COFFEE HEDGING: EVIDENCE OF ROBUSTA FROM VIETNAM45402

Green financial system in Vietnam - Challenges and impacts on the economy COFFEE HEDGING: EVIDENCE OF ROBUSTA FROM VIETNAM Nguyen Thi Nhung*, Tran Thi Van Anh, Nguyen Nhu Ngan, Tran Thi Hong, Nguyen Dinh Cuong Faculty of Finance and Banking, University of Economics and Business-VNU ABSTRACT For a long time, hedging for coffee has been a big preoccupation of Vietnamese Government and citizens Considering the important role of hedging in forward trading, the Government allowed forward transactions for coffee through commercial banks since 2004 and officially established Buon Ma Thuot Coffee Exchange Center (BECE) in 2008 and Vietnam Commodity Exchange (VNX) in 2011 However, trading results on coffee are still modest in terms of value and volume In other word, both BECE and VNX experienced a very low liquidity On 08th June 2018, the Ministry of Industry and Trade issued License No 486/ GP-BCT, resulting to create Mercantile Exchange of Vietnam - the nationally centralized commodities market Using daily data of domestic spot and futures price of types of contracts including: (i) Liffe-Robusta Coffee Jan 2020 (P1); (ii) Liffe-Robusta Coffee Mar 2020 (P2); (iii) Liffe-Robusta Coffee May 2020 (P3); (iv) Liffe-Robusta Coffee Jul 2020 (P4); (v) Liffe-Robusta Coffee Sept 2020 (P5); (vi) Liffe-Robusta Coffee Nov 2020 (P6) during the period of years, and vector error correction model, the research investigates if coffee futures trading on international commodity exchange (ICE EU) complete mission of hedging for the domestic spot market or not The research result indicates that there is no relationship between domestic spot and future prices for Robusta in Vietnam This unexpected evidence brought a question mark about hedging for agricultural products in general and coffee in particular in Vietnam Keywords: Hedging, Futures Contract, Derivatives, Coffee, Vietnam * Corresponding author Email address: nguyenthinhung.1684@gmail.com 162 Vietnam National University - University of Economics and Business INTRODUCTION Agricultural production is the most significant manufacturing sector for the development of Vietnam (OECD, 2015) Coffee is one of the most produced agricultural products in this country, second only to rice According to statistics from General Statistics Office of Vietnam (2019), coffee yield reached 1626.2 tons in 2018, increased by 3.1% year over year The result shows that Vietnam is currently ranked in the top five largest coffee exporters globally (WASI, 2019) Exporting coffee has played an important role in Vietnam’s economy, which not only contributes to government budget but also provides a wide range of job opportunities and incomes for residents in rural, midland and mountainous regions In 2018, Vietnam exported 1.9 million tons of coffee with export turnover reached $3.5 billion, accounted for 14% of the market share and 10.4% of the global export coffee value, ranked second behind Brazil but far ahead of other competitors (MOIT, 2019) Besides, exporting coffee also contributes to economic restructuring, promotes the development of the processing industry and the machinery manufacturing industry, gives an impulse to expand international trade relationship and supports the industrialization and modernization in Vietnam Since 2000, in an attempt to the economy to the market, the Government of Vietnam has committed to innovating to develop the agricultural market through establishing and developing Commodity Exchanges for some agricultural products such as coffee, rice, rubber… which play a very important role for the national economy and are greatly affected by price fluctuations on international markets In 2004, the State Bank of Vietnam allowed a number of commercial banks to carry out forward transactions on international exchanges and simultaneously set up domestic commodity exchanges In terms of coffee, BuonMaThuot Coffee Exchange Center (BECE) was established in 2008 and Vietnam Commodity Exchange (VNX) was introduced in 2011 However, Nguyễn (2017) indicated that trading results on coffee are modest in terms of value and volume In other word, both BECE and VNX experienced very low liquidity They didn’t complete their role and mission Few years after, on 08th June 2018, the Ministry of Industry and Trade issued License No 486/GP-BCT, allowing Mercantile Exchange of Vietnam operate This is the only nationally centralized commodities market organizer in Vietnam On 17th August 2018, MXV officially launched the nationally centralized commodities market It is clearly seen that risk management for agricultural products in general and coffee in particular, has been a big preoccupation in Vietnam since many years The need of hedging coming from Vietnamese producers/enterprises dramatically has increased With introduction of MXV since more than a year, Vietnamese participants have an opportunity of hedging their positions by selling or buying coffee futures on ICE EU for Robusta The above practice provides us a high motivation to examine if coffee futures trading on international commodity exchanges complete mission of hedging for the domestic spot market or not In case coffee futures offered by ICE EU are useful financial instruments of risk management for Vietnamese producers/enterprises, optimal hedging ratios as well as optimal number of contracts will be determined In line with this objective, this research applies Vector Error Correction Model (VECM) 163 Green financial system in Vietnam - Challenges and impacts on the economy to investigate if the long-run relationship between futures price and domestic price exists or not and which futures are the best hedging tools for Vietnamese participants In addition, by applying theorical framework developed by John C Hull (2012) the working paper establishes optimal hedging ratio and optimal number of each type of futures To our best knowledge, the contributions of the study are framed in providing empirical evidences indicating if futures trading on Robusta coffee on ICE EU plays an important role of hedging for Vietnamese participant or not as well as what are the optimal hedging ratio and optimal number of contracts for each kind of futures that they should refer when trading on international commodity exchanges These important results will be germane to have more efficient risk management in terms of price After Introduction that briefly presents our working paper, literature review is presented in the second part with researches related to futures role of hedging and optimal hedge ratio The third part describes methodology and data that the article uses to achieve research objectives, following with the results explanation in the 4th part and a discussion in 5th part Some conclusions are provided in the last part LITERATURE REVIEW The commodity futures market is a public market where commodities are contracted to buy or sell at an agreed price for delivery on a given day Thus, the main difference between the futures market and the spot market is that in the futures market, deliveries take place not in real time but in a future time and the terms of the transaction are standardized in a contract For its part, the futures contract is a legal contract attached to an actual commodity and the people involved in these contracts perform the economic function of setting the market price for the goods (Lerner, 2000) The commodity futures markets for agricultural products are formed and developed from the activities of buying and selling agricultural goods of countries specializing in agricultural production Due to the characteristics of agricultural production such as seasonality, supply exceeding demand in the harvest season, the price reduction when high harvest crops make agricultural producers have many difficulties due to the price fluctuations For people using agricultural products as raw materials for their production activities, the instability of raw material prices is also the risk they often face By participating in futures transaction, the enterprises who produce or deal in cash commodities could hedge price risk, reduce operational risks and unpredictable price changes (Cifarelli and Paladino, 2015; Wang and Wen, 2014; Lerner, 2000) Thus, hedging against risks is the original reason for the appearance of commodity futures market When investors who recognize that engaging in the futures trading can earn certain profits, although they have no demand for agricultural goods, they still participate in the commodity futures market In other words, they enter the commodity futures markets for speculative purposes These two activities, however, are not necessarily regarded as two separate behaviors It is possible that typical risk hedges carry out speculative acts to enjoy 164 Vietnam National University - University of Economics and Business price differences In other cases, speculators may find it beneficial to engage in hedging activities (Stulz, 1996; Irwin et al, 2009) The participation of two groups of participants has facilitated the commodity futures market in general and commodity futures markets for agricultural products in particular to develop in higher form This can be understood as the futures trading can stimulate market transparency, establish price discovery mechanisms, reduce collusion between market participants, mitigate price fluctuations and bubbles appearance as well as provide more accurate pricing information for all market participants (UNCTAD, 2009; Rashid et al, 2010) The hedging behavior in the commodity futures market for agricultural products can bring some specific benefits as follow Anticipatory hedging: Agricultural producers participating in the spot market may be at risk because prices fluctuate in an unfavorable direction for them from the time a decision is made to plant a particular agricultural crop until it is harvested Price risk in the spot market makes it difficult for producers to know exact income and thus it is difficult to plan their specific activities accurately However, by signing the contract on the futures market with the position opposed to the spot market, the manufacturers may have the ability to cover losses for transactions on the spot market by profits earned on the future market Flexibility in pricing: The commodity futures market provides a variety of contracts for each type of agricultural products, thus can increase the level of flexibility in pricing So, each trader can choose the price that best suits his purpose and needs (Thompson, 1985) In addition, in the commodity futures market, transactions must go through a broker who can be considered as a trading agent on behalf of the producers This can guarantee a minimum price for manufacturers’ outputs and thus reduce their risk (WB, 1999) Inventory management: the price difference between futures contracts or the price spread can be seen as a signal of the availability of stocks in the market The difference between futures and spot prices is often called the basis, which is a measure of the storage and interest costs incurred by a spot market trader in holding stocks right now to sell later Therefore, it can be considered that the level of inventory held in the spot market will be determined by the basis This will ensure a smoother price model in the spot market and may reduce price volatility (Netz, 1995; Morgan, 1999) The optimal hedging behavior are analyzed by Stein (1961) and McKinnon (1967) They associated it with the minimization of the variance of the return of the portfolio of a hedger, constructed with cash and futures contracts They show how to calculate an optimal cover ratio - (the minimum variance hedge ratio or MVH ratio), which is defined as the percentage of cash contracts matched by futures positions that minimizes the variance of the hedged portfolio The improvements of MVH strategy are strategies based on hedged portfolio return mean and variance expected utility maximization (Lence, 1995) or based on the generalized semi variance (Lien and Tse, 2000) Further improvements are the result of adopting new estimation techniques, taking into account for the non-stationarity and the heteroscedasticity of the time series 165 Green financial system in Vietnam - Challenges and impacts on the economy Many authors use different methods to understand hedging behavior for agricultural products in different commodity futures markets For example, Wang and Chidmi (2011) use the results of ordinary least squares, bivariate vector autoregressive and error correction models to estimate the hedge ratios for cotton production across different countries Cifarelli and Paladino (2015) investigate the dynamic behavior of futures returns on five commodity markets namely copper, cotton, oil, silver, and soybeans The relationship between expected spot and futures returns and timevarying volatilities is estimated using a non-linear in mean CCC-GARCH approach Qiang and Fan (2016) consider the contemporaneous causality between China’s oil markets with other commodity markets both domestically and internationally using an error correction model combined with a directed acyclic graph technique The authors note that due to the lack of a future market, the Chinese crude market has little effect on other commodity markets Öztek and Öcal (2017) model time-varying correlations between commodity and stock markets to uncover the dynamic nature of correlations during the financialization of commodity markets and in the aftermath of the recent financial crisis METHODOLOGY 3.1 Research Design The Robusta Coffee futures contract is used as the global benchmark for the pricing of physical Robusta Coffee It is actively traded by producers, exporters, trade houses, importers and roasters as well as by managed funds and both institutional and short-term investors After investigating if Robusta coffee futures (including that are traded on ICE EU can serve as a hedging instrument for Vietnamese participants, the research tries to compare the significance various futures contracts in terms of hedging instruments and then points out which contracts are the most significant hedging tools for Vietnamese participants 3.2 Data Source The research totally uses the secondary data including spot prices and futures prices In detail, spot price in Vietnam Futures prices are divided into 06 series, including: (i) Liffe-Robusta Coffee Jan 2020 (P1); (ii) Liffe-Robusta Coffee Mar 2020 (P2); (iii) Liffe-Robusta Coffee May 2020 (P3); (iv) Liffe-Robusta Coffee Jul 2020 (P4); (v) Liffe-Robusta Coffee Sept 2020 (P5); (vi) Liffe-Robusta Coffee Nov 2020 (P6) The underlying asset of all contracts is Coffee-ICO Robusta Average c/lb (COFICRB) In addition, series are collected since futures contracts are introduced for trading on commodity exchange That’s why, the research periods for various contracts are different All data are collected from Thomson Reuters - Liffe-Robusta Coffee Jan 2020: Sample of 395 observations, from 1rstJuly, 2018 to 3rd December 2019 166 Vietnam National University - University of Economics and Business - Liffe-Robusta Coffee Mar 2020: Sample of 356 observations, from 26th July, 2018 to 3rd December 2019 - Liffe-Robusta Coffee May 2020: Sample of 313 observations, from 25th September, 2018 to 3rd December 2019 - Liffe-Robusta Coffee Jul 2020: Sample of 268 observations, from 27th November, 2018 to 3rd December 2019 - Liffe-Robusta Coffee Sept 2020: Sample of 224 observations, from 28th January, 2019 to 3rd December 2019 - Liffe-Robusta Coffee Nov 2020: Sample of 180 observations, from 29 Mars, 2019 to 3rd December 2019 3.3 Methods of Data Analysis ▪ Step 1: Stationarity Test The study applies Augmented Dickey-Fuller Test or Unit Root Test to the residuals from this regression in order to test stationary of price data, including spot prices of coffee on domestic market and futures prices on international commodity exchanges, including ICE London and ICE United State Call: Po: Spot price of coffee (domestic price) Pj: Future price of coffee on international commodity exchanges There are also three basic regression models as follows: No constant, no trend: ∆Po,t = βPj, t-1 + ut Constant, no trend: ∆Po,t = α + βPj, t-1 + ut Constant and trend: ∆Po,t = α + βPj, t-1 +δt + ut Where: ∆Po,t: Change in spot rate on domestic market at time t There are two hypotheses: H0: β = → the time series in non - stationary H1: β < → the time series is stationary If t-Statistic is bigger than τ on Kendall’s tau table, the hypothesis H0 is rejected (accepted) when t-Statistic is bigger (smaller) than τ on Kendall’s tau table ▪ Step 2: Determining optimal Lag Optimal Lag is pointed out by the Akaike Selection Criterion In theory, the lag length is figured out only when this criterion has the smallest value, and it can make sure the stability of the model ▪ Step 3: Co-integration Test In attempt to confirm if spot prices and futures prices are co-integrated, the research applies Johansen Co-integration Test through criteria such as maximal eigen value test and trace test We have two hypotheses: (i) H0: No co-integrating equation between spot prices and futures prices; 167 Green financial system in Vietnam - Challenges and impacts on the economy (ii) H1: Co-integrating equation between spot prices and futures prices So, hypothesis will be rejected if the Trace and Max statistics is more than 5% critical value otherwise ▪ Step 4: Vector Error Correction Model (VECM) We have to investigate price vectors: PT = [P0;Pj] , referring to the spot prices on domestic market and futures prices on international commodity exchanges An estimated VECM is as below: * * ∆𝑃𝑃 !,# = 𝛽𝛽$ + ∑%+) 𝛽𝛽% ∆𝑃𝑃&,'() + ∑%+) 𝛿𝛿% ∆𝑃𝑃$,'() + 𝜔𝜔𝜇𝜇'() + 𝑣𝑣' [1] Cointegrating equation (long-run model): µ!"# = 𝐸𝐸𝐸𝐸𝐸𝐸$"# = 𝑃𝑃%,$"# − 𝛽𝛽' − 𝛽𝛽# 𝑃𝑃',$"# [2] Where: P0,t: Spot price of coffee on domestic market at time t; Pj,t: Future price of coffee on international commodity exchanges at time t; ∆ is the difference in price; μt-1 is the lagged value of the error correction term; vt is a white noise error term The first equation describes both the short-run and long-run dynamics between spot price and future price while the second one only focuses on long-run interconnection P0,t and Pj,t experience a long-run relationship when the coefficient of the co-integrating equation is between -1 and at a statistical significance In fact, the coefficient of ETC w indicates how quickly the dependent variable (Pj,t) returns to equilibrium after a change in independent variable (P0,t) In addition, a shot-run a relationship between the two kinds of price is point out by Wald Test and Breusch-Godfrey Serial Correction LM Test Stability Diagnostics/Recursive Estimates (OLS only) allows to demonstrate a dynamical stability of the model RESULTS 4.1 Data Description Table reports summary statistics (about max, average, and standard deviation of return) for futures prices and spot price for Robusta Coffee It is clearly seen that there is a downward trend during research period while domestic price of Robusta in Viet Nam doesn’t change a lot during the period from June 2018 to December 2019 [Figure 1] 168 Vietnam National University - University of Economics and Business Table Statistics of Futures Prices and Spot Price of Robusta Coffee Contracts Max Min Average Standard Deviation LIFFE-ROBUSTA COFFEE JAN 2020 (USD) 1,861.00 1,227.00 1,546.26 157.53 LIFFE-ROBUSTA COFFEE MAR 2020 (USD) 1,875.00 1,252.00 1,541.86 143.94 LIFFE-ROBUSTA COFFEE MAY 2020 (USD) 1,887.00 1,277.00 1,547.31 143.37 LIFFE-ROBUSTA COFFEE JUL 2020 (USD) 1,751.00 1,304.00 1,529.75 112.24 LIFFE-ROBUSTA COFFEE SEPT 2020 (USD) 1,720.00 1,330.00 1,522.00 96.02 LIFFE-ROBUSTA COFFEE NOV 2020 (USD) 1,640.00 1,358.00 1,508.38 67.33 DOMESTIC SPOT (VND) 36,950.00 29,950.00 33,415.19 1,489.87 Source: Calculation by authors based on data collected from Thomson Reuters Source Calculation by authors based on data collected from Thomson Reuters Figure Movements of Futures Prices and Domestic Spot 4.2 Empirical Results Appendix show the results of Augmented Dickey-Fuller Test Domestic spot, Price of Liffe-Robusta Coffee Jan 2020, Price of Liffe-Robusta Coffee Mar 2020, Price of Liffe-Robusta Coffee May 2020, Price of Liffe-Robusta Coffee Jul 2020, Price of Liffe-Robusta Coffee Sept 2020 and Price of Liffe-Robusta Coffee Nov 2020 have p-value and t-Statistic as bellow: Domestic price: p-value = 0.4499 > α = 10%, t-Statistic = -3.981521 |τ1% | = |−3.4361 | Price of Liffe-Robusta Coffee Jan 2020: p-value = 0.2864 > α = 10%, t-Statistic = -3.981521 |τ1% | = |−3.4361 | Price of Liffe-Robusta Coffee Mar 2020: p-value = 0.3802 > α = 10%, t-Statistic = -3.984120 |τ1% | = |−3.4361 | Price of Liffe-Robusta Coffee May 2020: p-value = 0.0492 < α = 10%, t-Statistic = -3.987745 |τ1% | = |−3.4361 | Price of Liffe-Robusta Coffee Jul 2020: p-value = 0.1740 > α = 10%, t-Statistic = -3.992801 |τ1% | = |−3.4361 | Price of Liffe-Robusta Coffee Sept 2020: p-value = 0.2803 > α = 10%, t-Statistic 169 = -3.999740 |τ | = |−3.4361 | 1% Price of Liffe-Robusta Coffee Nov 2020: p-value = 0.2452 > α = 10%, t-Statistic Price of Liffe-Robusta Coffee May 2020: p-value = 0.0492 < α = 10%, t-Statistic = -3.987745 |τ1% | = |−3.4361 | Price of financial Liffe-Robusta Coffee Jul 2020: p-value = and 0.1740 > α =on 10%, Green system in Vietnam - Challenges impacts the t-Statistic economy = -3.992801 |τ1% | = |−3.4361 | Price of Liffe-Robusta Coffee Sept 2020: p-value = 0.2803 > α = 10%, t-Statistic = -3.999740 |τ1% | = |−3.4361 | Price of Liffe-Robusta Coffee Nov 2020: p-value = 0.2452 > α = 10%, t-Statistic = 4.010143 |τ1% | = |−3.4361 | It is clearly seen all series (including Domestic spot, Price of Liffe-Robusta Coffee Jan 2020, Price of Liffe-Robusta Coffee Mar 2020, Price of Liffe-Robusta Coffee May 2020, Price of Liffe-Robusta Coffee Jul 2020, Price of Liffe-Robusta Coffee Sept 2020 and Price of Liffe-Robusta Coffee Nov 2020) experience constant and trend stationarity with negative t-Statistic at about the 1% level Appendix shows that is the best optimal lag for all series couple (Domestic spot and Price of Liffe-Robusta Coffee Jan 2020, Domestic spot and Price of LiffeRobusta Coffee Mar 2020, Domestic spot and Price of Liffe-Robusta Coffee May 2020, Domestic spot and Price of Liffe-Robusta Coffee Jul 2020, Domestic spot and Price of Liffe-Robusta Coffee Sept 2020, Domestic spot and Price of LiffeRobusta Coffee Nov 2020) In terms of co-integration, Trace and Max-eigenvalue test indicate no cointegration at the 0.05 level between Domestic spot and Price of Liffe-Robusta Coffee May 2020, Domestic spot and Price of Liffe-Robusta Coffee Jul 2020, Domestic spot and Price of Liffe-Robusta Coffee Sept 2020, Domestic spot and Price of LiffeRobusta Coffee Nov 2020 [Appendix 3] However, there is co-integrating equations at the 0.05 level between Domestic spot and Price of Liffe-Robusta Coffee Jan 2020, Domestic spot and Price of Liffe-Robusta Coffee Mar 2020 In the longrun, only Futures contracts (including Liffe-Robusta Coffee Jan 2020 and LiffeRobusta Coffee Mar 2020) experience a positive impact on domestic spot The null hypothesis of no co-integration is rejected against the alternative of a co-integrating relationship in the model for Liffe-Robusta Coffee Jan 2020 and Liffe-Robusta Coffee Mar 2020 In the other words, these indices exhibit a long-run relationship, satisfying requirements of Vector Error Correction Model (VECM) Appendix shows error correction model of Domestic spot and Price of LiffeRobusta Coffee Jan 2020, Domestic spot and Price of Liffe-Robusta Coffee Mar 2020 Table summarizes equations about estimated VECM and co-integrating (long-run model) Table Estimated VECM and cointegrating equation (long-run model) Estimated VECM Cointegrating equation (long-run model) Domestic spot ∆P1,t = - 0,009827 × ETCt-1 - 0,029624 μt-1 = ETCt-1 = P1,t-1 - 4.173551 × P0,t-1 and Price of × ∆P1,t-1 + 0,001007 × ∆P0,t-1 + 4457.532 Liffe-Robusta 1.144732 Coffee Jan 2020 Estimated equation: D(P1) = C(1) × ( P1(-1) - 4.1735–57.5324 ) + C(2) × D(P1(-1)) + C(3) × D(P0(-1)) + C(4) 170 Vietnam National University - University of Economics and Business Cointegrating equation (long-run model) Estimated VECM Domestic spot ∆P1,t = 0,010041 × ETCt-1 - 0,024469 μt-1 = ETCt-1 = P2,t-1 - 4.568667 × P0,t-1 and Price of × ∆P1,t-1 + 0,011209 × ∆P0,t-1 + 4980.191 Liffe-Robusta - 0,957390 Coffee Mar 2020 Estimated equation: D(P2) = C(1) × ( P2 (-1) - 4.5686 × P0(-1) + 4980.19087818 ) + C(2) × D(P2(-1)) + C(3) × D(P0 (-1)) + C(4) Source Calculation by authors based on data collected from Thomson Reuters It can be obviously seen that coefficient of the co-integrating equation is 0.009827 and 0.010041, bigger than [Table and Table 4], resulting the conclusion that there is any long-run relationship between Domestic spot and Price of Liffe-Robusta Coffee Jan 2020 In other word, there is any evidence for a long-run relationship between Domestic spot and Futures Price Table Error Correction Model for Domestic spot and Price of Liffe-Robusta Coffee Jan 2020 Dependent Variable: D(LIFFE_ROBUSTA_COFFEE_JAN) Method: Least Squares Date: 12/10/19 Time: 21:01 Sample (adjusted): 6/05/2018 12/03/2019 Included observations: 393 after adjustments D(LIFFE_ROBUSTA_COFFEE_JAN) = C(1) × ( LIFFE_ROBUSTA_COFFE         E_JAN(-1) - 4.17355117723 × DOMESTIC_SPOT USD_(-1) +         4457.5324554 ) + C(2) × D(LIFFE_ROBUSTA_COFFEE_JAN(-1)) +         C(3) × D(DOMESTIC_SPOT USD_(-1)) + C(4) Coefficient Std Error t-Statistic Prob.   C(1) 0.009827 0.004315 2.277567 0.0233 C(2) -0.029624 0.051629 -0.573791 0.5664 C(3) 0.001007 0.070241 0.014341 0.9886 C(4) -1.144732 0.882510 -1.297133 0.1954 R-squared 0.013280     Mean dependent var -1.109415 Adjusted R-squared 0.005670     S.D dependent var 17.49696 S.E of regression 17.44729     Akaike info criterion 8.566371 Sum squared resid 118414.6     Schwarz criterion 8.606817 Log likelihood -1679.292     Hannan-Quinn criter 8.582400 Durbin-Watson stat 1.996459 Source: Calculation by authors based on data collected from Thomson Reuters 171 Vietnam National University - University of Economics and Business Null Hypothesis: LIFFE_ROBUSTA_COFFEE_NOV has a unit root Exogenous: Constant, Linear Trend Lag Length: (Automatic - based on SIC, maxlag = 13) Augmented Dickey-Fuller test statistic Test critical values: t-Statistic   Prob.* -2.682424  0.2452 1% level -4.010143 5% level -3.435125 10% level -3.141565 *MacKinnon (1996) one-sided p-values Augmented Dickey-Fuller Test Equation Dependent Variable: D (LIFFE_ROBUSTA_COFFEE_NOV) Method: Least Squares Date: 12/10/19 Time: 16:10 Sample (adjusted): 4/01/2019 12/03/2019 Included observations: 179 after adjustments Variable Coefficient Std Error t-Statistic Prob.   LIFFE_ROBUSTA_COFFEE_NOV (-1) -0.081028 0.030207 -2.682424 0.0080 C 127.1139 48.22744 2.635717 0.0091 @TREND ("3/29/2019") -0.063002 0.039334 -1.601710 0.1110 R-squared 0.040646     Mean dependent var -0.793296 Adjusted R-squared 0.029744     S.D dependent var 18.93177 S.E of regression 18.64809     Akaike info criterion 8.705983 Sum squared resid 61204.25     Schwarz criterion 8.759403 Log likelihood -776.1854     Hannan-Quinn criter 8.727644 F-statistic 3.728389     Durbin-Watson stat 2.035560 Prob(F-statistic) 0.025950 Source Calculation by authors based on data collected from Thomson Reuter and EViews 179 Green financial system in Vietnam - Challenges and impacts on the economy Appendix Optimal Lag VAR Lag Order Selection Criteria Endogenous variables: LIFFE_ROBUSTA_COFFEE_JAN DOMESTIC_SPOT USD_  Exogenous variables: C  Date: 12/10/19 Time: 16:36 Sample: 6/01/2018 12/03/2019 Included observations: 387  Lag LogL LR FPE AIC SC HQ -4557.914 NA   58802408  23.56545  23.58590  23.57356 -3180.337  2733.796  48586.71  16.46686   16.52823*  16.49119 -3178.137  4.342844  49040.94  16.47616  16.57845  16.51672 -3173.906  8.309437  48982.61  16.47497  16.61817  16.53175 -3170.587  6.483906  49155.70  16.47849  16.66260  16.55149 -3158.234  24.00306  47079.64  16.43532  16.66035  16.52455 -3138.276  38.57596  43353.53  16.35285  16.61879   16.45830* -3133.320   9.527067*   43141.13*   16.34791*  16.65476  16.46958 -3131.562  3.362553  43645.32  16.35949  16.70726  16.49739  * indicates lag order selected by the criterion  LR: sequential modified LR test statistic (each test at 5% level)  FPE: Final prediction error  AIC: Akaike information criterion  SC: Schwarz information criterion  HQ: Hannan-Quinn information criterion VAR Lag Order Selection Criteria Endogenous variables: LIFFE_ROBUSTA_COFFEE_MAR DOMESTIC_SPOT USD_  Exogenous variables: C  Date: 12/10/19 Time: 16:34 Sample: 7/26/2018 12/03/2019 Included observations: 348  Lag LogL LR FPE AIC SC HQ -4074.298 NA   51199851  23.42700  23.44914  23.43582 -2877.324  2373.311  53910.13  16.57083   16.63724*  16.59727 -2875.505  3.586409  54590.18  16.58336  16.69405  16.62743 -2871.723  7.411402  54659.06  16.58461  16.73959  16.64631 -2868.919  5.463184  55036.70  16.59149  16.79074  16.67081 -2857.473  22.16734  52732.36  16.54870  16.79223  16.64565 -2838.695   36.15267*  48440.04  16.46377  16.75157   16.57835* -2834.033  8.923729   48257.66*   16.45996*  16.79204  16.59217 -2832.078  3.718903  48830.57  16.47171  16.84807  16.62155 180 Vietnam National University - University of Economics and Business  * indicates lag order selected by the criterion  LR: sequential modified LR test statistic (each test at 5% level)  FPE: Final prediction error  AIC: Akaike information criterion  SC: Schwarz information criterion  HQ: Hannan-Quinn information criterion VAR Lag Order Selection Criteria Endogenous variables: LIFFE_ROBUSTA_COFFEE_MAY DOMESTIC_SPOT USD_  Exogenous variables: C  Date: 12/10/19 Time: 16:33 Sample: 9/25/2018 12/03/2019 Included observations: 305  Lag LogL LR FPE AIC SC HQ -3577.383 NA   53522485  23.47137  23.49576  23.48112 -2534.517  2065.217  58886.94  16.65913   16.73231*   16.68840* -2533.017  2.951004  59860.50  16.67552  16.79750  16.72431 -2528.824  8.193718  59785.36  16.67425  16.84502  16.74256 -2526.223  5.048967  60337.34  16.68343  16.90299  16.77125 -2516.150  19.41776  57983.72  16.64361  16.91196  16.75094 -2500.123   30.68827*  53588.56  16.56474  16.88188  16.69159 -2495.850  8.126507   53495.54*   16.56295*  16.92888  16.70931 -2494.184  3.146025  54324.57  16.57825  16.99298  16.74414  * indicates lag order selected by the criterion  LR: sequential modified LR test statistic (each test at 5% level)  FPE: Final prediction error  AIC: Akaike information criterion  SC: Schwarz information criterion  HQ: Hannan-Quinn information criterion VAR Lag Order Selection Criteria Endogenous variables: LIFFE_ROBUSTA_COFFEE_JUL DOMESTIC_SPOT USD_  Exogenous variables: C  Date: 12/10/19 Time: 16:30 Sample: 11/27/2018 12/03/2019 Included observations: 260  Lag LogL LR FPE AIC SC HQ -2937.748 NA   22695820  22.61345  22.64084  22.62446 -2140.609  1575.882  50849.58  16.51238   16.59455*   16.54541* -2137.352  6.390183  51141.13  16.51809  16.65504  16.57314 -2133.734  7.039913  51292.73  16.52103  16.71276  16.59811 181 Green financial system in Vietnam - Challenges and impacts on the economy -2131.949  3.446429  52175.70  16.53807  16.78458  16.63717 -2124.006  15.21380  50619.26  16.50774  16.80903  16.62886 -2117.429   12.49671*   49628.90*   16.48792*  16.84398  16.63106 -2114.623  5.288491  50091.26  16.49710  16.90795  16.66227 -2112.982  3.067153  51014.54  16.51525  16.98087  16.70244  * indicates lag order selected by the criterion  LR: sequential modified LR test statistic (each test at 5% level)  FPE: Final prediction error  AIC: Akaike information criterion  SC: Schwarz information criterion  HQ: Hannan-Quinn information criterion VAR Lag Order Selection Criteria Endogenous variables: LIFFE_ROBUSTA_COFFEE_SEP DOMESTIC_SPOT USD_  Exogenous variables: C  Date: 12/10/19 Time: 16:28 Sample: 1/28/2019 12/03/2019 Included observations: 216  Lag LogL LR FPE AIC SC HQ -2415.890 NA   18112490  22.38787  22.41912  22.40049 -1790.080  1234.235  57217.86  16.63037   16.72413*   16.66825* -1787.782  4.490712  58127.16  16.64613  16.80239  16.70926 -1784.097  7.130705  58298.76  16.64904  16.86781  16.73743 -1782.620  2.830424  59679.91  16.67241  16.95368  16.78604 -1773.940  16.47532  57153.74  16.62908  16.97286  16.76797 -1768.823   9.619310*   56571.83*   16.61873*  17.02501  16.78287 -1765.930  5.384171  57163.62  16.62898  17.09777  16.81837 -1764.692  2.281476  58656.28  16.65455  17.18585  16.86920  * indicates lag order selected by the criterion  LR: sequential modified LR test statistic (each test at 5% level)  FPE: Final prediction error  AIC: Akaike information criterion  SC: Schwarz information criterion  HQ: Hannan-Quinn information criterion VAR Lag Order Selection Criteria Endogenous variables: LIFFE_ROBUSTA_COFFEE_NOV DOMESTIC_SPOT USD_  Exogenous variables: C  Date: 12/10/19 Time: 16:24 Sample: 3/29/2019 12/03/2019 Included observations: 172 182 Vietnam National University - University of Economics and Business  Lag LogL LR FPE AIC SC HQ -1886.192 NA   11757075  21.95572  21.99232  21.97057 -1444.778  867.4286   72676.28*   16.86952*   16.97931*   16.91406* -1443.425  2.629132  74949.09  16.90029  17.08328  16.97453 -1439.972  6.624609  75431.79  16.90665  17.16284  17.01059 -1438.945  1.946911  78092.77  16.94122  17.27060  17.07486 -1431.835   13.30980*  75331.46  16.90506  17.30764  17.06840 -1427.822  7.419806  75337.05  16.90490  17.38069  17.09794 -1425.319  4.569712  76683.61  16.92231  17.47129  17.14505 -1424.626  1.248209  79722.61  16.96077  17.58295  17.21320  * indicates lag order selected by the criterion  LR: sequential modified LR test statistic (each test at 5% level)  FPE: Final prediction error  AIC: Akaike information criterion  SC: Schwarz information criterion  HQ: Hannan-Quinn information criterion Source Calculation by authors based on data collected from Thomson Reuter and EViews 183 Green financial system in Vietnam - Challenges and impacts on the economy Appendix Trace and Max-Eigenvalue Test Date: 12/10/19 Time: 19:36 Sample (adjusted): 6/05/2018 12/03/2019 Included observations: 393 after adjustments Trend assumption: Linear deterministic trend Series: LIFFE_ROBUSTA_COFFEE_JAN DOMESTIC_SPOT USD_  Lags interval (in first differences): to Unrestricted Cointegration Rank Test (Trace) Hypothesized Trace 0.05 No of CE(s) Eigenvalue Statistic Critical Value Prob.** None *  0.036283  16.87110  15.49471  0.0308 At most  0.005953  2.346570  3.841466  0.1256 Max-Eigen 0.05  Trace test indicates cointegrating eqn(s) at the 0.05 level  * denotes rejection of the hypothesis at the 0.05 level  **MacKinnon-Haug-Michelis (1999) p-values Unrestricted Cointegration Rank Test (Maximum Eigenvalue) Hypothesized No of CE(s) Eigenvalue Statistic Critical Value Prob.** None *  0.036283  14.52453  14.26460  0.0455 At most  0.005953  2.346570  3.841466  0.1256  Max-eigenvalue test indicates cointegrating eqn(s) at the 0.05 level  * denotes rejection of the hypothesis at the 0.05 level  **MacKinnon-Haug-Michelis (1999) p-values  Unrestricted Cointegrating Coefficients (normalized by b'*S11*b=I):  LIFFE_ROBUSTA_COFFEE_JAN DOMESTIC_ SPOT USD_ -0.004902  0.020460  0.007108 -0.002789  Unrestricted Adjustment Coefficients (alpha):  D(LIFFE_ROBUSTA_COFFEE_JAN) -2.004485 -1.076297 D(DOMESTIC_SPOT USD_) -1.954788  0.545459 Cointegrating Equation(s):  Log likelihood -3223.315 Normalized cointegrating coefficients (standard error in parentheses) LIFFE_ROBUSTA_COFFEE_JAN DOMESTIC_ SPOT USD_  1.000000 -4.173551  (0.81337) Adjustment coefficients (standard error in parentheses) D(LIFFE_ROBUSTA_COFFEE_JAN)  0.009827  (0.00431) 184 Vietnam National University - University of Economics and Business D(DOMESTIC_SPOT USD_)  0.009583  (0.00306) Date: 12/10/19 Time: 19:46 Sample (adjusted): 7/30/2018 12/03/2019 Included observations: 354 after adjustments Trend assumption: Linear deterministic trend Series: LIFFE_ROBUSTA_COFFEE_MAR DOMESTIC_SPOT USD_  Lags interval (in first differences): to Unrestricted Cointegration Rank Test (Trace) Hypothesized Trace 0.05 No of CE(s) Eigenvalue Statistic Critical Value Prob.** None *  0.037586  16.01403  15.49471  0.0418 At most  0.006902  2.451907  3.841466  0.1174  Trace test indicates cointegrating eqn(s) at the 0.05 level  * denotes rejection of the hypothesis at the 0.05 level  **MacKinnon-Haug-Michelis (1999) p-values Unrestricted Cointegration Rank Test (Maximum Eigenvalue) Hypothesized Max-Eigen 0.05 No of CE(s) Eigenvalue Statistic Critical Value Prob.** None  0.037586  13.56212  14.26460  0.0643 At most  0.006902  2.451907  3.841466  0.1174  Max-eigenvalue test indicates no cointegration at the 0.05 level  * denotes rejection of the hypothesis at the 0.05 level  **MacKinnon-Haug-Michelis (1999) p-values  Unrestricted Cointegrating Coefficients (normalized by b'*S11*b=I):  LIFFE_ROBUSTA_COFFEE_ MAR DOMESTIC_SPOT_ USD_ -0.004512  0.020613  0.007277 -0.001278  Unrestricted Adjustment Coefficients (alpha):  D(LIFFE_ROBUSTA_COFFEE_ MAR) -2.225633 -1.109810 D(DOMESTIC_SPOT USD_) -2.004420  0.657527 Cointegrating Equation(s):  Log likelihood -2921.311 Normalized cointegrating coefficients (standard error in parentheses) LIFFE_ROBUSTA_COFFEE_ MAR DOMESTIC_SPOTUSD_  1.000000 -4.568667  (1.00989) 185 Green financial system in Vietnam - Challenges and impacts on the economy Adjustment coefficients (standard error in parentheses) D(LIFFE_ROBUSTA_COFFEE_ MAR)  0.010041  (0.00421) D(DOMESTIC_SPOT USD_)  0.009043  (0.00310) Date: 12/10/19 Time: 19:50 Sample (adjusted): 9/27/2018 12/03/2019 Included observations: 311 after adjustments Trend assumption: Linear deterministic trend Series: LIFFE_ROBUSTA_COFFEE_MAY DOMESTIC_SPOT USD_  Lags interval (in first differences): to Unrestricted Cointegration Rank Test (Trace) Hypothesized Trace 0.05 No of CE(s) Eigenvalue Statistic Critical Value Prob.** None  0.037822  14.07222  15.49471  0.0810 At most  0.006671  2.081521  3.841466  0.1491  Trace test indicates no cointegration at the 0.05 level  * denotes rejection of the hypothesis at the 0.05 level  **MacKinnon-Haug-Michelis (1999) p-values Unrestricted Cointegration Rank Test (Maximum Eigenvalue) Hypothesized Max-Eigen 0.05 No of CE(s) Eigenvalue Statistic Critical Value Prob.** None  0.037822  11.99070  14.26460  0.1110 At most  0.006671  2.081521  3.841466  0.1491  Max-eigenvalue test indicates no cointegration at the 0.05 level  * denotes rejection of the hypothesis at the 0.05 level  **MacKinnon-Haug-Michelis (1999) p-values  Unrestricted Cointegrating Coefficients (normalized by b'*S11*b=I):  LIFFE_ROBUSTA_COFFEE_ MAY DOMESTIC_ SPOT USD_ -0.005576  0.019953  0.006473  0.002064  Unrestricted Adjustment Coefficients (alpha):  D(LIFFE_ROBUSTA_COFFEE_ MAY) -1.946929 -1.219108 D(DOMESTIC_SPOT USD_) -2.251703  0.564959 Cointegrating Equation(s):  Log likelihood Normalized cointegrating coefficients (standard error in parentheses) 186 -2583.736 Vietnam National University - University of Economics and Business LIFFE_ROBUSTA_COFFEE_ MAY DOMESTIC_ SPOT USD_  1.000000 -3.578292  (0.84998) Adjustment coefficients (standard error in parentheses) D(LIFFE_ROBUSTA_COFFEE_ MAY)  0.010857  (0.00569) D(DOMESTIC_SPOT USD_)  0.012556  (0.00423) Date: 12/10/19 Time: 19:54 Sample (adjusted): 11/29/2018 12/03/2019 Included observations: 266 after adjustments Trend assumption: Linear deterministic trend Series: LIFFE_ROBUSTA_COFFEE_JUL DOMESTIC_SPOT USD_  Lags interval (in first differences): to Unrestricted Cointegration Rank Test (Trace) Hypothesized Trace 0.05 No of CE(s) Eigenvalue Statistic Critical Value Prob.** None  0.043040  13.73881  15.49471  0.0904 At most  0.007627  2.036607  3.841466  0.1536  Trace test indicates no cointegration at the 0.05 level  * denotes rejection of the hypothesis at the 0.05 level  **MacKinnon-Haug-Michelis (1999) p-values Unrestricted Cointegration Rank Test (Maximum Eigenvalue) Hypothesized Max-Eigen 0.05 No of CE(s) Eigenvalue Statistic Critical Value Prob.** None  0.043040  11.70220  14.26460  0.1223 At most  0.007627  2.036607  3.841466  0.1536  Max-eigenvalue test indicates no cointegration at the 0.05 level  * denotes rejection of the hypothesis at the 0.05 level  **MacKinnon-Haug-Michelis (1999) p-values  Unrestricted Cointegrating Coefficients (normalized by b'*S11*b=I):  LIFFE_ROBUSTA_COFFEE_ JUL DOMESTIC_ SPOT USD_  0.000803  0.021421  0.009262 -0.007769  Unrestricted Adjustment Coefficients (alpha):  187 Green financial system in Vietnam - Challenges and impacts on the economy D(LIFFE_ROBUSTA_COFFEE_JUL) -2.549539 -1.066032 D(DOMESTIC_SPOT USD_) -1.898456  0.764426 Cointegrating Equation(s):  Log likelihood -2182.843 Normalized cointegrating coefficients (standard error in parentheses) LIFFE_ROBUSTA_COFFEE_ JUL DOMESTIC_ SPOT USD_  1.000000  26.68950  (7.98947) Adjustment coefficients (standard error in parentheses) D(LIFFE_ROBUSTA_COFFEE_JUL) -0.002046  (0.00085) D(DOMESTIC_SPOT USD_) -0.001524  (0.00062) Date: 12/10/19 Time: 19:56 Sample (adjusted): 1/30/2019 12/03/2019 Included observations: 222 after adjustments Trend assumption: Linear deterministic trend Series: LIFFE_ROBUSTA_COFFEE_SEP DOMESTIC_SPOT USD_  Lags interval (in first differences): to Unrestricted Cointegration Rank Test (Trace) Hypothesized Trace 0.05 No of CE(s) Eigenvalue Statistic Critical Value Prob.** None  0.031414  10.13196  15.49471  0.2707 At most  0.013628  3.046130  3.841466  0.0809  Trace test indicates no cointegration at the 0.05 level  * denotes rejection of the hypothesis at the 0.05 level  **MacKinnon-Haug-Michelis (1999) p-values Unrestricted Cointegration Rank Test (Maximum Eigenvalue) Hypothesized Max-Eigen 0.05 No of CE(s) Eigenvalue Statistic Critical Value Prob.** None  0.031414  7.085835  14.26460  0.4790 At most  0.013628  3.046130  3.841466  0.0809  Max-eigenvalue test indicates no cointegration at the 0.05 level  * denotes rejection of the hypothesis at the 0.05 level  **MacKinnon-Haug-Michelis (1999) p-values  Unrestricted Cointegrating Coefficients (normalized by b'*S11*b=I):  188 Vietnam National University - University of Economics and Business LIFFE_ROBUSTA_COFFEE_ DOMESTIC_SPOT SEP USD_  0.003191  0.020261  0.010079 -0.009008  Unrestricted Adjustment Coefficients (alpha):  D ( L I F F E _ R O B U S TA _ C O F FEE_SEP) -2.456057 -1.304341 D(DOMESTIC_SPOT USD_) -1.397321  1.212561 Cointegrating Equation(s):  Log likelihood -1835.664 Normalized cointegrating coefficients (standard error in parentheses) LIFFE_ROBUSTA_COFFEE_ DOMESTIC_SPOT SEP USD_  1.000000  6.349145  (2.59683) Adjustment coefficients (standard error in parentheses) D(LIFFE_ROBUSTA_COFFEE_SEP) -0.007838  (0.00381) D(DOMESTIC_SPOT USD_) -0.004459  (0.00280) Date: 12/10/19 Time: 19:57 Sample (adjusted): 4/02/2019 12/03/2019 Included observations: 178 after adjustments Trend assumption: Linear deterministic trend Series: LIFFE_ROBUSTA_COFFEE_NOV DOMESTIC_SPOT USD_  Lags interval (in first differences): to Unrestricted Cointegration Rank Test (Trace) Hypothesized Trace 0.05 No of CE(s) Eigenvalue Statistic Critical Value Prob.** None  0.031006  9.842474  15.49471  0.2930 At most *  0.023517  4.235958  3.841466  0.0396  Trace test indicates no cointegration at the 0.05 level  * denotes rejection of the hypothesis at the 0.05 level  **MacKinnon-Haug-Michelis (1999) p-values Unrestricted Cointegration Rank Test (Maximum Eigenvalue) Hypothesized Max-Eigen 0.05 No of CE(s) Eigenvalue Statistic Critical Value Prob.** None  0.031006  5.606516  14.26460  0.6641 At most *  0.023517  4.235958  3.841466  0.0396 189 Green financial system in Vietnam - Challenges and impacts on the economy  Max-eigenvalue test indicates no cointegration at the 0.05 level  * denotes rejection of the hypothesis at the 0.05 level  **MacKinnon-Haug-Michelis (1999) p-values  Unrestricted Cointegrating Coefficients (normalized by b'*S11*b=I):  LIFFE_ROBUSTA_COFFEE_ NOV DOMESTIC_SPOT USD_  0.009939  0.014980  0.011389 -0.013314  Unrestricted Adjustment Coefficients (alpha):  D(LIFFE_ROBUSTA_COFFEE_NOV) -3.206775 -0.689632 D(DOMESTIC_SPOT USD_) -0.538622  2.162631 Cointegrating Equation(s):  Log likelihood -1499.241 Normalized cointegrating coefficients (standard error in parentheses) LIFFE_ROBUSTA_COFFEE_ NOV  1.000000 DOMESTIC_SPOT USD_  1.507234  (0.85455) Adjustment coefficients (standard error in parentheses) D(LIFFE_ROBUSTA_COFFEE_NOV) -0.031871  (0.01393) D(DOMESTIC_SPOT USD_) -0.005353  (0.01087) Source: Calculation by authors based on data collected from Thomson Reuter and EViews 190 Vietnam National University - University of Economics and Business Appendix Vector Error Correction Model - VECM Vector Error Correction Estimates  Date: 12/10/19 Time: 20:25  Sample (adjusted): 6/05/2018 12/03/2019  Included observations: 393 after adjustments  Standard errors in ( ) & t-statistics in [ ] Cointegrating Eq:  CointEq1 LIFFE_ROBUSTA_COFFEE_JAN(-1)  1.000000 DOMESTIC_SPOT USD_(-1) -4.173551  (0.81337) [-5.13117] C  4457.532 D(LIFFE_ROBUSTA_COF- D(DOMESTIC_SPOT FEE_JAN) USD_)  0.009827  0.009583  (0.00431)  (0.00306) [ 2.27757] [ 3.13262] -0.029624  0.076761  (0.05163)  (0.03661) [-0.57379] [ 2.09697]  0.001007  0.008166  (0.07024)  (0.04980) [ 0.01434] [ 0.16397] -1.144732 -0.233011  (0.88251)  (0.62572) [-1.29713] [-0.37239]  R-squared  0.013280  0.042826  Adj R-squared  0.005670  0.035444  Sum sq resids  118414.6  59528.48  S.E equation  17.44729  12.37051  F-statistic  1.745084  5.801588  Log likelihood -1679.292 -1544.152  Akaike AIC  8.566371  7.878634  Schwarz SC  8.606817  7.919080  Mean dependent -1.109415 -0.328127  S.D dependent  17.49696  12.59575 Error Correction: CointEq1 D (LIFFE_ROBUSTA_COFFEE_JAN(-1)) D(DOMESTIC_SPOT USD_(-1)) C  Determinant resid covariance (dof adj.)  46552.98  Determinant resid covariance  45610.16  Log likelihood -3223.315 191 Green financial system in Vietnam - Challenges and impacts on the economy  Akaike information criterion  16.45453  Schwarz criterion  16.55565  Vector Error Correction Estimates  Date: 12/10/19 Time: 21:09  Sample (adjusted): 7/30/2018 12/03/2019  Included observations: 354 after adjustments  Standard errors in ( ) & t-statistics in [ ] Cointegrating Eq:  CointEq1 LIFFE_ROBUSTA_COFFEE_MAR (-1)  1.000000 DOMESTIC_SPOT USD_ (-1) -4.568667  (1.00989) [-4.52395] C  4980.191 D(LIFFE_ROBUSTA_COFFEE_MAR) D(DOMESTIC_SPOT USD_)  0.010041  0.009043  (0.00421)  (0.00310) [ 2.38309] [ 2.91480] -0.024469  0.076991  (0.05441)  (0.04006) [-0.44973] [ 1.92178]  0.011209  0.006845  (0.07153)  (0.05267) [ 0.15670] [ 0.12996] -0.957390 -0.116917  (0.93559)  (0.68889) [-1.02330] [-0.16972]  R-squared  0.015983  0.041020  Adj R-squared  0.007549  0.032800  Sum sq resids  108068.2  58590.97  S.E equation  17.57175  12.93842  F-statistic  1.894983  4.990330  Log likelihood -1514.960 -1406.604  Akaike AIC  8.581697  7.969515  Schwarz SC  8.625418  8.013236  Mean dependent -0.935028 -0.195974  S.D dependent  17.63845  13.15598 Error Correction: CointEq1 D(LIFFE_ROBUSTA_COFFEE_MAR(-1)) D(DOMESTIC_SPOT USD_(-1)) C 192 Vietnam National University - University of Economics and Business  Determinant resid covariance (dof adj.)  51614.31  Determinant resid covariance  50454.47  Log likelihood -2921.311  Akaike information criterion  16.56108  Schwarz criterion  16.67038 Source: Calculation by authors based on data collected from Thomson Reuter and EViews 193 ... Price of Liffe -Robusta Coffee Jan 2020, Price of Liffe -Robusta Coffee Mar 2020, Price of Liffe -Robusta Coffee May 2020, Price of Liffe -Robusta Coffee Jul 2020, Price of Liffe -Robusta Coffee Sept... of Augmented Dickey-Fuller Test Domestic spot, Price of Liffe -Robusta Coffee Jan 2020, Price of Liffe -Robusta Coffee Mar 2020, Price of Liffe -Robusta Coffee May 2020, Price of Liffe -Robusta Coffee. .. (i) Liffe -Robusta Coffee Jan 2020 (P1); (ii) Liffe -Robusta Coffee Mar 2020 (P2); (iii) Liffe -Robusta Coffee May 2020 (P3); (iv) Liffe -Robusta Coffee Jul 2020 (P4); (v) Liffe -Robusta Coffee Sept