correction Model for the period January 1991 to December 2009. The analysis provides the evidence of feedback causality between the variables. It infers that the Gold prices Granger-causes stock market returns and stock market returns also Granger-causes the gold prices in India during the sample period. Thus, both the variables contain some significant information for the prediction of one in terms of another.
Keywords: Gold Price, Stock Market Return, BSE 100 Index, India, Volatility, Causality JEL Classification Codes: C22, C32, E44
The study of the capital market of a country in terms of a wide range of macro-economic and financial variables has been the subject matter of many researches since last few decades. Empirical studies reveal that once financial deregulation takes place, the stock markets of a country become more sensitive to both domestic and external factors. And, one such factor is the price of gold. From 1900 to 1971, with the global systems of gold standard and USD standard, gold price was regulated. But, since 1972, gold has been disconnected from the USD. Particularly in 1976 when the International Monetary Fund (IMF) passed Jamaica Agreement, did gold begin to evolve from currency to ordinary merchandise and since then gold price has been determined by market supply and demand. And, in India, the government started the process of globalization and liberalization since 1991 which allowed prices to be determined by the market forces.
the government has been taking a number of steps to reform the gold sector and ensure that India benefits from the demand-influence that it has on the gold business internationally. The twin factors. the total annual supply of gold across the globe has also decreased from 4037 tons in 2002 to 3380 tons in 2008. From 2000 to 2009. If we look at the rolling . Since the gold prices in India are influenced by international factors.monthly. Moreover. But. There are several reasons gold has high demand in India. There is no credit risk attached to gold. Volatility involves short term . gold prices are continuously increasing. Dollar-Rupee rate and local taxes and levies. Any change in the global prices gets transmitted very quickly and gets reflected in domestic prices. particularly for countries like India who are price takers in gold with a major part of the demand met by imports. Gold also has taken the role of an asset of last resort. the average annual prices of gold also showed an increasing trend from the year 1991 to 1996. gold prices in India showed an increasing trend (Fig. its volatility is very important. allowing traders. The third reason for holding gold is to build a diversified portfolio. In the post liberalization period. World Economic History shows that countries have repeatedly used gold as security against loans when they have had difficulties with their Balance of Payments and have felt the need to borrow on the international capital markets. and (ii) appreciation of USD against INR. India is a large buyer of gold at about 700-800 tons per annum. (i) increase in global spot gold prices (as the commodity becomes dearer to those looking for safe haven during times of economic crisis. It also recycles about 200 tons of gold out of old jewellery. The first reason is security. The domestic gold price in India is continuously increasing due to its heavy demand in the country. Figure 1: Annual Price Movement of Gold in Indian Market Prior to the introduction of liberalization and globalization policies. Secondly. A large chunk of Indian imports is used for jewellery exports. first allowing a number of banks to import gold – braking the monopoly of the State Trading Corporations. it showed a decreasing trend in 1997 and 1998 and again showed an increasing trend in the year 2000. manufacturers as well as investors to trade in gold futures in India itself.Gold Price Volatility and Stock Market Returns in India 48 Since then. The liberalisation of the gold sector has been made in stages. then considerably reducing the import duty – destroying a lucrative parallel smuggling channel and now. led to sharp rise in gold prices in India in the recent past. weekly or even hourly fluctuations in gold prices as measured by their absolute percentage changes during a particular period. namely. The domestic gold prices in India are associated strongly with the import parity prices which are determined by the global spot prices. gold offers full security as long as it is retained by central banks. gold is able to maintain its liquidity even at times of crisis situations like high global inflation or political turbulence.1).
At the same time gold continues to forge ahead. gold continues to be a safe haven investment because gold prices rise in such circumstances (Gaur and Bansal. while the stock market declined from 2000 to 2003 and then again in 2008 (Fig. In 2008 when the market was suffering from bearish phase worldwide. the prices are more volatile after July 2007 which is almost the same time when the slow down started in USA as a result of the sub-prime crisis (Fig.49 P K Mishra. So far since March 2009 in India signs of recovery in the stock markets have emerged. J R Das and S K Mishra standard deviation of monthly gold prices since 2000. gold prices spiked as panic spread across global markets. 2010). big and small have chosen to hedge their investments through gold at the time of crises.3).2). It is no surprise that many investors. Figure 2: Standard Deviation of Gold Price in India A look at the historic data brings out that when the stock market crashes or when the dollar weakens. . Figure 3: Movement of Gold Price and BSE 100 Index 20000 16000 12000 8000 4000 0 92 94 96 98 00 02 04 06 08 BSE100 GOLDPRICE Gold prices have been on an uptick since 2000.
It is with this backdrop. i. In India gold has been held by individuals for years and have passed hands of many generations. The ADF unit root test has been performed by estimating the regression: ΔYt = α 0 + α1Yt −1 + ∑ γ jΔYt − j + ε t j=1 p The ADF unit root test is based on the null hypothesis H 0 : Yt is not I(0) . The gold investing habits of Indians strongly ingrained in the Indian Social Psyche. to a large extent. the equity culture in India is not as developed as in some other parts of the world. Section III makes the analysis. the Karl Pearson’s correlation coefficient between the aforesaid time series has been calculated and its significance has been tested by the t-test. Data and Methodology This paper aims at investigating the dynamic relationship between gold prices and stock market returns in India for the period 1991 to 2009. the two assets prices – equity and gold.Gold Price Volatility and Stock Market Returns in India 50 albeit at a slower pace. The study analyses the monthly data on domestic gold prices and stock market returns in India for the aforesaid period. during each of the two prolonged bear phases (lasting at least a year) over the past decade. The rest of the paper is organized as follows: Section II explains the data and methodology. t t −1 closing value of monthly BSE 100 Index at time ‘t’ and‘t-1’ respectively. However. guided by the individual sentiments.(∑ X)(∑ Y) r= N ∑ X 2 . If the calculated ADF statistic is less than the critical value. then the null hypothesis is rejected. At the outset. Then the Granger causality between the variables has been investigated in the Vector Error Correction framework. gold has provided an effective hedge. Gold has not yet lost its prime importance as a hedge against loss of wealth in times of crises. The reason for holding gold is. And. displaying the ability of the yellow metal to protect one's portfolios at the time of a dip. were moving in opposite directions. In fact. Wherever data were missing. The correlation coefficient has been calculated by using the formula: N ∑ XY . the significance of this correlation coefficient has been tested by the t-test using the tr n−2 under the null hypothesis H 0 : ρ = 0 against the alternative hypothesis of statistic t n − 2 = 1− r2 H1 : ρ ≠ 0 with n-2 degrees of freedom. otherwise accepted. In 2008. the averages of the data of the previous month and next month have been taken. 2. and Section IV concludes.e. The monthly stock market returns ( Rt ) based on BSE 100 Index have been calculated by the ⎛ I ⎞ Rt = log ⎜ t ⎟ ⎝ I t −1 ⎠ where I and I are the logarithmic difference change in the BSE 100 Index. otherwise accepted. In addition. the stationarity and cointegration between variables have been found out.(∑ Y)2 And. The Augmented Dickey-Fuller unit root test has been used to examine the stationarity of the time series of the study and to find the order of integration between them. This study is mainly based on secondary data that have been collected from the database on Indian economy maintained by Reserve Bank of India. as the essential steps of Granger Causality test.(∑ X)2 N ∑ Y 2 . If the calculated value of t exceeds the critical value of t. If the .. this paper proceeds to investigate the direction of causality between domestic gold prices and stock market returns in India. in India stocks do not seem to be perceived as an alternative to gold. then the null hypothesis will be rejected.
. In the Granger Causality test. where τ trace = −T ∑ log(1 − λi ). the variable is said to be co-integrated of order one. So.. the vector of endogenous variables is divided in two sub-vectors. The value of Pearson’s correlation coefficient (r) between these two time series over the period 1991 to 2009 is 0. divided by the number of restrictions pK1K 2 . In the second step. k λr +1 is the (r + 1)th largest squared eigenvalue.3 that the direction of movements of gold prices and BSE 100 Indices in India is same. the null hypothesis is rejected. Y2t with dimensions K1 and.. the Johansen’s cointegration test has been applied to check whether the long run equilibrium relation exists between the variables. the Granger Causality test has been used to determine whether one time series is useful in forecasting another thereby finding out the direction of relationship between the variables of the study. Since the t-statistic at 226 degrees of freedom is 26. the null hypothesis is that the number of distinct cointegrating vector(s) is less than or equal to the number of cointegration relations ( r ). The null hypothesis that Y1t is not Granger-causal for Y2t is tested by checking the null hypothesis α 21... The maximum eigenvalue test examines the null hypothesis of exactly r cointegrating relations against the alternative of r + 1 cointegrating relations with the test statistic: τ max = −T log(1 − λr +1 ). The Johansen approach to cointegration test is based on two test statistics. i = r +1 The trace test statistic can be specified as: where λi is the i th largest eigenvalue of matrix Π ..i ⎤ ⎡ Y1. and the maximum eigenvalue test statistic.i ⎦ ⎣ 2. is used in conjunction with an F(pK1K 2 . 1991).873.. To test whether this value of ‘r’ shows a significant relationship between two time series. Empirical Analysis It is clear from the Fig. K 2 respectively. In the trace test. In this case. and T is the number of observations. including the parameters of the deterministic term. . Y1t and. the null hypothesis of r = 0 is tested against the alternative of r + 1 cointegrating vectors.i = 0. Of course. the variable is said to be integrated of order zero. the test is based on a model with p + 2 lags of the endogenous variables. In the trace test..i α 22.i α12..51 P K Mishra. If the variable is non-stationary at level... p + 1 A Wald test statistic. i = 1. the levels VAR following form without exogenous variables of the model is considered. 2.9 and the critical value of t at 5% level of significance is less than it. J R Das and S K Mishra variable is stationary at level. 3..t −i ⎤ ⎡ u1t ⎤ ⎥ ⎢ Y ⎥ + CD t + ⎢ ⎥ ⎢ Y ⎥ = ∑ ⎢α ⎣ 2t ⎦ i =1 ⎣ 21.. The null hypothesis of the test is r = 0 against the alternative of r ≠ 0. so that K = K1 + K 2 .. I(1). For testing this property.t −i ⎦ ⎣ u 2t ⎦ as proposed by Dolado and Lütkepohl (1996).. + Bq X t − q + C*D*t + u t If that model contains p + 1 lags of the endogenous variables as in the above model. Here n * is the total number of parameters in the system (Lütkepohl. it can be said that the correlation between gold prices and BSE 100 indices is statistically significant. student’s t-test has been used. A 0 Yt = A1Yt −1 + . the ADF test can be utilised and the first difference of the variable can be used for testing a unit root. KT − n * ) distribution for testing the restrictions. ⎡ Y1t ⎤ p + 2 ⎡ α11. At the end. the trace test statistic. + A p +1Yt − p −1 + B0 X t + . The sub-vector Y1t is said to be Granger-causal for Y2t if it contains useful information for predicting the latter set of variables. the role of Y1t and Y2t can be reversed to test Granger-causality from Y2t to Y1t . viz. I(0)..
49471 0..0143 between them which is not statistically significant for the t-statistic of 0.e. Table 2: Results of Johansen’s Cointegration Test Sample: January 1991 to December 2009 Included observations: 225 after adjustments Trend assumption: Linear deterministic trend Series: Gold Prices and Stock Market Returns Lags interval (in first differences): 1 to 2 Unrestricted Cointegration Rank Test (Trace) Trace 0. computation reveals that the value of ‘r’ is 0. So it can be said that although gold prices and BSE 100 Indices are significantly correlated. The Augmented Dickey-Fuller unit root test has been used for this purpose. The Trace test indicates the existence of two cointegrating equations at 5% level of significance.61 Critical Values At 1%: -3. the cointegration between the stationary variables has been tested by the Johansen’s Trace and Maximum Eigenvalue tests. Thus.05 level * denotes rejection of the hypothesis at the 0.459 At 5%: -2. The results of these tests are shown in Table 2. the maximum eigenvalue test makes the confirmation of this result. the two variables of the study have long-run or equilibrium relationship between them.01 Reject Null hypothesis of no unit root It is clear from the Table 1 that the hull hypothesis of no unit roots for both the time series are rejected at their first differences since the ADF test statistic values are less than the critical values at 10%. And.459 At 5%: -2.841466 Max-eigenvalue test indicates 2 cointegrating eqn(s) at the 0.841466 Hypothesized No.Gold Price Volatility and Stock Market Returns in India 52 Thus. of CE(s) None * At most 1 * Prob. 5% and 1% levels of significances.062248 14.062248 14.05 Eigenvalue Statistic Critical Value 0.05 level * denotes rejection of the hypothesis at the 0. Table 1: Results of Augmented Dickey-Fuller Unit Root Test ADF Statistic -14.46069 3.05 level ** MacKinnon-Haug-Michelis (1999) p-values Unrestricted Cointegration Rank Test (Maximum Eigenvalue) Hypothesized Max-Eigen 0.46069 3.217 at 226 degrees of freedom.05 No. And. I(1).** 0.69901 15. the results of such test are reported in Table 1. Thus. it seems that gold prices and stock market returns based on BSE 100 Index are significantly correlated. i. The Granger causality test presumes that the given time series are stationary.23832 14.26460 At most 1 * 0.05 level ** MacKinnon-Haug-Michelis (1999) p-values Prob. of CE(s) Eigenvalue Statistic Critical Value None * 0.573 Decision Reject Null hypothesis of no unit root Variables in their First Differences Gold Prices Stock Market Returns -12.0001 .874 At 10%: -2.874 At 10%: -2. the correlation between gold prices and stock market returns based on BSE 100 Index is not significant.573 At 1%: -3. the variables are stationary and integrated of same order.0000 0.0000 0. But much interesting results have been obtained from the Granger Causality test.** 0. And. In the next step.264883 83.0001 Trace test indicates 2 cointegrating eqn(s) at the 0.264883 69.
038. gold remains an integral part of social and religious customs. a very low degree of correlation holds between them. gold has been more attractive than bank deposits. Thus. The extent of holding of gold in India is widespread but stocks are not held by all. The overall AUM in Gold ETFs at the end of December 2009 was Rs 1. New York and Sydney. Now. It shows that Indian investors are gradually moving into gold ETFs for investment instead of physical form.e. It is thus inferred that Indians have started considering gold more than jewellery and as good as investments on bonds and equities.000 0..114 which have been increased to 9. the first Gold Exchange Traded Fund. Perhaps. futures and options have become very popular and have been traded on various exchanges around the world and over-the-counter directly in the private market.53 Table 3: Results of Granger Causality Test P K Mishra. During the period of global financial crisis. both the variables contain some significant information such that they cause each other. people have often trusted gold as a better investment.. i.e. Recently derivatives such as gold forwards. The volume of Gold futures traded in this exchange during January to August 2007 was 4. .795 in January to August 2008. Indians consider gold the safe haven investment as a financial asset and as jewellery. In India the first gold ETF was launched in March 2007 by Benchmark Mutual Fund. this explains the co-movement of gold prices and stock prices in the aftermath of global financial crisis. i. J R Das and S K Mishra Null Hypothesis Gold Prices do not Granger Cause Stock Market Returns Stock Market Returns do not Granger Cause Gold Prices F-Statistic (73. Gold Bullion Securities was launched on the Australian Stock Exchange. the UTI gold ETF has emerged as the best performer since May 2009. gold futures are primarily traded on the New York Commodities Exchange. stocks and bonds. besides being the basic form of savings. stock markets crashed but gold price continues to increase in the country. the National Commodity and Derivatives Exchange introduced 100 gram gold futures in November 2006. up from Rs 717 crore in April 09.352 crore. In the USA. The number of new accounts created by Gold ETFs in India surged 57% between March and September 2009.000 Decision Reject Reject Now. World Gold Council Report says that India stands today as the world’s largest single market for gold consumption. In many countries including India. And. though retail participation in the Stock Markets might have gone up in the last few years. In India. But it is very interesting to note that these two variables are insignificantly correlated. But recently many innovative financial products have been lunched relating to gold. The results of the Granger causality test are reported in Table 3. 12) 11.997 Probability 0. In developing countries. In March 2003. gold exchange traded funds are being traded like shares on the major stock exchanges including London. This could be explained as follows. the Granger causality test can be performed to determine the direction of causation between these two variables in the Vector Error Correction Model.678 32.479. It is inferred that the null hypothesis of “Gold Prices do not Granger Cause Stock Market Returns” and “Stock Market Returns do not Granger Cause Gold Prices” are here clearly rejected. Traditionally.
Cheung..15. Econometrica. (1988). Vol. Asuncion. Gaur. BSE Review of Markets. Engle.. (1990). Dipak. P. (1982). Federal Reserve Bank of Richmond. Department of Economics. J. & DeGennaro. The Journal of Futures Markets. Vol. Bhattacharya. and Bansal. Levin. Then application of Granger causality test in the vector error correction model suggests the evidence of feedback causality running between the gold prices and BSE 100 Index based stock returns in India. (2007). “Bring Back the Gold Standard”. Princeton: Princeton University Press.26. pp. 115. Sampson (1980). Germany: Deutsche Bank Research. Journal of Business Finance and Accounting. pp.. “Gold as an Inflation Hedge: A Comparative Study of Six Major Industrial Countries”. “Deregulation of Gold in India .. 21. E. Vol.25. J. Vol. M. “The Value of Gold as a Reserve Asset”. (2002). Dickey. 3. Thus. pp. (1981). “The Nature and Efficiency of the Gold Market”. No. “Likelihood Ratio Statistics for Auto-Regressive Time Series with a Unit Root”. & Leal.A Case Study in Deregulation of a Gold Market”. “The Economics of Gold Price Movements”.49. Andrews. J. The Augmented Dickey-Fuller test says that the time series of the study are stationary and all integrated of order one. R. C. References                     Abken A. D. Inclan. Applied Econometrics Time Series. Vol. pp 3-13. & MacKinlay. A.. John Wiley and Sons.. K. & Schmedders.2. The study uses monthly data on the defined time series. A. “India's Capital Markets: Unlocking the Door to Future Growth”. 1057-1072. C. R. Dun’s Review (1980). 257-278. E. pp. No.1-28 Blose. “What Moves the Gold Market?”. Brodsky. University of St. Y. 14. and Gray P. 18-21. C. Economic Review. World Gold Council.55. “Making Wald Tests Work for Cointegrated VAR Systems”. 203-214. each variable contains some significant information so that one can be used to predict the other. The Econometrics of Financial Markets. R. Conclusion This paper examines the gold price volatility and the causality between domestic gold prices and stock market returns in India for the period 1991 to 2009.32-37. World Development. & Granger. Vol. and Woodward. Y. Aggrarwal. Inc. Ensers.Gold Price Volatility and Stock Market Returns in India 54 4. Research Study No 27. . Himadri. Vol. R.46-48. and Soenen. “Gold Price Risk and the Returns on Gold Mutual Funds”. R. Journal of Financial and Quantitative Analysis. (1996). R. 33-55. R. Volume 8. No. The required data have been collected from the database of Reserve Bank of India. Issue 3. David A.4. Econometrica. (1996). Sanjeev (2004). pp. (2004). A. Vol. A. T. The Journal of Portfolio Management. J. Agarwal. Discussion Paper Series No. L. W. Econometric Reviews. Dolado. “Volatility in Emerging Stock Markets”. and Lütkepohl. pp. pp. (1999). “A Comparative Study of Gold Price Movements in Indian and Global Markets”. Peter (1980). (2005). Vol. “Cointegration and Error Correction: Representation. A. A. 559-587. 48. W. (1997).251-276. “Bullion Markets”. J..0021. W. 369-386..34. Journal of Economics and Business. A. 9. J. Vol. Vol. M (2001).58-67. Laurence E. Journal of Financial and Quantitative Analysis. Pages 175-192. 191-197. H. Economic Theory.. Campbell. and Wong. Ó John Wiley & Sons. Cai. Singapore. Macmillan. Estimation and Testing”. pp. Indian Journal of Finance. “Excess Price Volatility and Financial Innovation”. & Fuller.2. pp. “Stock Returns and Volatility”. Second Edition. Peter and Wright. J. Chua. Citanna. Walter (1995). Lo. Vol. March /April. (1987). March 1980. 499-513. (2010).. The Johansen’s cointegration test reveals that there exists long run equilibrium relation between gold prices and stock market returns in India.. Aggarwal R. Ghosh. Baillie. “Gold as an Inflation Hedge”.
F. (2006).. pp. Vol. 11-25. S. Tschoegl. (1986). J R Das and S K Mishra Granger. MA: MIT Press. B. W. “Spurious Regressions in Econometrics”. and Neil Alan (1996).syr. (1963)..An Argument”. C. H. (1974). Vol. Econometrica.48. 1-27. “Developments in the Study of Cointegrated Economic Variables”. Taylor. “Gold in India”. pp.424-438. Journal of Economic Literature. Vol. MacKinnon.XLI. Vol. Gold Analysis. (1989). (1978). Vol. New York: John Wiley and Sons. “The Price of Gold and the Exchange Rate”. F.. New Delhi. (1993). 5th ed. 371-379.. pp. James G. “Forecasting Volatility in Financial Markets: A Review”. France and the Coming of the Depression: 1919-1932”.16. (2003). & Granger. Econometric analysis. Yale University. Vol.2. C. Vol. 478-539. & Philips. C. C.pdf Lütkepohl. Emerging Markets (May) . R. “The Role and the Rule of Gold. web. “The Impact of Political and Economic Events on Stock Behaviour”. Y. 47. Modelling Financial Time Series. Market Volatility. Economic and Political Weekly. Sjaastad. 14. C.. (1980). 4. (1986). The Research Foundation of CFA Institute. (1965). A. 4. Introduction to Multiple Time Series Analysis. pp.6. Youngje (2002). H. Kendall. No. Alfred A. J. 15. Doctoral Dissertation. “New Perspectives on Investing in Emerging Markets”. C. M. Princeton Paper No. pp. 1367-1393. and Henderson.61. (1991). “Market Anticipation of Government Policies and the Price of Gold”. W. and Leo Michelis (1999). Kyklos. D. “Efficiency in the Gold Market”. and Scacciavillani. Vol. Vol. Journal of the Royal Statistically Society. H. “Vector Autoregression and Causality”. (1953). Johnson. O. Vol. L. J. “Institutional Change on india's Capital Markets”. Vol.86. Reserve Bank of India Bulletin. S. Oxford Bulletin of Economics and statistics.XXXIV (3-4). RBI (1997). New Delhi: TMH Publishing Company Ltd. No. “The Gold Standard. Gujarati. Granger. S. Journal of International Money and Finance.” Journal of Applied Econometrics. Journal of Political Economy.edu/~ykim30/file/gold. “A Study of Gold Futures Price Spreads”.. Haug. Toda. & Morgenstern.              .55             P K Mishra. Granger. H. Columbia University. & Sangeetha. (1999). Springer Verlag. W. N. IIM. 86. Poon. (1998).. Ahmedabad . 227. William H.. “Spectral Analysis of New York stock Market Prices”. Granger. Subramaniam.249 Schill. and Hirsch. Johansen.. 879-897 Shiller. 563-577. Salent. A. Series A. W. Basic Econometrics. S. E. M. pp. “Numerical Distribution Functions of Likelihood Ratio Tests For Cointegration. “Investigating Causal Relation by Econometric Models and Cross Spectral Methods”.37. J. Kim. “The Analysis of Economic Time Series”.111-120. Clark (1994). Journal of Econometrics. (1996). Oxford: Oxford University Press. Journal of banking and Finance. S. (1995). Poitras. Berlin. (2003). June. 213-228. D. W.183-194. (2007). Econometrica. Cambridge. Greene. Shah. Rueff. G. Pearson Education Publishers. Likelihood-Inference in Cointegrated Vector Autoregressive Models. (1969).
This action might not be possible to undo. Are you sure you want to continue?
We've moved you to where you read on your other device.
Get the full title to continue listening from where you left off, or restart the preview.