______________________________________ Agriculture Commodity Futures and Risk Management: Evidence from NMCE1 ______________________________________ By Tarun K Soni* *Assistant Professor at FMS- WISDOM, Banasthali Vidyapith (An MHRD A Category and NAAC A Grade Institution), Newai, Rajasthan, India.
E-mail:
[email protected]
1
Suggested Citation: Soni, T.K. (2015). Agricultural Commodity Futures and Risk Management: Evidence from NMCE, NIFM Journal of Public Financial Management,7(1),119-128
Agricultural Commodity Futures and Risk Management: Evidence from NMCE Abstract The paper aims to study the market efficiency, unbiasedness among fifteen agricultural commodities futures contracts traded at National Multi-Commodity Exchange of India ltd. (NMCE). The paper uses a two step approach by first testing long run relationship using Johansen's cointegration approach and subsequently, the dynamic OLS approach proposed by Stock and Watson (1993) was used to estimate the coefficients in the cointegration equation, followed by Wald test to test the statistical significance of each coefficient. The Wald chi-square test statistics indicate that futures markets are not efficient in predicting the future ready prices. The results also testify the fact that the futures contracts are not perfect hedge against the variations in ready prices. The results have important implications on the previous research done on the same issue which have simply tested the efficiency on the basis of cointegration results, ignoring the restrictions on cointegrating vectors may result in incorrect assessment of price discovery and risk management functions of commodity exchanges. Further, the results also urge for further reforms in the agricultural commodity futures through increasing awareness, wider participation, better infrastructure etc. so as to make futures market more efficient in the long run and perform their role of price discovery and risk management more efficiently and effectively. Keyword: Market efficiency, Commodity markets, NMCE, India JEL Classification – C14, C32, G14
INTRODUCTION In the process of economic liberalization and deregulation in Indian agriculture sector, futures trading in agricultural commodities were reintroduced in 2003. Along with liberalization of commodity futures markets, the government also felt a need to modernize the systems and structures of the existing regional exchanges. However, failure of existing regional exchanges to modernize and provide fair and transparent trading platform lead to the setting up of new modern, demutualised nation-wide multicommodity Exchanges with investment support by public and private institutions. National Multi Commodity Exchange of India Ltd. (NMCE) was the first such exchange to be granted permanent recognition by the government. NMCE was promoted by commodity relevant public institutions, viz., Central Warehousing Corporation (CWC), National Agricultural Cooperative Marketing Federation of India (NAFED), Gujarat AgroIndustries Corporation Limited (GAICL), Gujarat State Agricultural Marketing Board (GSAMB), National Institute of Agricultural Marketing (NIAM), and Neptune Overseas Limited (NOL). NMCE functions on a national scale and the basket of commodities has grown considerably since its inception including cash crops, food grains, plantations, spices, oil seeds, metals & bullion etc. NMCE is unique in many other respects. It is a zero-debt company; following widely accepted prudent accounting and auditing practices. It has robust delivery mechanism making it the most suitable for the participants in the physical commodity markets. The exchange does not compromise on its delivery provisions to attract speculative volume. Public interest rather than commercial interest guide the functioning of the Exchange.2 Since the setup of NMCE, the exchange has also established fair and transparent rule based procedures and demonstrated total commitment towards eliminating any conflicts of interest. While many economists approve the price discovery and risk transfer function of futures exchanges, some policy makers on the other hand are skeptical 2
Source: http://www.nmce.com/aboutus.aspx
about its efficiency and link them to price rise and inflation of agricultural commodities (Iyer and Pillai, 2010; Ali and Gupta, 2011) due to which the commodity markets have also faced temporary policy reversals with the ban of certain actively traded commodity contracts. Amidst of doubts and skepticism on the role of commodity futures, the aim of this study is to test the efficiency of the agricultural commodities futures contracts traded on NMCE. Although previous studies have examined price efficiency in Indian commodity markets, the study intends to extend these previous studies by examining efficiency and unbiasedness in more recent years and for a complete sample of fifteen commodities actively traded on the exchange. In the first step efficiency has been estimated through Johansen's cointegration approach. Subsequently, the dynamic OLS approach proposed by Stock and Watson (1993) is used to estimate the coefficients in the cointegration equation, followed by Wald test to test the unbiasedness of the commodity contracts. The remainder of the article is organized as follows. Section 2 presents a brief review of literature. Section 3 describes the data source and the methodology adopted in the study. Section 4 presents the empirical results. Finally, Section 5 provides a summary and concludes. LITERATURE REVIEW Several studies in the past have studied the relationship between spot and futures prices of commodities with the objective of investigating the issue of market efficiency. Examples of research on the efficiency of the Indian commodities markets include Sahadevan, 2004; Raizada and Sahi, 2006; Lokare, 2007; Bose, 2008; Kumar and Pandey, 2008; Sahoo and Kumar, 2009; Iyer and Pillai, 2010; Ali and Gupta, 2011; Sehgal, Rajput and Dua, 2012. The existence of cointegrating relationships in the Indian commodities futures market has been supported by majority of studies. However, results supporting the existence of cointegrating vectors are rare as very few studies in Indian context have tested the efficiency and unbiasedness by imposing restrictions on the cointegrating vectors. Sahadevan (2004) 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 βo= 0, β1=1. Raizada and Sahi (2006), however, tested futures and spot prices for wheat contract traded at NCDEX between July 2004 to July 2006 and obtained results rejecting βo= 0, β1=1. Further, several studies have examined whether spot and futures prices are cointegrated, using Johansen's cointegration approach and found significant cointegration in spot and futures prices of various commodities contracts traded on Indian commodity exchanges (see Lokare, 2007; Kumar, Singh and Pandey, 2008; Sahoo and Kumar, 2009; Ali and Gupta, 2011; Sehgal, Rajput and Dua, 2012.) whereas Bose (2008) obtained results against market efficiency and price discovery of Indian agricultural indices between June 2005 to September 2007. Further, Iyer and Pillai (2010) found evidence for price discovery in futures market in five out of six commodities using a two-regime threshold vector auto regression (TVAR) and a two-regime threshold autoregression model from October 2005 to March 2008. Furthermore, Ali and Gupta (2011) examined the price discovery of 12 major agricultural commodities contracts using cointegration and granger causality analysis between July 2004 to January 2007 and found significant cointegration between futures and spot prices for all the selected agricultural commodities excluding wheat and rice. From the empirical literature cited above, it is clear that the Indian literature is limited to small sample period spanning between 2 to 3 years, or to very fewer commodities traded on national exchanges (Sahadevan, 2004; Raizada and Sahi, 2006; Lokare, 2007; Bose, 2008; Kumar and Pandey, 2008; Sahoo and Kumar, 2009; Iyer and Pillai, 2010; Soni, 2012; Soni, 2013; Soni, 2014). Further, most of the studies have not tested the cointegrated vectors (βo, β1) (Lokare, 2007; Ali and Gupta, 2011; Sehgal, Rajput and Dua, 2012) which might have lead to wrong conclusions regarding efficiency and price discovery. As though cointegration between the spot price and futures price is a necessary condition for market efficiency however, if the restriction β1=1 is not rejected then there is strong evidence against the efficiency and unbiasedness of the markets (Kellard et al., 1999). The present research examines the cointegrating relationships
and then tests cointegrating vectors for unbiasedness for fifteen main agricultural commodities contracts traded on NMCE from June 2004 to September 2012. DATA AND METHODS Data To analyze the efficiency of futures market in Indian scenario we focus on fifteen main agricultural contracts namely Cardamom, Castor Seed, Castor Seed 10MT, Chickpea, Coffee REP bulk, Copra, Guar Seed, Guar Gum, Isabgol Seed, Pepper, Mustard seed, Rubber, Sacking, Soya Oil and Turmeric which are among the most traded agricultural commodity contracts on the exchange. The period of study is from 2004 to September 2012; however data period is different for different commodities due to their late introduction for trading on exchange and few agricultural commodities were banned from trading time to time. The data comprises of near month series of daily closing spot and futures prices of the sample commodities collected from NMCE website. The descriptive statistics such as mean, standard deviation, skewness etc. for spot series and future series for all the commodities are presented in Table I. Econometric methods This study empirically analyzes the weak form of efficiency for fifteen agriculture futures contracts traded on NMCE. The conventional process of testing for efficiency requires first testing the presence of cointegration and second testing that futures price at contract purchase is an unbiased predictor of the spot price at the contract termination (Chowdhury, 1991; Lai and Lai, 1991; Kellard et al., 1999). Before testing for cointegration, each individual price series should be examined to determine whether they are I (1). Augmented Dickey-Fuller (ADF) test and non-parametric Phillips-Perron (PP) approaches were employed to examine the stationarity of all the futures and spot price series. In the second step we test for the presence of cointegration among future and spot series for all fifteen commodities. Subsequently, the dynamic OLS approach proposed by Stock and Watson (1993) was used to estimate the coefficients in the cointegration equation, followed by the Wald test conducted to test the statistical significance of cointegrating vector (see Kawamato and Hamori, 2011).
Johansen Method of Cointegration The cointegration between the spot price and futures price is a necessary condition for market efficiency. It ensures that there exists a long-run equilibrium relationship between the two series (Wang and Ke, 2005). The maximum likelihood approach of Johansen and Juselius (1990) was used to establish whether there is a long-run relationship between future and spot prices of the selected commodities. Johansen’s cointegrating analysis involves estimating the following Vector Error Correction Model in reduced form k 1
ΔYt =
Y
t-1+
Y t-k +
(1)
t
i 1
Where, Y
is a vector of non-stationary variables, and Γ, Π, and λ are matrices of
t
parameters to be estimated. The rank of the matrix determines the long-run relationship and can be decomposed as = α β', where α and β contain adjustments and the cointegrating vectors respectively. Δ and t refer change and error term respectively. Johansen’s methodology requires the estimation of the vector autoregression regression (VAR) equation and the residuals are then used to compute two likelihood ratio (LR) test statistics that can be used in the determination of the unique cointegrating vectors of Xt. The first test which considers the hypothesis that the rank of Π is less than or equal to r cointegrating vectors is given by the trace test below: n
Trace = -T
ln(1 ) i
(2)
i r 1
The second test statistic is known as the maximal eigen value test which computes the null hypothesis that there are exactly r cointegrating vectors in Xt and is given by:
max = -T ln (1- r )
(3)
The distributions for these tests are not given by the usual chi-squared distributions. The asymptotic critical values for these likelihood ratio tests are calculated via numerical simulations (see Johansen and Juselius 1990; and Osterwald-Lenum 1992). To test whether the futures price at contract purchase date is an unbiased predictor of spot at contract maturity date. The cointegrating regression is generally specified as: =
+
+
(4)
Where ft-1 is the logarithm of the lagged futures price (that is, the futures price at contract purchase) and st is the logarithm of the spot price that is matched with the settlement date of the futures contract. The dynamic OLS approach proposed by Stock and Watson (1993) was used to estimate the coefficients in the cointegration equation. The unbiasedness hypothesis is that βo= 0, β1=1 and ut is white noise. If the restriction β1=1 is not rejected then there is strong evidence against the hypothesis that in long run equilibrium, the spot price is equal to the futures price plus a (possibly zero) constant. RESULTS AND DISCUSSION The descriptive statistics such as mean, standard deviation, skewness etc. for spot and future series for each commodity contract are presented in Table 1. Put Table 1 here The ADF and PP unit root tests were used to examine the stationarity of spot and futures prices. These two methods have been adopted to assess the unit root test using parametric and non-parametric approaches. Table 2 presents the result of unit root tests for major agricultural commodities by both the approaches. ADF tests suggested that the null of a unit autoregressive root, i.e. integration of order I (1), could not be rejected for all the commodities. Put Table 2 here After testing the precondition of non-stationary time series of price information, cointegration test has been carried out to determine the existence of a long-run relationship between the spot and futures prices. Table 3 presents the cointegration
results from the application of the Johansen method of reduced rank regression using the vector error correction model. The Johansen λ trace (trace statistics) and λ max (maximal eigen value), analysis indicates that null hypothesis of non-cointegration (R=0) is rejected at 0.05 level of significance for all the commodities excluding castor seed (CSSSPOT/CSSFUT) and guar seed (GUASPOT/GUAFUT). This shows that the castor seed and guar seed futures prices are not cointegrated with spot prices, indicating no long-run equilibrium relationship between the futures prices and spot prices of these two commodities. However, lack of cointegration may be caused by non-stationary components of transportation and carrying costs, including factors like the interest rate, risk premiums, convenience yields and physical storage costs (Wang and Ke, 2005). Put Table 3 here Further, the existence of cointegration between the spot and futures prices for the rest fifteen commodities contracts confirms the first necessary condition for long-term market efficiency. Subsequently, the dynamic OLS approach proposed by Stock and Watson (1993) was used to estimate the coefficients in Equation (4), followed by the Wald test conducted to test the statistical significance of each coefficient. The Wald statistics of unbiasedness β0 = 0 & β1 = 1 and market efficiency β1 = 1 are shown in Table 4. For all commodities, the restriction of unbiasedness β0 = 0 & β1 = 1 on the cointegrating vector was rejected providing evidence that future prices were not a very good predictor of spot prices. However, the unbiasedness hypothesis may be rejected with the existence of a risk premium or a transportation cost even when the market is efficient (Wang and Ke, 2005). Therefore, it is more important to test the restriction of market efficiency i.e. β1 = 1. The results of the restriction of market efficiency were also consistent with the previous restriction where the null hypothesis was rejected for all baring one commodity i.e. Soybean contract. Put Table 4a here Put Table 4b here The results obtained from a statistical analysis of the selected agricultural commodities on price discovery provides sufficient evidence that the future price of these
commodities are not efficient and unbiased predictor of the spot prices in the long run. The findings are in contrast to the earlier studies (Lokare, 2007; Bose, 2008; Sahoo and Kumar, 2009; Sen and Paul, 2010; Iyer and Pillai, 2010; Mukherjee, 2011; Ali and Gupta, 2011) which have reported markets are on the right track by performing an effective role in price discovery and risk management. SUMMARY AND CONCLUSION In the present paper we tested for long run efficiency and unbiasedness between spot and futures prices of fifteen agricultural commodities contracts traded at NMCE. In the first step efficiency has been estimated through Johansen's cointegration approach. Subsequently, the dynamic OLS approach proposed by Stock and Watson (1993) was used to estimate the coefficients in the cointegration equation, followed by the Wald test conducted to test the statistical significance of each coefficient. The significant Wald chisquare test statistics indicate that futures markets are not efficient in predicting the future ready prices. The results also testify the fact that the futures contracts are not perfect hedge against the variations in ready prices. A perfect hedge guarantees that the profit or loss on the futures contracts fully offsets the loss or profit on the physical transactions in the ready market. Any disparity between the futures price for a specific maturity contract and the ready prices in physical market on the day of the maturity of futures contract exposes the participants to basis risk. The users of futures markets face this risk because the specific physical commodity they wish to hedge does not have the same price development as that of the standardized futures contract. This result is rather expected given the fact that the exchange has thin trade volume and infrequent trading. In spite of a developed ready market in most of these commodities, futures markets do not attract traders. These conclusions have important implications on the previous research done on the same issue which have simply tested the efficiency on the basis of cointegration results, ignoring the restrictions on cointegrating vectors which may result in incorrect assessment of price discovery and risk management functions of commodity exchanges. The results also provide support to the need for further reforms in the Indian agriculture commodity futures to encourage wider participation by means of spreading
awareness about the importance of commodity exchanges to small and marginal farmers, educating the owners/ managers about how they can hedge the price risk and minimize the risk of higher input cost using derivatives. Further improving infrastructure like storage warehouses, road connectivity can also facilitate delivery system which can further improve the acceptance of these exchanges. Furthermore use of latest and user friendly technology like mobile trading, installing kiosks at strategic locations can encourage mass participation leading to better price discovery and dissemination.
References Ali, J & Gupta, KB 2011, Efficiency in agricultural commodity futures markets in India: Evidence from cointegration and causality tests, Agricultural Finance Review, 71, 2, 162-178. Beck, SE 1994, Cointegration and market efficiency in commodities futures markets, Applied Economics, 26, 3, 249-257. Bose, S 2007, Commodity Futures Market in India: A Study of Trends in the Notional Multi-Commodity Indices, Money \& Finance, ICRA Bulletin, 3, 3. Chowdhury, AR 1991, Futures market efficiency: evidence from cointegration tests, Journal of Futures Markets, 11, 5, 577-589. Iyer, V & Pillai, A 2010, Price discovery and convergence in the Indian commodities market, Indian Growth and Development Review, 3, 1, 53-61. Johansen, S & Juselius, K 1990, Maximum likelihood estimation and inference on cointegration with applications to the demand for money, Oxford Bulletin of Economics and statistics, 52, 2, 169-210. Kaur, G & Rao, D 2010, Efficiency of Indian Commodities Market: A Study of Agricultural Commodity Derivatives Traded on NCDEX, Available at SSRN 1600687. Kawamoto, K & Hamori, S 2011, Market efficiency among futures with different maturities: Evidence from the crude oil futures market, Journal of Futures Markets, 31, 5, 487-501. Kellard, N, Newbold, P, Rayner, T & Ennew, C 1999, The relative efficiency of commodity futures markets, Journal of Futures Markets, 19, 4, 413-432. Kumar, B, Singh, P & Pandey, A 2008, Hedging Effectiveness of Constant and Time Varying Hedge Ratio in Indian Stock and Commodity Futures Markets, Available at SSRN 1206555. Lai, KS & Lai, M 1991, A cointegration test for market efficiency, Journal of Futures Markets, 11, 5, 567575. Lokare, S 2007, Commodity derivatives and price risk management: an empirical anecdote from India, Reserve Bank of India Occasional Papers, 28, 2, 27-76. McKenzie, AM & Holt, MT 2002, Market efficiency in agricultural futures markets, Applied Economics, 34, 12, 1519-1532. Osterwald-Lenum, M 1992, A Note with Quantiles of the Asymptotic Distribution of the Maximum Likelihood Cointegration Rank Test Statistics, Oxford Bulletin of Economics and statistics, 54, 3, 461472. Sahadevan, K 2002, Price Discovery, Return and Market Conditions: Evidence from Commodity Futures Markets, The ICFAI Journal of Applied Finance, 8, 5, 25-39. Sahi, G & Raizada, G 2006, Commodity Futures Market Efficiency in India and Effect on Inflation, Available at SSRN 949161.
Sahoo, P & Kumar, R 2009, Efficiency and futures trading-price nexus in Indian commodity futures markets, Global Business Review, 10, 2, 187-201. Sehgal, S, Rajput, N & Dua, RK 2012, Price Discovery in Indian Agricultural Commodity Markets, International Journal of Accounting and Financial Reporting, 2, 2, Pages--34. Soni, T.K (2012). Testing for Linear and Nonlinear Granger Causality in Notional Indian Multi-Commodity Indices. ArthAnvesan, 17(1), 61-76. Soni, T.K & Singla, H.K. (2013) A Study of the Efficiency and Unbiasedness in NCDEX: A Case Study of Guar Gum, Indian Journal of Finance, 7(11), 28-37. Soni, T.K. (2014). Cointegration, linear and nonlinear causality: analysis using Indian Agriculture Futures contracts. Journal of Agribusiness in Developing and Emerging Economies, 4(2),157-71 Soni, T.K (2013).Nonlinearity in Indian Commodity Markets: Evidence from a battery of tests. International Journal of Financial Engineering and Risk Management (IJFERM), 1(1), 73-89. Soni, T.K (2013). Testing Efficiency of Guar seed Futures: Empirical Evidence from India. Romanian Economic Journal, 16(47), 211-228.
Wang, HH & Ke, B 2005, Efficiency tests of agricultural commodity futures markets in China, Australian Journal of Agricultural and Resource Economics, 49, 2, 125-141.
Table 1. Descriptive statistics for log of daily spot and futures prices Series
Mean
Std. Dev.
Skewness
Kurtosis
J-Bera
Prob.
CDMFUT
337.72
107.59
1.37
5.51
500.14
0.00
CDMSPOT
322.15
97.53
1.56
6.75
860.38
0.00
CFRFUT
8483.09
2257.46
0.65
3.01
102.00
0.00
CFRSPOT
8384.63
2242.13
0.67
2.96
109.29
0.00
CHNFUT
2551.82
519.23
1.14
3.79
324.42
0.00
CHNSPOT
2543.13
518.61
1.32
4.16
462.18
0.00
CPSFUT
4054.52
797.56
1.47
5.35
625.77
0.00
CPSSPOT
4008.41
802.52
1.44
5.18
577.74
0.00
CSSFUT
335.33
32.34
-0.15
1.94
5.07
0.08
CSSSPOT
326.56
27.65
0.10
2.38
1.79
0.41
CSTFUT
2956.50
1001.90
0.44
2.54
60.59
0.00
CSTSPOT
2947.78
1005.77
0.42
2.60
52.55
0.00
GUAFUT
1392.52
315.59
0.16
1.20
39.63
0.00
GUASPOT
1386.98
325.60
0.13
1.14
41.80
0.00
GUGFUT
5597.44
1349.84
2.08
6.94
520.34
0.00
GUGSPOT
5636.77
1325.66
2.19
7.55
634.26
0.00
GUTFUT
2804.62
2455.35
4.50
26.19
33522.57
0.00
GUTSPOT
2781.70
2482.91
4.52
26.30
33875.10
0.00
ISBFUT
5310.75
575.37
0.06
2.89
1.60
0.45
ISBSPOT
5243.39
561.13
-0.01
2.82
1.77
0.41
PEPFUT
12649.49
6200.11
1.63
6.73
2255.17
0.00
PEPSPOT
12727.36
6199.97
1.58
6.43
2008.00
0.00
RBRFUT
11879.09
5765.24
0.63
1.93
275.96
0.00
RBRSPOT
11833.35
5763.07
0.63
1.93
277.28
0.00
RPSFUT
506.62
79.38
0.58
3.94
127.08
0.00
RPSSPOT
504.81
77.54
0.61
3.90
131.21
0.00
SCKFUT
2865.62
583.82
-0.62
2.15
131.94
0.00
SCKSPOT
2837.80
586.97
-0.61
2.17
127.94
0.00
SYOFUT
429.75
28.91
-0.62
3.26
32.18
0.00
SYOSPOT
427.71
33.72
-0.51
2.63
23.86
0.00
SYTFUT
558.38
109.13
0.56
1.97
103.70
0.00
SYTSPOT
556.59
109.00
0.53
1.90
105.90
0.00
TRMFUT
10354.16
3782.47
-0.38
1.91
43.80
0.00
TRMSPOT
10871.24
4111.48
-0.39
1.86
46.68
0.00
Table 2. Unit root test on spot and futures prices of selected agricultural commodities Augmented Dickey-Fuller (ADF)
Phillips-Perron (PP)
Levels
First Differences
Levels
First Differences
CDMFUT
-1.583 (0.491)
-27.682 (0.000)*
-1.489 (0.538)
-27.648 (0.000)*
CDMSPOT
0.484 (0.984)
-26.962 (0.000)*
0.730 (0.990)
-26.870 (0.000)*
CFRFUT
0.650 (0.988)
-30.688 (0.000)*
0.076 (0.964)
-31.010 (0.000)*
CFRSPOT
0.527 (0.985)
-34.554 (0.000)*
0.020 (0.960)
-35.334 (0.000)*
CHNFUT
0.224 (0.9736)
-38.813 (0.000)*
-0.015 (0.957)
-38.618 (0.000)*
CHNSPOT
0.815 (0.9919)
-37.11 (0.000)*
0.653 (0.988)
-37.060 (0.000)*
CPSFUT
-1.874 (0.3443)
-29.346 (0.000)*
-1.857 (0.352)
-29.192 (0.000)*
CPSSPOT
-1.722 (0.4200)
-28.107 (0.000)*
-1.817 (0.372)
-27.912 (0.000)*
CSSFUT
-1.398 (0.5832)
-7.857 (0.000)*
-1.727 (0.417)
-7.813 (0.000)*
CSSSPOT
-1.188 (0.6786)
-6.187 (0.000)*
-1.780 (0.390)
-6.176 (0.000)*
CSTFUT
-1.388 (0.5881)
-34.890 (0.000)*
-1.537 (0.515)
-35.090 (0.000)*
CSTSPOT
-1.446 (0.5603)
-36.517 (0.000)*
-1.460 (0.553)
-36.512 (0.000)*
GUAFUT
-0.361 (0.9164)
-17.509 (0.000)*
-0.210 (0.937)
-17.617 (0.000)*
GUASPOT
-0.269 (0.9297)
-17.776 (0.000)*
-0.190 (0.939)
-17.799 (0.000)*
GUGFUT
1.175 (0.9958)
-21.183 (0.000)*
1.617 (0.997)
-21.186 (0.000)*
GUGSPOT
1.357 (0.9969)
-19.358 (0.000)*
1.756 (0.998)
-19.365 (0.000)*
GUTFUT
27.791 (1.000)
-19.922 (0.000)*
20.328 (1.000)
-21.842 (0.000)*
GUTSPOT
9.828 (1.000)
-40.608 (0.000)*
14.967 (1.000)
-40.417 (0.000)*
ISBFUT
1.357 (0.9969)
-31.321 (0.000)*
1.357 (0.9969)
-31.255 (0.000)*
ISBSPOT
-1.982 (0.2944)
-37.833 (0.000)*
-1.982 (0.2944)
-37.789 (0.000)*
PEPFUT
-1.338 (0.6116)
-36.795 (0.000)*
-1.422 (0.571)
-36.889 (0.000)*
PEPSPOT
-1.104 (0.7135)
-35.348 (0.000)*
-1.269 (0.643)
-36.083 (0.000)*
RBRFUT
-2.251 (0.1882)
-36.912 (0.000)*
-2.283 (0.177)
-36.909 (0.000)*
RBRSPOT
-1.982 (0.2944)
-34.181 (0.000)*
-2.174 (0.215)
-34.230 (0.000)*
RPSFUT
-2.027 (0.2748)
-36.164 (0.000)*
-2.132 (0.231)
-36.167 (0.000)*
RPSSPOT
-1.153 (0.6935)
-35.973 (0.000)*
-1.285 (0.636)
-35.990 (0.000)*
SCKFUT
-1.939 (0.3139)
-36.765 (0.000)*
-1.923 (0.321)
-36.971 (0.000)*
SCKSPOT
-2.295 (0.1737)
-33.835 (0.000)*
-2.187 (0.211)
-34.232 (0.000)*
SYOFUT
-2.174(0.2158)
-21.506 (0.000)*
-2.155 (0.222)
-21.512 (0.000)*
SYOSPOT
-1.719 (0.4215)
-21.470 (0.000)*
-1.831 (0.365)
-21.512 (0.000)*
SYTFUT
-0.910 (0.7848)
-40.488 (0.000)*
-0.465 (0.898)
-41.035 (0.000)*
SYTSPOT
-0.818 (0.8138)
-42.440 (0.000)*
-0.323 (0.922)
-43.073 (0.000)*
TRMFUT
-2.131 (0.2324)
-20.940 (0.000)*
-2.131 (0.232)
-20.991 (0.000)*
TRMSPOT
-2.297 (0.1728)
-17.386 (0.000)*
-2.193 (0.208)
-17.720 (0.000)*
Notes: Significant at:* 0.01and ** 0.05 levels; values in parentheses indicate MacKinnon (1996) p-values
Table 3. Johansen’s cointegration tests statistics for selected agricultural commodities Commodity
CDMFUT
Lag
4
Trace statistics
Max-Eigen Statistics
H o: R
λ trace
Prob.
λ max
Prob.
0
59.498
0.004*
0.065
0.00*
1
0.678
0.463
0.008
0.410
0
43.654
0.000*
43.654
0.000*
1
0.072
0.788
0.072
0.788
0
38.126
0.000*
36.787
0.000*
1
1.338
0.247
1.338
0.247
0
91.879
0.000*
88.643
0.000*
1
3.235
0.072
3.235
0.072
0
8.026
0.462
5.822
0.636
1
2.203
0.137
2.203
0.137
0
43.663
0.000*
41.759
0.000*
1
1.903
0.167
1.903
0.167
0
6.896
0.589
6.738
0.520
1
0.157
0.691
0.157
0.691
0
22.955
0.003*
21.882
0.002*
1
1.073
0.300
1.073
0.300
0
256.729
0.000*
172.784
0.000*
1
83.945
0.000*
83.945
0.000*
0
72.281
0.000*
63.248
0.000*
1
9.0322
0.000*
9.0322
0.002*
0
134.921
0.000*
132.237
0.000*
1
2.684
0.101
2.684
0.101
0
245.966
0.000*
244.887
0.000
Comment Rank=1 reject noncointegration
CDMSPOT CFRFUT
3
Rank=1 reject noncointegration
CFRSPOT CHNFUT
4
Rank=1 reject noncointegration
CHNSPOT CPSFUT
3
Rank=1 reject noncointegration
CPSSPOT CSSFUT
3
Rank=0 Accept noncointegration
CSSSPOT CSTFUT
4
Rank=1 reject noncointegration
CSTSPOT GUAFUT
4
Rank=0 Accept noncointegration
GUASPOT GUGFUT
4
Rank=1 reject noncointegration
GUGSPOT GUTFUT
4
Rank=1 reject noncointegration
GUTSPOT ISBFUT
4
Rank=1 reject noncointegration
ISBSPOT PEPFUT
3
Rank=1 reject noncointegration
PEPSPOT RBRFUT
3
Rank=1 reject non-
RBRSPOT RPSFUT
4
1
1.079
0.298
1.079
0.298
cointegration
0
48.743
0.000*
46.335
0.000*
Rank=1 reject non-
1
2.407
0.120
2.407
0.120
0
152.281
0.001*
147.322
0.000*
1
4.958
0.026*
4.958
0.026*
0
22.126
0.004*
17.530
0.014*
1
4.596
0.032*
4.596
0.032*
0
33.707
0.00*
33.679
0.000*
1
0.035
0.85
0.035
0.851
0
23.710
0.002*
19.217
0.007*
RPSSPOT
cointegration
SCKFUT
4
SCKSPOT
Rank=1 reject noncointegration
SYOFUT
3
Rank=1 reject noncointegration
SYOSPOT SYTFUT
4
Rank=1 reject noncointegration
SYTSPOT TRMFUT
4
TRMSPOT
Rank=1 reject noncointegration
1
4.492
0.034
4.492
0.034
*
*
Notes: Significant at:* 0.01and ** 0.05 levels
Table 4a. Testing Restrictions in Cointegrating Regression Variables
Cardamom
Castor
Castor Seed
Seed
10MT
Chickpea
Coffee
Copra
Guar Seed
Guar Gum
REP bulk
Guar Seed 10MT
β0
0.330
.0500
0.091
0.187
-0.717
0.161
0.376
-0.251
0.041
Β1
0.948
.995
0.988
0.978
1.128
0.980
0.948
1.028
0.996
Wald
0.0001*
0.000*
0.0001*
0.0001*
0.000*
0.000*
0.0004*
0.000*
0.000*
.000*
0.000*
0.000*
0.000*
0.000*
0.000*
0.000*
0.000*
0.000*
Test β1=1 Wald Test β0=0 and β1=1
Significant at:* 0.01and ** 0.05 levels
Table 4b. Testing Restrictions in Cointegrating Regression Variables
Isabgol Seed
Pepper
Mustard
Rubber
Sacking
Soya Oil
Soya Oil 10MT
Turmeric
seed β0
0.021
-0.002
0.038
0.112
0.114
1.052
0.061
0.413
Β1
0.998
0.999
0.996
0.982
0.986
0.827
0.990
0.950
Wald Test
0.0001*
0.000*
0.000*
0.000*
0.000*
0.6794
0.0000*
0.0000*
0.0000*
0.000*
0.000*
0.000*
0.000*
0.000*
0.000*
0.000*
β1=1 Wald Test β0=0 and β1=1
Notes: Significant at:* 0.01and ** 0.05 levels