Soy Oil Efficiency in India

IRACST- International Journal of Research in Management & Technology (IJRMT), ISSN: 2249-9563 Vol. 4, No.2, April 2014 Efficiency Tests of Refined Soy Oil Futures in India Alok Kumar Sahai Ilead School of Business Indore, India Abstract— The concepts of market efficiency and unbiasedness are difficult to distinguish empirically. Market efficiency implies that future prices will equal expected future spot prices plus or minus a risk premium, while futures prices will be unbiased forecasters of futures spot prices only if the markets are both efficient and have no risk premia. The hypothesis that futures prices are unbiased forecasters of spot prices is thus a joint hypothesis of market efficiency and risk neutrality. Further, a market may be efficient and unbiased in the long run, but may exhibit short term inefficiencies. This paper attempts two-stage Engle Granger (E-G) cointegration procedure to test for long run market efficiency and unbiasedness in the refined soy oil (RSO) futures market. Wald test of coefficients restriction is used to test the efficiency and existence of risk premium. Four price series for futures and corresponding spot prices of different maturities are used to test the efficiency. Granger causality is employed to test the direction of price adjustment. RSO prices in two periods were studied. E-G test and Wald test of coefficient restriction showed that while the cointegration has increased across pairs of different maturities over time, the tests of coefficient restrictions do not support the unbiasedness hypothesis. RSO futures markets in India are found to be inefficient in price discovery and unsuitable for optimal hedging. Keywords- Market efficiency, unbiasedness, Wald test, Engle Granger, Soybean oil. Cointegration I. INTRODUCTION Futures markets have two important economic functions namely price risk management and price discovery. Producers and consumers can take suitable positions in the futures market to minimize their risk exposure in the spot market. Apart from these hedgers, there are speculators and other traders in the markets who aim to take risk and profit by taking such positions in the futures markets. With the presence of market participants with various objectives and information, the futures market enables the current futures price to act as an accurate indicator of the spot price expected at the maturity of the contract. This is the price discovery function of the futures markets. Only an efficient futures market can ensure price discovery. The market is weak form efficient as postulated by Fama [7], if the futures price reflects all the available information for predicting the future spot price and any participant can not make profits consistently. The efficiency of commodity futures market has been an issue of debate for sometime. An efficient commodity market should act as an effective and unbiased predictor for the futures spot price and reflect the equilibrium value of supply and demand in the market. Cointegration between spot and futures prices is conventionally regarded as one of the necessary conditions for market efficiency. It ensures at least a long run equilibrium relationship between the two prices. Without this equilibrium the spot and futures prices will drift apart and futures prices will provide little information about the future spot prices. The commodity market scene in India improved after liberalization in 1991 and subsequent lifting of ban on the trading of commodities in 2003. The commodities futures markets are relatively new in India and not as mature as the equity markets. The study of market efficiency in agricultural commodity futures markets is important to both the government and the producers/marketers. For the government, an efficient market means a better alternative to the market interventions and policy measures. For producers/marketers, it provides a reliable forecast of spot prices in the future to allow them to effectively manage their risks in production or marketing. This paper attempts to test the efficiency of commodity market by studying the refined soy oil (RSO) spot and future price series. RSO is the second most important edible oil in India and forms over a quarter of the agricultural commodity traded by value at the three major national exchanges namely National Commodity and Derivatives Exchange (NCDEX), ACE Commodity Exchange (ACE) and Indian Commodities Exchange (ICEX). The ticker symbol of refined soy oil on NCDEX is SYOREFIDR. 96 IRACST- International Journal of Research in Management & Technology (IJRMT), ISSN: 2249-9563 Vol. 4, No.2, April 2014 II. THE UNBIASEDNESS HYPOTHESIS A non stationary time series is said to be integrated of order one, denoted by I (1), if the series is stationary after the first order differencing. A pair of time series is said to be cointegrated if each of the series taken individually is I (1) and a linear combination of the two series is stationary. The unbiasedness hypothesis is a joint assumption of both market efficiency and risk neutrality (Beck, 1994) and is represented as follows: St = β0+β1Ft-1+β2Rt-1+vt (1) Where St and Ft-1 are the natural logarithms of the spot and futures prices at time t and t-1, Rt-1 is the zero mean risk premium and vt is the random error. The risk premium can be ignored in this equation as it is considered to be stationary in theory. The cointegrating equation can now be specified as: St = β0 + β1Ft-i+ ut (2) Where ut = β2Rt-1+vt and must be integrated in level. If both St and Ft-i are I (1), a condition which normally holds good for financial time series, the vector process (St, Ft-i) is cointegrated. This cointegration between St and Ft-i is a necessary condition for market efficiency [12]. Cointegration ensures that a long run equilibrium relationship holds between the two series. If however, St and Ft-i are not cointegrated, the prices will drift apart and futures price will provide little information about the movement of the spot price. The unbiased hypothesis requires that β0=0, β1=1 and ut should be serially uncorrelated. Rejection of the null hypothesis can therefore be explained by one of the following: 1) The futures market is inefficient, 2) A non-zero risk premium exists, other, futures prices would not be helpful in predicting the spot price [12]. Cointegration is a necessary condition for market efficiency. Market efficiency also requires that futures price be an unbiased predictor of the spot price, indicating the presence of a cointegrating vector between the two price time series. The studies by Garbade and Silber [8] and Kawaller, Koch and Koch [10] and Shroeder and Goodwin [19] indicate that price discovery occurs more significantly in the futures market compared to the cash market. Studies on market efficiency have been conducted for mainly developed countries so far [5, 14]. The study of the lead lag relationship among future and spot prices using Johansen's cointegration framework have abounded in the financial markets literature across the world [3,6,12,14,20]. Similar studies in India also have been reported after commodity market reforms were allowed by the government. Research in India on price discovery and efficiency of the commodities markets include [1,4,9,11,13,,15,16,17,18]. However, no studies have covered the future and spot relationship of oilseeds. Sahadevan [15] performed tests on futures and spot prices for six agricultural commodities traded at different regional exchanges between January 1999 to August 2001 and obtained results rejecting β0=0, β1=1. Raizada and Sahi [16] tested spot and futures prices for wheat contracts traded on NCDEX between July 2004 to July 2006 and obtained results rejecting β0=0 and β1=1. Ali and Gupta [1] in their study examined the price discovery of 12 major agricultural commodities using cointegration and Granger causality analysis between July 2004 to January 2007 and found significant cointegration between futures and spot prices for all selected agricultural commodities excluding wheat and rice. Sehgal, Rajput and Dua [18] examined the price discovery of 10 major agricultural commodities using cointegration and Granger causality analysis between November 2003 to March 2012 and found significant cointegration between futures and spot prices in 9 out of 10 selected commodities. 3) The markets are inefficient and also a risk premium exists. III. LITERATURE REVIEW The existence of price discovery, market efficiency, and market stability associated with spot and futures markets continues as a prominent discussion among academics, practitioners and regulators. Numerous papers examine the role of price discovery in the futures markets for various commodities and financial assets. If the non stationary spot and futures prices are cointegrated, it indicates a long term equilibrium relationship among them. However, if the prices did not move in tandem with each From the empirical literature cited above, it is clear that studies on future contracts on agricultural commodities are few. None of the existing available research has tested the efficiency and unbiasedness of RSO futures with different maturities. Further, large number of studies which have tested the cointegrating relationships have not tested the cointegrated vectors, which might have led to incomplete conclusions regarding efficiency. This study attempts to fill the research gap by testing the price discovery and unbiasedness of futures prices on predicting the spot prices data on refined soy oil, the most important edible oil in India. Unlike the existing studies this study makes use of futures prices of different maturities using daily closing prices. 97 IRACST- International Journal of Research in Management & Technology (IJRMT), ISSN: 2249-9563 Vol. 4, No.2, April 2014 IV. DATA This study uses data from refined soy oil futures markets in India. Daily futures price data of refined soy oil during the period of October 2004 to January 2014 are obtained from the National Commodity and Derivatives Exchange (NCDEX). Matching spot market closing prices are also obtained from NCDEX. On expiry the futures prices are supposed to converge into the spot prices. Soy oil prices are not available due to a break in 2008 when trading in soy oil was banned due to excessive volatility. Accordingly, data is collected for two periods namely October 2004 to April 2008 and December 2008 to January 2014. Futures market efficiency is tested for f o u r forecasting horizons, ranging from one week to three months. Accordingly, futures prices are taken at one week, one month, t w o months and three months prior to the maturity of each contract. In summary, we have four futures price series and four corresponding spot series for each period. Each futures price series consists of prices taken at a particular period prior to the expiry date of each contract. The label of each series is listed in Table 1. F1W, F1M, F2M and F3M indicate futures prices one week, one month, two months and three months prior to expiry of the contract. The corresponding spot series are called S1W, S1M, S2M and S3M respectively. The number of observations is 256 in 2004-2008 period and 396 in the 2009-2014 period across eight price series. V. METHODOLOGY Raw price series are first converted to logarithmic series for both the time periods. Next each series is tested for the order of integration using Augmented Dickey Fuller (ADF) test. Cointegration between the eight pairs is tested using Engle Granger two-step methodology. The E-G methodology starts with establishing a linear equation between the spot and futures prices. Cointegration between the two price series is established if the residual series is stationary in level. Market efficiency is thus established. The price discovery function is tested by testing the parameter restrictions on equation 2. The parameter restrictions β0=0 and β1=1 are tested in multiple settings. We conduct parameter testing separately for β0=0 and β1=1 and also jointly as β0, β1 = (0, 1). VI. RESULTS The table A1 shows the descriptive statistics of the time series data. The standard deviation of the 2009-2014 series has increased by more than twice from the 2004-2008 series indicating a fourfold rise in volatility. Table A2 presents the unit root testing data on the eight series. Augmented Dickey Fuller (ADF) test has been applied under conditions of intercept only, intercept and trend & no intercept and no trend. The difference series have p values below 0.01 indicating that all the price series are integrated of first order, I (1). Table A3 presents the Engle Granger two step test of cointegration for the period of 2004-2008. One week to expiry and one month to expiry futures series are cointegrated (p<0.05) but the longer expiry futures of two month and three month expiry are not cointegrated with the spot prices. This means that futures prices contain better information about the supply and demand of the commodity when it gets closer to maturity. Table A4 presents the results of Wald test. We fail to reject the hypothesis β1 =1 in all but the farthest maturity (F3M) indicating that futures prices are unbiased predictor of the spot prices. Results of the Engle Granger test of cointegration for the second period of 2008-2014 are presented in Table A5. All the residual series have p value less than 0.05 indicating cointegrated futures and spot series across all maturities. A look at Wald test of coefficient restrictions in Table A6 reveals that hypothesis β1 =1 is rejected in all but the farthest maturity (Spot-F3M) pair. Pairwise Granger causality tests exhibited no causality except one month maturity pair in the 2008-2014 period (Table 7). This further confirms that the future price series can not be used to predict the spot price at maturity. Cointegration is confirmed to exist but the test of coefficient restrictions do not support the price discovery and the unbiasedness hypothesis for refines soy oil. The soy oil futures, therefore, can not be used for optimal portfolio hedging and the RSO futures markets are not fully efficient. The constraint β1=1 is a more important indicator for market efficiency because β0 can be non zero under the existence of risk premium and transportation costs even when the markets are efficient. Granger causality is conducted for the eight pairs of price series to confirm the findings. 98 IRACST- International Journal of Research in Management & Technology (IJRMT), ISSN: 2249-9563 Vol. 4, No.2, April 2014 APPENDIX Table A1: Descriptive Statistics for RSO spot and future series Oct 2004- Apr 2008 F1M F1W F2M F3M S1M S1W S2M S3M Mean 414.41 410.36 417.14 417.21 409.50 409.08 409.82 Median 406.35 404.35 413.20 414.75 396.45 399.58 398.05 408.68 406.00 Maximum 510.75 501.50 512.05 504.55 507.05 502.30 507.05 504.75 Minimum 344.75 342.05 346.95 350.35 342.05 340.95 342.05 342.20 Std. Dev. 47.35 47.99 48.39 47.49 46.83 47.81 47.77 45.42 Skewness 0.3187 0.3916 0.2665 0.3208 0.4035 0.4167 0.4255 0.3643 Kurtosis 1.8653 1.9716 1.7415 1.7700 2.0123 2.0398 2.0156 2.1192 Jarque-Bera 2.2585 2.2280 2.4908 2.5660 2.1692 2.1552 2.2577 1.7423 Probability 0.3233 0.3282 0.2878 0.2772 0.3380 0.3404 0.3234 0.4185 F1W F1M F2M F3M S1W S1M S2M Dec 2008-Jan 2014 S3M Mean 594.10 576.82 587.58 577.42 593.31 578.23 591.24 Median 625.25 598.63 625.95 600.53 625.05 592.63 624.13 578.03 584.50 Maximum 760.65 725.30 810.50 788.90 761.75 733.50 791.90 785.75 Minimum 420.75 435.75 429.30 429.35 417.30 429.95 439.35 429.95 Std. Dev. 109.12 106.23 110.19 111.29 110.76 111.21 116.03 114.91 Skewness -0.1984 -0.0369 0.0512 0.1573 -0.1774 0.0152 0.0199 0.1893 Kurtosis 1.4524 1.2912 1.6139 1.5767 1.4490 1.2874 1.3886 1.4789 Jarque-Bera 4.8924 5.6074 3.7024 4.0724 4.8522 5.6233 4.9799 4.7096 Probability 0.0866 0.0606 0.1570 0.1305 0.0884 0.0601 0.0829 0.0949 F1W Futures prices one week before expiry F1M Futures prices one month before expiry F2M Futures prices two months before expiry F3M Futures prices three months before expiry S1W Spot price one week before expiry S1M Spot price one month before expiry S2M Spot price two months before expiry S3M Spot price three months before expiry 99 IRACST- International Journal of Research in Management & Technology (IJRMT), ISSN: 2249-9563 Vol. 4, No.2, April 2014 Table A2: Unit Root Test Results Oct 2004-April 2008 Dec 2008-Jan 2014 Exogenous-Constant Exogenous-Constant Level t statistics First Difference p value t statistics Level p value t statistics First Difference p value t statistics p value F1W -0.092549 0.9418 -5.098557 0.0003 F1W -1.259513 0.6401 -7.814523 0.0000 F1M -0.589237 0.8591 -6.115337 0.0000 F1M -0.181696 0.9338 -3.618929 0.0087 F2M -0.679401 0.8376 -5.344667 0.0001 F2M -1.02926 0.7351 -7.177804 0.0000 F3M -0.388552 0.8993 -5.311543 0.0001 F3M -1.319625 0.6138 -7.314288 0.0000 S1W -0.003223 0.9512 -5.240793 0.0002 S1W -1.213685 0.6605 -7.558161 0.0000 S1M -0.246232 0.9220 -5.823575 0.0000 S1M -0.529534 0.8764 -10.78675 0.0000 S2M -0.159721 0.9337 -5.180877 0.0002 S2M -0.805532 0.8081 -7.472813 0.0000 S3M 0.081697 0.9590 -4.783082 0.0006 S3M -0.895771 0.7819 -7.615995 0.0000 Exogenous-Constant, Linear Trend Level t statistics Exogenous-Constant, Linear Trend First Difference p value t statistics Level p value t statistics F1W -2.286293 0.4286 -5.184689 0.0012 F1W F1M -2.678896 0.2514 -6.022737 0.0001 F2M -2.648457 0.2632 -5.129637 0.0013 F3M -2.306381 0.4184 -5.583394 First Difference p value -2.04111 0.5635 F1M -2.00261 F2M -2.071275 0.0004 F3M -2.023749 t statistics p value -7.730251 0.0000 0.5856 -3.55651 0.0442 0.5476 -7.119689 0.0000 0.5748 -7.242261 0.0000 S1W -2.25365 0.4452 -5.260863 0.0010 S1W -1.982424 0.5949 -7.467915 0.0000 S1M -2.387559 0.3783 -5.904071 0.0002 S1M -1.301851 0.8761 -10.67558 0.0000 S2M -2.286128 0.4287 -5.208929 0.0011 S2M -2.100564 0.5318 -7.379734 0.0000 S3M -2.014635 0.5708 -5.52057 0.0005 S3M -2.305319 0.4236 -7.535299 0.0000 Exogenous-None Exogenous-None Level t statistics First Difference p value t statistics Level p value First Difference t statistics p value t statistics p value F1W 1.176858 0.9349 -5.097695 0.0000 F1W 1.077662 0.9243 -7.704345 0.0000 F1M 0.857345 0.8903 -5.9289 0.0000 F1M 2.666394 0.9978 -2.340647 0.0200 F2M 0.849453 0.8889 -5.183326 0.0000 F2M 1.087413 0.9256 -6.967119 0.0000 F3M 0.932002 0.9023 -5.288838 0.0000 F3M 1.033139 0.9188 -7.248017 0.0000 S1W 1.179945 0.9352 -4.986286 0.0000 S1W 1.093078 0.9262 -7.470749 0.0000 S1M 1.002904 0.9128 -5.615065 0.0000 S1M 3.670773 0.9999 -3.741085 0.0004 S2M 1.093707 0.9249 -4.998298 0.0000 S2M 0.998974 0.9137 -7.283325 0.0000 S3M 0.913316 0.8994 -4.753663 0.0000 S3M 0.957269 0.908 -7.542932 0.0000 100 IRACST- International Journal of Research in Management & Technology (IJRMT), ISSN: 2249-9563 Vol. 4, No.2, April 2014 Table A3: Engle Granger Cointegration Test (2004-2008) 2004-2008 Residual Series Regression Results Coefficient Spot-F1W Spot-F1M Spot-F2M Spot-F3M Std. Error t-Statistic p value β0 0.016017 0.057286 0.279604 0.7817 β1 0.99682 0.009529 104.6064 0.0000 β0 0.051018 0.129997 0.392453 0.6975 β1 0.989552 0.021588 45.83724 0.0000 β0 0.11015 0.206908 0.532364 0.5984 β1 0.978785 0.034324 28.5158 0.0000 β0 0.318955 0.273283 1.167122 0.2524 β1 0.943704 0.045333 20.81738 0.0000 R-squared t stat prob* 0.997266 -3.743442 0.0082 0.985922 -3.052948 0.0410 0.964419 -2.383607 0.1554 0.935256 -2.713682 0.0831 Table A4: Wald Test of Coefficient Restrictions (2004-2008) Wald Test of Coefficient Restrictions Spot-F1W χ2 Spot-F1M χ2 p value Spot-F2M p value χ2 Spot-F3M p value χ2 p value β0=0 0.078178 0.7798 0.154019 0.6947 0.283411 0.5945 1.362173 0.2432 β1=1 0.111369 0.7386 0.234212 0.6284 0.382005 0.5365 6.210492 0.0127 β0=0 & β1=1 8.293136 0.0158 24.64022 0.0000 21.08302 0.0000 17.93793 0.0001 Table A5: Engle Granger Cointegration Test (2008-2014) 2008-2014 Residual Series Regression Results Coefficient Std. Error t-Statistic p value Spot-F1W β0 β1 1.014598 0.005901 171.9264 0.0000 Spot-F1M β0 -0.206088 0.088543 -2.327555 0.0240 β1 1.032625 0.013887 74.36004 0.0000 Spot-F2M β0 -0.200429 0.155076 -1.292457 0.2028 β1 1.032151 0.024361 42.36919 0.0000 Spot-F3M β0 -0.210833 0.18049 -1.168113 0.2482 β1 1.034192 0.028361 36.46501 0.0000 -0.094878 0.037606 -2.522928 R-squared 0.0153 t stat prob* 0.998514 -6.887553 0.0000 0.991038 -8.775275 0.0000 0.975545 -3.658891 0.0081 0.963062 -3.470033 0.0128 Table A6: Wald Test of Coefficient Restrictions (2008-2014) 2008-2013 Wald Test of Coefficient Restrictions Spot-F1W χ2 Spot-F1M p value χ2 Spot-F2M p value χ2 Spot-F3M p value χ2 p value β0=0 6.365166 0.0116 5.417511 0.0199 1.670446 0.1962 1.364489 0.2428 β1=1 6.118844 0.0134 5.519402 0.0188 59.92151 0.0000 1.453427 0.2280 β0=0 & β1=1 9.022299 0.0110 5.96815 0.0506 2.552652 0.2791 3.036753 0.2191 101 IRACST- International Journal of Research in Management & Technology (IJRMT), ISSN: 2249-9563 Vol. 4, No.2, April 2014 Table A7: Pairwise Granger Causality Tests Pairwise Granger Causality Tests 2004-2008 Lags: 5 Null Hypothesis: Obs LS1W does not Granger Cause LF1W 27 LF1W does not Granger Cause LS1W LS1M does not Granger Cause LF1M 27 LF1M does not Granger Cause LS1M LS2M does not Granger Cause LF2M 27 LF2M does not Granger Cause LS2M LS3M does not Granger Cause LF3M 27 LF3M does not Granger Cause LS3M F-Statistic Prob. 1.53965 0.23318 2.22508 0.10222 0.45549 0.80332 0.47637 0.78861 1.37934 0.28373 1.10208 0.39746 1.01727 0.43984 1.39098 0.27972 2008-2014 Null Hypothesis: LS1W does not Granger Cause LF1W Obs 41 LF1W does not Granger Cause LS1W LS1M does not Granger Cause LF1M 47 LF1M does not Granger Cause LS1M LS2M does not Granger Cause LF2M 42 LF2M does not Granger Cause LS2M LS3M does not Granger Cause LF3M LF3M does not Granger Cause LS3M 48 F-Statistic Prob. 1.45847 0.23265 1.60072 0.1902 1.35083 0.2657 12.398 0.0000 0.45896 0.80356 1.02695 0.41905 0.59906 0.70084 0.70143 0.62588 102 IRACST- International Journal of Research in Management & Technology (IJRMT), ISSN: 2249-9563 Vol. 4, No.2, April 2014 REFERENCES [1] Ali, Jabir and Gupta, K.B., (2011), “Efficiency in agricultural commodity futures markets in India: Evidence from cointegration and causality tests.” Agricultural Finance Review. 71 (2). [2] Beck, S.E. (1994), “Cointegration and market efficiency in futures markets.” Applied Economics, 26(3) http://dx.doi.org/10.1080/00036849400000006 commodities [3] Bhattacharya, P and Harminder, S. (2012), “Efficiency and unbiasedness in the Indian stock index futures market.” Review of Futures Markets, 18(3). [4] Bose, S. (2007), “Commodity Futures Market in India; A Study of Trends in the National Multi Commodity Indices.”, Money and Finance, ICRA Bulletin, 3 (3). [5] Chowdhury, A.R. (1991), “Futures Market Efficiency: Evidence from Cointegration tests,” Journal of Futures Markets, 11(5). http://dx.doi.org/10.1002/fut.3990110506 [6] Crowder, W. J. and Hamed, A. (1993), “A cointegration test for oil futures market efficiency”. Journal of Futures Markets, 13(8). http://dx.doi.org/10.1002/fut.3990130810 [7] Fama, Eugene (1970). "Efficient Capital Markets: A Review of Theory and Empirical Work". Journal of Finance 25 (2): 383– 417. http://dx.doi.org/10.2307/2325486 [8] Garbade, K.D. and Silber, W.L. (1983), “Dominant Satellite Relationship between Live Cattle Cash and Futures Markets”, The Journal of Futures Markets, Vol. 10 (2). [9] Kaur, G. and Rao, D. (2010), “Efficiency of Indian Commodities Market: A Study of Agricultural Commodity Derivatives Traded on NCDEX. http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1600687 [10] Kawaller, I.G., Koch, P.D. and Koch, T.W. (1987), “The Temporal Price Relationship between S&P 500 Futures and the S&P 500 Index”, Journal of Finance, 42. [11] Kumar, B., Singh P. and A. Pandey (2008), “Hedging Effectiveness of Constant and Time Varying Hedge Ratio in Indian Stock and Commodity Futures Markets.” http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1206555 [12] Lai, K.S. and Lai, M. (1991), “A cointegration test for market efficiency.” Journal of Futures Markets, 11(5). http://dx.doi.org/10.1002/fut.3990110505 [13] Lokare, S. M., 2007, Commodity Derivatives and Price Risk Management: An Empirical Anecdote from India, Reserve Bank of India, Occasional Papers, Vol. 28, No 2. [14] McKenzie, A. M. and Holt, M.T. (2002), “Market Efficiency in Agricultural Futures Markets.” Applied Economics, 34(12). http://dx.doi.org/10.1080/00036840110102761 [15] Sahadevan, K. G. (2002), “Price Discovery, Return and Market Conditions: Evidence from Commodity Futures Markets.” The ICFAI Journal of Applied Finance, 8(5). [16] Sahi, G. and Raizada, G. (2006), “Commodity Futures Market Efficiency in India and Effect on Inflation.” http://papers.ssrn.com/sol3/papers.cfm?abstract_id=949161. [17] Sahoo, P. and Kumar, R. (2009), “Efficiency and Futures Trading Price nexus in Indian Commodity Futures Markets.” Global Business Review, 10(2). http://dx.doi.org/10.1177/097215090901000204 [18] Sehgal, S., Rajput, N. and Dua, R. K. (2012), “Price Discovery in Indian Agricultural Commodity Markets.” International Journal of Accounting and Financial Reporting, 2(2). http://dx.doi.org/10.5296/ijafr.v2i2.2224 [19] Shroeder, T.C. and Goodwin, B.K. (1991), “Price Discovery and Cointegration for Live Hogs,” Journal of Futures Markets, 11. [20] Wang Hong and Bingfan Ke,(2005), “ Efficiency Test of Agricultural Commodity Future Markets in China”, AUTHORS PROFILE Alok Kumar Sahai obtained his Post Graduate Diploma in Management from IIM Bangalore and has worked in industrial marketing for Fujifilm in MENA region for 10 years He has worked as a faculty at iLead School of Business at Indore and runs his own commodity advisory at Indore. 103