Do the Spot and Futures Markets for Commodities in India Move Together?

Vol. 4, No. 3, 2015, 150-159 Do the Spot and Futures Markets for Commodities in India Move Together? Ranajit Chakraborty 1, Rahuldeb Das 2 Abstract The objective of this paper is to study the relationship between spot and futures prices of commodities in the Indian commodity market. Few agricultural and non-agricultural commodities have been involved in analyzing the co-movements of the spot and the futures prices. Long-run and short-run cointegartions between spot and futures prices have been tested for the selected commodities. The lead/lag relationship between spot and futures prices has been examined by using Granger causality test. The study shows that spot and futures prices are cointegrated in long-run for most of the commodities. Information spillover observed form spot market to spot market for most of the commodities. Moreover, the Granger causality test shows that spot prices are the Granger cause of futures prices for all the commodities except the index MCXAGRI. Bi-directional information is also observed flow for a number of commodities. Keywords: Long-run cointegration, information spillover, Granger Causality, bi-directional information flow JEL Classification: G1, C1, C5. 1. Introduction In India, commodity prices vary widely from location to location. Also the prices vary according to their quality characteristics and preferred end-uses. At any time point there is no single unique spot price for a commodity valid throughout the country. Seasonal effect is also there on the prices of most of the agricultural commodities. So, the commodity traders face prices risk due to the variations in commodity prices over varieties, space and time. Hence, an active futures market is essential for efficient reference pricing and effective risk management. Futures contracts have been introduced in the commodity market to manage effectively the price risk associated with spot prices of a commodity. This process of reducing price risk is called hedging. The primary function of a commodity futures market is hedging. Producers, merchants, stockists and importers require hedging against price decline. Processors, manufacturers and exporters require hedging against abnormal price increase. For an effective hedging there must be a stable and predictable relationship between cash and futures price movements. Another purpose behind establishing futures markets is to contribute positively to overall price discovery in the commodity market. In a commodity market, price discovery by the futures market is very much important. The futures price of a commodity serves as a reference price for the transaction in the physical market. The factors that can influence commodity prices are too many in number and complex as well. The supply of a commodity depends on the factors such as the amount of production, availability of infrastructure, weather condition, availability of ground water, power source available, seeds quality, imports and exports. These factors are influenced by several other factors. The factors, population growth, changes in income level and exports can influence the demand of a commodity. Furthermore, commodity markets also face pressures of speculative demand. Because of the factors affecting supply and demand of a commodity, 1 Ex-Professor, Department of Business Management, University of Calcutta, 1, Reformatory Street, Kolkata 2 Assistant Professor, Department of Basic Science, Techno India College of Technology, New Town, Rajarhat, Kolkata 2015 Research Academy of Social Sciences http://www.rassweb.com 150

price discovery is of great importance for the smooth functioning of production, merchandising, stocking and manufacturing process. The essence of price discovery function of the futures market hinges on whether new information is reflected first in changed futures prices or in changed cash prices. To serve both the purposes of the futures market, effective hedging and price discovery, spot and futures market of a commodity have to be closely related. Both spot and futures market should move together. However, several evidences from different markets all over the world have shown that one market moves faster than the other. As a result a lead-lag relationship exists between the spot and futures markets. Whenever futures market leads the spot market, the information flows from futures market to spot market. Then futures market helps in the price discovery of spot markets. When spot market leads the futures market, spot market contributes to the price discovery of futures, market. This relationship between spot and futures price of the commodities has received considerable attention in the developed countries. However, in developing countries like India work examining this issue is less extensive. In this chapter the association between spot and futures prices has been investigated for some selected commodities and commodity index. To examine their co-movements the Johansen s Cointegration test has been used. The Granger causality test has been used to check the lead-lag relationship between spot and futures prices. Hence the information flow between spot and futures market and its direction has been explained. 2. Literature Review Numerous efforts have been employed to study the lead-lag relationship between spot and futures market around the globe. However, relatively little work has been done in the Indian commodity market. Garbade and Silber (1983) showed that though the future market dominates the spot market in price discovery, there are reverse information flows too. Their evidence suggests that in case of wheat, corn and orange juice 75 percent of new information incorporated first in future prices and then transmitted to the spot prices. Same observation found for gold. Price discovery for silver, oats and copper was more divided between spot and future markets. Silvapulle and Moosa (1999) examined the relationship between spot and future prices of WTI crude oil using a sample of daily data. They have got two different results for different test procedures. The linear causality test has shown that feature price leads spot price. But nonlinear causality test has shown a bi-directional effect. They concluded that both future and spot markets reacts simultaneously to new information. Thomas and Karande (2001) examined the flow of information between spot and future markets. They have chosen two different markets of the castor seed, one export oriented and other production oriented. They have concluded that future dominates the spot price. Export oriented market dominates production oriented market, except in the harvest season. McKenzie and Holt (2002) have tested market efficiency and unbiasedness for four commodities, live cattle, hogs, corn and soybean meal. They have found that the market is unbiased in the long run, although cattle, hogs and corn futures markets exhibit short run inefficiencies and price bias. Asche and Guttormsen (2002) used data from the International Petroleum Exchange (IPE) on the gas oil contract. Their result indicates that future price leads spot price. Another observation found by them was that the contracts with longer time to expiration lead the contracts with shorter time to expiration. Giot (2003) compared the incremental information content of lagged implied volatility to skewed student GARCH models for a collection of agricultural commodity (cocoa, coffee, and sugar nearby futures contract) traded on the New York Board of Trade. The results indicate that the implied volatility for option on future contracts has a high information content regarding conditional variance and VaR forecasts of the underlying future forecasts. Zapata et al (2005) finds using Granger causality test that information flow directed from futures prices to cash prices for world sugar on the New York Exchange. He finds that a shock in the future price innovation generates a quick and positive response in futures and cash prices. Bagchi (2007) using CNX Midcap 200 found that there has been a relatively high impact of the volatility computed on high-low-closing prices and the lowest impact is found for volatility computed on high prices of the securities. The above results also confirm that the entropy of volatility is a valuable indicator for evaluating the performance of the volatility. He has analysed the relative impact of the volatility 151

R. Chakraborty & R. Das measure computed from the four parameters, such as closing, high and low quotes of the day as well as a combination of these three on the stock index return. Reddy and Sebastin (2008) examined the dynamic relationship between derivatives market and the underlying spot market. The study observed that price innovations appeared first in the derivatives market and were then transmitted to the equity market. They have used the concept of entropy to study the information flow between the markets. Iyer and Pillai (2010) have found that for Chickpea and copper future market dominates the spot market in the pre expiration week. For the commodities Chickpea, gold, copper and rubber the same case happens in the expiration weeks. Nickel was the only commodity where the spot market plays a dominant role. Data Data for this study has been collected from the Multi Commodity Exchange (MCX) and the National Commodity Derivative Exchange (NCDEX) of India. The period of study is from 1 st January 2004 to 31 December 2012. Spot and futures prices of the agricultural-commodity index MCXAGRI is taken from the website of MCX. Daily spot and futures prices of Barley, Chickpea, Chilli, Cumin, Maize, Mustard Seed, Pepper, Brent Crude Oil and Gold obtained from the website of NCDEX. Daily closing prices are considered for analysis in this chapter. 3. Methodology Johansen s Long-Run Cointegration Engle and Granger (1987) suggested in their cointegration theory that two non-stationary series having a same stochastic trend, tend to move together over the long-run. In the short-run, deviation from long-run equilibrium may occur. To explore the relationship between commodity futures prices and spot prices, cointegration is tested. The Johansen s full information multivariate cointegrating procedure is used to perform the cointegration analysis. The Johansen s Cointegration test is conducted through the k th order vector error correction model (VECM) represented by the equation k 1 X t = X t 1 + i=1 Г i X t 1 + γ + ε t Equation 1 Where, X t is (n 1) vector to be tested for cointegration, X t = X t X t 1, γ is the vector of deterministic term, П and Ѓ are coefficient matrices. γ is the intercept term and ε t represents the error term. The rank of the coefficient matrix П indicates the existence of cointegration between endogenous variable. If the rank of the matrix П is zero, no cointegration exists between the variables. If П is a full rank matrix, then variables in vector X t are stationary. If the rank lies between zero and n, cointegration exists between the variables under study. The lag length k is selected for a minimum value of Akaike Information Criterion (AIC). In this study, to test the cointegration between futures prices and spot prices the value of n is 2. Two separate tests are performed with null hypotheses rank = 0 and rank = 1. If rank = 0 is rejected and r = 1 is not rejected, the conclusion will be that the two series are cointegrated. To test the long-run relationship likelihood ratio test is used. The trace statistics, tests the null hypothesis of at most r cointegrating vectors against a general alternative hypothesis of more than r cointegrating vectors. The trace statistics is given by n trace = T ln(1 λ i=r+1 i ) Equation 2 where T is the number of observations and λ i is the eigen value. Vector Error Correction Model (VECM) for Short-run Cointegration The short-run integration or information spillover between commodity futures prices and spot prices is investigated through Vector Error Correction Model (VECM). VECM specifications provide a long-run equilibrium error correction in prices in the conditional mean equations. This approach is used to model the short run relationship of cointegrated variables. The VECM specification for commodity futures prices and the factors is represented as 152

P Spot,t = C Spot + α Spot P Spot,t 1 + α Future P Future,t 1 + k k i=2 β Spot,i P Spot,t i + j=2 β Future,j P Future,t j + ε Spot,t Equation 3 Where P Spot,t is the log of spot price of the Indian commodity market and P Future,t is the log of futures prices. P Spot,t i and P Future,t j are the first differenced value of spot and futures prices respectively. C Spot,t and ε Spot,t are the intercept term and residual term of the model respectively. The error correction term is given by the expression α Spot P Spot,t 1 + α Future P Future,t 1. The short-run parameter estimates, β Spot,i and β Future,j, measure the short-run integration or return spillover. After checking the cointegration between spot and futures prices of the commodities and commodity index, the returns have been calculated using the following formula R t = ln ( Y t Y t 1 ) 100 Equation 4 Where, R t is the return of a commodity at time t, Y t and Y t-1 are the prices of a commodity at time t and t-1 respectively. Augmented Dickey-Fuller Test (ADF) Stationarity of all return series have been checked by Augmented Dickey-Fuller (ADF) test. The equation of ADF test is given below: p Y t = µ + bt + βy t-1 j=1 α j Y t j + є t Equation 5 Here Y t is a time series which is to be tested for stationarity, Y t is the first order differenced series based on Y t, i.e., Y t = Y t - Y t-1, µ is the intercept term, b is the coefficient of a time trend, β is the coefficient of the lagged value of Y t, α j, j = 1(1)p, are the coefficient of the autoregressive process of Y t, and є t is white noise residual. The hypothesis under consideration is H 0: b = β = 0. Therefore, an F-test is performed on the hypothesis H 0. If β = 0 then there is a unit root. If Unit Root is present in the series, then it is non-stationary. Granger Causality Test To test the Granger Causality running from the variable X t to the variable Y t, the following equations have been used: p Y t = 0 + i=1 β i Y t i + ε 0t Equation 6 p q Y t = 1 + i=1 β i Y t i + j=1 γ j X t j + ε 1t Equation 7 where ε 0t and ε 1t are white noise residuals. 0 and 1 are the intercept terms. β i s are the coefficients of the lagged values of the variable Y t. γ j s are the coefficients of the lagged values of the variable X t. The lag length p can be determined from the results of cross correlation. The Granger test based on equations 7 and 8 is equivalent to testing the following null hypothesis: H 0: γ 1 = γ 2 = γ 3 = = γ q 153

The test statistic for this hypothesis is R. Chakraborty & R. Das F = (SSE 1 SSE 2 )/q SSE 1 /(N p q 1) where SSE 1 and SSE 2 are the sum of squared errors from least squared regression on equation 7 and 8. 4. Empirical Analysis Spot and futures prices of a market are supposed to move together. Flow of information from one market to other help in price discovery of both the markets.in the Indian commodity market this fact is tested in this chapter by inspecting the cointegration between the spot and the futures market. Johansen s Cointegration Rank Test is used in this purpose. The existence of cointegration can be tested by examining the rank of coefficient matrix П in equation 1. If the rank of the matrix П is one, cointegration exists between the variables under investigation. Table 1 represents the Johansen s cointegration rank test results for the commodities considered in this study. For MCXAGRI, Chickpea and Gold no cointegration exist between spot and futures price. For all other commodities under study, spot and futures prices are found to be cointegrated. In these cases, null hypothesis for the rank r = 0 is rejected at the 5% level and r =1 is accepted at the same level of significance. So, the long-run equilibrium relationship exists between futures prices and spot prices of all the commodities except MCXAGRI, Chickpea and Gold. Table 1: Johansen s Co-integration Rank Test Commodity/ Index Lag H 0: r = 0 H 0: r = 1 MCXAGRI 2 10.8767 2.2536 Barley 2 18.63088* 0.2672 Chickpea 2 42.0308* 6.6946* Chilli 2 107.6968* 2.6572 Cumin 2 323.8863* 1.4311 Maize 2 180.2758* 2.4968 Mustard Seed 2 19.8858* 3.0593 Pepper 2 97.8043* 0.1576 Brent Crude Oil 2 626.4609 * 2.7280 Gold 2 749.5024* 5.7888* To test the short-run co-movements, the Vector Error Correction Model (VECM) is used. Table 2 and Table 3 represent the results of VECM for the commodities under study. Table 2 represents the information spill over form futures market to spot market, whereas, Table 3 depicts the spill over effect form spot market to the futures market. In the Table 2 the short-run coefficients β Future, 1 measures the information spillover from futures market to spot market. β Spot, 1 in Table 3 measures the information spillover from the spot market to the futures market. From table 2 it has been found that significant short run effect of futures price on the spot price is present for the index MCXAGRI. For the commodities Pepper and Brent crude Oil, a similar effect is also observed. In the Table 3 it has been found that for Barley, Chilli, Cumin, Maize, Mustard Seed, Pepper and Gold, the short run effect of spot prices is there of futures prices. For all of these cases the β Spot, 1 coefficient is significant at 1% or 5% level. So, the results suggest that there is information spill over between spot to futures prices for all the commodities under study except Chickpea. But the direction of the spillover effect depends on the type of commodity. For some of the commodities like Pepper the short run effect from both the markets is observed. Therefore, in the Indian commodity market, spot and futures prices are cointegrated in both long and short run in most cases. 154

After testing the long and short run cointegration between the spot and futures prices of the commodities, the causal relationship between them is checked. However, to check the causal relationship between the spot and futures prices the data series are needed to be stationary. In this regard, returns have been calculated for both spot and futures prices of commodities using the equation 4.To test stationarity of the return series ADF test has been applied. From Table 4 it has been found that the return series of spot and futures for all the commodities are stationary. In all the cases p values are significantly low and hence the null hypothesis of existence of unit root is rejected. Table 2: Parameter Estimation of the VECM to Test the Information Spillover from Futures Market to Spot Market Commodity/ Index C α Spot α Future β Spot,1 β Futre,1 MCXAGRI 0.93398 0.00123-0.00264 0.04871-0.08505* Barley 0.38295-0.0117 0.01172 0.03794 0.03521 Chickpea -1.94084* -0.00496 0.00495 0.00123-0.01949 Chilli -1.16929 0.0146-0.01485 0.25487** 0.00975 Cumin -14.44387** 0.05881-0.05897 0.12722** -0.00972 Maize 0.88278** 0.01918-0.02014 0.11291** -0.00944 Mustard Seed -0.3938-0.00155 0.00157 0.05819 0.00894 Pepper 17.85426** 0.04861-0.0492 0.20521** -0.04686** Brent Crude Oil 2.33932-0.27657 0.27466-0.25424** 0.33997** Gold 26.59091** -0.39252 0.38998 0.08427-0.04285 Note: Significance codes: `*` indicates 5% level and `**`indicates 1% level. Table 3: Parameter Estimation of the VECM to Test the Information Spillover from Spot Market to Futures Market Commodity/ Index C α Future α Spot β Future,1 β Spot,1 MCXAGRI 3.83461 0.00321-0.00688 0.0348 0.0412 Barley 0.28126* 0.00271-0.00271 0.03451 0.13742** Chickpea 1.06869 0.01848-0.01845 0.11763** -0.00523 Chili -2.32536 0.0775-0.07886 0.26975** 0.0591* Cumin -16.13501** 0.08152-0.08175 1.06286** -0.18042** Maize 3.07696** 0.05378-0.05647 0.10938* 0.13435** Mustard Seed -0.25707 0.01315-0.01334 0.14077** 0.12373** Pepper 42.63264** 0.07499-0.07591 1.13789** 0.3373** Brent Crude Oil -2.89857 0.08-0.07945-0.01334 0.02746 Gold -41.77451** 0.27285-0.27109 0.23028** -0.14415** Note: Significance codes: `*` indicates 5% level and `**`indicates 1% level. Spot and futures markets are supposed to move together and help each other in price discovery. But, it has been observed by several researchers that in this causal relationship one market leads other market. In order to estimate this lead/lag relationship between the spot and futures market Granger causality test is used. That means which series affects the other one first and which series is the conclusion of which series. If the spot price is the Granger cause of futures price than spot market leads the futures market and If the futures price is the Granger cause of spot price than the futures market leads the spot market. Lag length 5 has been used to test the causality in both directions since for lag length greater than 5, no significant causality is observed for the selected commodities. Table 5 and Table 6 represent the result of Granger causality Test for the commodities under study. In each of these tables, for a commodity, causality in the direction futures price to spot price and the spot price to futures price is depicted side by side. 155

R. Chakraborty & R. Das Table 4: ADF Test for Stationarity Commodity/ Index Spot Future ADF Stat p-value ADF Stat p-value MCXAGRI -0.464 <.0001-0.427 <.0001 Barley -38.63 <.0001-40.55 <.0001 Chickpea 0.113 <.0001 0.015 <.0001 Chili 0.043 <.0001 0.196 <.0001 Cumin 0.059 <.0001 0.431 <.0001 Maize -0.419 <.0001 0.125 <.0001 Mustard Seed 0.005 <.0001-0.063 <.0001 Pepper -0.039 <.0001 0.182 <.0001 Brent Crude Oil -0.120 <.0001-0.031 <.0001 Gold 0.020 <.0001 0.004 <.0001 Commodity MCXAGRI Barley Chickpea Chili Cumin Table 5: Granger Causality Test between Spot and Futures Prices Lag Spot ~ Future Future ~ Spot F-stat Pr(>F) F-stat Pr(>F) 1 8.4898 0.003627** 1.9548 0.1623 2 5.0753 0.006363** 0.9165 0.4002 3 3.4918 0.01517* 0.6889 0.5589 4 2.7826 0.02552* 1.2738 0.2782 5 2.4051 0.0349* 0.9078 0.475 1 0.8071 0.3691 34.944 0.000000** 2 0.5627 0.5698 18.277 0.000000** 3 0.6614 0.5758 12.478 0.000000** 4 1.2593 0.2839 9.7286 0.000000** 5 1.1391 0.3375 7.7699 0.000000** 1 0.0784 0.779500 78.366 0.000000** 2 0.4044 0.667500 39.89 0.000000** 3 0.6903 0.557900 26.942 0.000000** 4 0.526 0.7167 20.71 0.000000** 5 0.6905 0.630700 16.936 0.000000** 1 0.0657 0.7978 65.482 0.000000** 2 4.0895 0.01694* 120.49 0.000000** 3 3.9255 0.008351** 86.324 0.000000** 4 3.2561 0.01139* 71.091 0.000000** 5 2.7839 0.01647* 56.904 0.000000** 1 10.762 0.001052** 567.17 0.000000** 2 14.458 0.000001** 292.59 0.000000** 3 15.901 0.000000** 202.35 0.000000** 4 12.633 0.000000** 154.28 0.000000** 5 10.44 0.000000** 125.73 0.000000** For the index MCXAGRI, the futures prices are Granger cause of Spot prices. Form lag 1 to lag 5 the p- values are significantly low. The null hypothesis of having no influence of future price on spot price is rejected at 1% and 5% level of significance. But, spot price can t be considered as a Granger cause of futures price in this case. The evidence suggests that movements in the index or spot prices do not provide any 156

information about the upcoming futures prices. So, in these cases, futures markets lead the spot market. The information transmitted from future market to spot market. For Barley, Chickpea and Mustard Seed the spot price is a Granger cause of futures price. The p-values are significantly low for all the lags up to 5 for each of these cases. Therefore, spot market leads the futures market in these cases. Futures market does not provide information to the spot market for these commodities. Chilli, Cumin, Pepper and Gold show a different scenario of causality. For all of these commodities, both spot and futures markets are Granger cause of one another. p-values are significantly low for both the tests. So, a bi-directional information flow is observed for these commodities. For Maize and Brent Crude Oil, the result shows that futures price provides information to the spot price. Spot price also provides information to the futures price at lag 4 and lag 5 in case of Maize and at lag 5 for Brent Crude Oil. Though bi-directional information flow exists for Maize but the flow form futures price to spot price is stronger. Table 6: Granger Causality Test between Spot and Futures Prices Commodity Lag Spot ~ Future Future ~ Spot F-stat Pr(>F) F-stat Pr(>F) 1 19.906 0.000000*** 2.8898 0.08928 2 12.019 0.000000** 1.9721 0.139400 Maize 3 9.721 0.000002** 2.4843 0.059070 4 7.8415 0.000003** 3.8348 0.004130** 5 8.6075 0.000000** 3.0546 0.009421** 1 0.2787 0.597600 15.555 0.000082** 2 0.29 0.748300 8.0136 0.000339** Mustard Seed 3 0.3619 0.780500 5.6676 0.000724** 4 0.5951 0.666200 4.6511 0.000966** 5 0.5528 0.736300 3.7455 0.002215** 1 27.195 0.000000** 1104.3 0.000000** 2 32.251 0.000000** 579.89 0.000000** Pepper 3 24.588 0.000000** 391.1 0.000000** 4 19.038 0.000000** 291.46 0.000000** 5 16.84 0.000000** 234.46 0.000000** 1 245.31 0.000000** 1.1933 0.2748 2 145.86 0.000000** 1.3542 0.2584 Brent Crude 3 97.742 0.000000** 1.6791 0.1695 Oil 4 74.118 0.000000** 2.2089 0.06577 5 61.955 0.000000** 4.4093 0.00053** 1 35.964 0.000000** 124.98 0.000000** 2 31.408 0.000000** 63.751 0.000000** Gold 3 24.617 0.000000** 45.758 0.000000** 4 21.915 0.000000** 34.327 0.000000** 5 18.164 0.000000** 27.667 0.000000** From the Granger causality test different patterns of causality have been found in the Indian commodity market. For some of the commodities futures markets transmit information to cash markets. That means the futures market leads the spot market. For few other commodity spot markets disseminate information to the futures markets. Here, spot market leads the futures market. So, future market is having predictability for the spot market. Moreover, there are some commodities for which a bi-directional information flow exits. Both spot and futures market provide information to one another. Therefore, in the Indian commodity market the pattern of causality is commodity specific. Both spot and futures market can play a dominant role depending on the type of commodity. 157

R. Chakraborty & R. Das In the Indian commodity market, for the Index MCXAGRI, the futures price influences the spot price. This fact has been confirmed by the short-run cointegration test using VECM and Granger causality test. That means the information is disseminated from futures market to spot market in this case. However, no long-run cointegration is observed between these prices. For the commodity Barley and Mustard Seed, longrun cointegration is observed between spot and futures prices. In these cases, both VECM and Granger causality show similar results. Information spillover is observed form spot market to futures market and spot price act as a Granger cause of futures price. Chickpea spot prices have been found to be the Granger cause of futures prices. So, in this case, spot market leads the futures market. For, Chilli, Cumin and Maize long run cointegration exists between spot and futures price. Information spillover has been observed form spot market to the market. However, for all of these commodities, bi-directional causality is observed between spot and futures price. Both the markets provide information to each other. Pepper is the only commodity for which long-run cointegration, bidirectional information spillover and bi-directional causality are observed between spot and futures prices. Spot and futures prices of Brent crude Oil are cointegrated in long-run. The information spillover effect is observed form futures market to spot market. Also, bi-directional Granger causality is found between spot and futures prices of this commodity. Gold spot and futures prices are not cointegrated on long-run. In shortrun information spillover is evident from the VECM in the direction from spot to futures market. Bidirectional causality is also observed in this case. So in the Indian commodity market, long-run cointegration exists between spot and futures prices for most of the commodities considered under this study. The results of this study indicate that spot market provides information to the futures market in most cases. That means for most of the commodities spot market is more active than the futures market. But, the reason behind the establishment of commodity derivatives market was to provide information to the spot market and to help the spot market in price discovery. So, the finding of this chapter may provide some helpful information in the process of decision making of the investors and policy makers of the Indian commodity market. 5. Conclusion In this chapter the causality running in the Indian commodity market between spot and futures price has been examined. Including few agricultural and non-agricultural commodities, the co-movements of the spot and futures prices also has been examined. For most of the commodities under consideration spot and futures prices are cointegrated. A long run cointegration has been found for those commodities. However, for some of the commodities sufficient evidence of long run cointegration has not been found. For most of the commodities information spillover is observed in the direction from spot market to futures market. The bidirectional causality between spot and futures prices has been found for a significant number of commodities. The results of this study indicate that in the Indian commodity market spot market is more active compared to the futures market. Though bi-directional causality has been observed for a number of cases, but both short-run cointegration and Granger causality test show that information flow is more prominent from the spot market to the futures market. References Asche. F. and A.G. Guttormemesen, 2002. Lead lag relationship between futures and spot prices. Discussion Paper D-15/2001, Department of Economics and Social Science, Agricultural University of Norway. Available at http://idtjeneste.nb.no/urn:nbn:no-bibsys_brage_22418. Date of access 15.08.20013. D Bagchi, 2007. An analysis of relative information content of volatility measures of stock index in India. ICFAI Journal of Derivatives Markets 4(4):-43. Garbede. K.D. and W.L. Silber, 1983. Price movement and price discovery in futures and cash market. Review of Economics and Statistics 64: 189-197. 158

P. Giot, 2003. The information content of implied volatility in agricultural commodity markets. Journal of Futures Markets 23 (5):441-454. Granger. C. and J. Lin, 1994. Using the mutual information coefficient to identify lags in nonlinear models. Journal of Time Series Analysis 15(4): 371-384. Iyer. V and A. Pillai, 2010. Price discovery and convergence in the Indian commodities market. Indian Growth and Development Review 3(1): 53 61. McKenzie. A.M. and M.T. Holt, 2002. Economics 34: 1519-32. Market efficiency in agricultural futures markets. Applied Reddy. Y. V. and A. Sebastin, 2008. Interaction between equity and derivatives markets in India: An entropy approach. ICFAI Journal of Derivatives Markets 5(1): 18-32. Silvapulle. P. and I. A. Moosa, 1999. The relationship between the spot and futures prices: Evidence from the crude oil market. Journal of Futures Markets 19: 175-193. Thomas. S. and K. Karande, 2001. Price discovery across multiple markets. Technical report, IGIDR, Bombay, India. Available at http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.28.5475&rep=rep1&type=pdf. Date of Access 20.10.2013. Zapata. H., Fortenbery. T.R. and D. Armstrong, 2005. Price discovery in the world sugar futures and cash markets: Implications for the Dominican Republic. Staff Paper No. 469, Department of Agricultural & Applied Economics, University of Wisconsian-Madison. Available at https://www.aae.wisc.edu/pubs/sps/pdf/stpap469.pdf Date of Access 12.04.2014. 159