SHALLOT SPATIAL MARKET INTEGRATION BETWEEN SURPLUS AND DEFICIT AREAS Текст научной статьи по специальности «Экономика и бизнес»

DOI https://doi.org/10.18551/rjoas.2020-09.18

SHALLOT SPATIAL MARKET INTEGRATION BETWEEN SURPLUS AND DEFICIT

AREAS

Olviana Tomycho*, Nendissa Doppy Roy

Department of Agribusiness, Faculty of Agriculture, University of Nusa Cendana, Indonesia

Khoiriyah Nikmatul

Department of Agribusiness, Faculty of Agriculture, University of Islam Malang, Indonesia

Sa'diyah Ana Arifatus

Department of Agribusiness, Faculty of Agriculture, University of Tribhuana Tunggadewi

Malang, Indonesia

*E-mail: [email protected]

ABSTRACT

Shallot is an important commodity in several countries so that it is a commodity that is widely traded internationally. Shallots for the Indonesian people are one of the staple foods that determine inflation. Shallot production potential for Indonesia is not evenly distributed, there are areas of surplus and there are areas of deficit so that there is an opportunity for an imbalance between supply and demand. This difference causes price disparities between surplus and deficit regions if markets are not integrated. The literature on the results of studies on spatial market integration between surplus and deficit areas in traditional markets and modern markets for shallots has not been widely found. Using time series data, shallot commodity prices for the period July 2016-May 2020. Using market prices spatially, namely prices in traditional markets and modern markets in the city of Surabaya (surplus) and prices in Kupang (deficit). Market integration analysis uses Johansen co-integration, Granger causality and VAR-VECM. The results of the study found that the Shallot price between the surplus and deficit markets are integrated into the long run but in the short run it is not perfectly integrated. There is no causality relationship between markets. Markets have a mechanism for adapting themselves to changing situations in the market. The influence of marketing infrastructure, transportation and the often uncertain drive of demand is driving the situation. Information asymmetry occurs as a result of these conditions. Policies on infrastructure improvement and market information disclosure to ensure a balance of supply and demand need to be a priority.

KEY WORDS

Shallots, surplus and deficit, cointegration, VAR-VECM.

There are many international trade transactions for the shallot commodity China controls the market share of shallots in the world. So that the price is mostly influenced by international prices other than domestic. The price of international shallots is always lower than Indonesia (Ministry of Trade of the Republic of Indonesia 2014).

Production of shallots in Indonesia still depends on the season and potential for the agro-climate, so that seasonal production is also not available in all regions in Indonesia. There are areas that are in surplus because of the potential for agro-climates, which are generally in the western part of Indonesia (such as Surabaya, the capital city of East Java Province) and many deficit areas are in eastern Indonesia, such as the province of East Nusa Tenggara (NTT). The province of NTT is known as a semi-framed archipelago which has constraints on the production and distribution of food commodities, particularly staple foods such as shallots. The main problem in NTT which is from the marketing side are the lack of food production. Shallots are always in deficit every year. So they must be supplied from to NTT, one of which is from Surabaya which is a surplus. The spatial relationship

between the two shallot markets between NTT (Kupang) and Surabaya often experiences ups and downs due to the imbalance of supply and demand, especially between the harvest season and the garden season. This imbalance causes price fluctuations. Speculators have the opportunity to use erratic fluctuations in prices by sett prices (acting as price makers), while traders in traditional and modern markets and consumers are only price takers. The consequence of rising food prices can have an impact on increasing poverty, (Sa'diyah et al, 2019).

These considerations, research on the spatial integration of the shallot market between surplus and deficit areas is important to do. Several literature findings from market integration study using the Johansen cointegration approach, Granger causality and VECM were carried out by Ghafoor and Aslam, 2012 on the rice market in Pakistan; Traub et al, 2014 South African and Mozambique maize markets; Akhter 2017, on rice markets in India, Bangladesh and Nepal; Roman 2020, the dairy market in Poland; Ozturk 2020, on the wheat market in Turkey; Nigatu and Adjemian 2020, on the markets for corn, soybeans and cotton between U.S. and international prices. A study conducted in Indonesia by Cahyaningsih et al. (2012), on the Indonesian rice market; Irawan, and Rosmayanti 2016, regarding the rice market in Bengkulu; Hanani et al, 2020, about the cayenne pepper market in Malang, Indonesia. The study found varied conclusions, depending on the types of commodities, policies of each country, geographical conditions, the behaviour of marketing institutions and the sociopolitical and infrastructure situations.

Information about the dynamics of spatial market integration between supposes areas and the deficit of shallots on prices in traditional and moderate markets have not been widely found. A spatially integrated market, if prices in one market are related to price movements away a positive way, it follows the law of one price "LOP" (the law of one price), so that the market is said to be efficient because every market participant gets a fair profit. The marketing system is unfair if the opposite happen, it is not integrated; therefore, a market position experiencing a deficit has the potential to come under pressure and losses from a surplus market. The results of the study on spatial market integration and price fluctuations of shallots are very useful for policy makers and those with an interest in predicting future price dynamics so as to anticipate the risk of large losses.

This study uses price data presented by the data centre of national strategic food prices for shallot commodities in traditional and modern markets. The data form is the monthly time series for the period June 20160-May 2020. The data objects include markets in areas where shallots are surplus and deficit, namely Surabaya and Kupang. East Java Province is the largest province with the head as a centre of shallot production of Indonesia. Meanwhile, NTT is the province with the most deficit (BPS, 2017). The price of shallots in East Java uses the price in (City) Surabaya and NTT uses (City) Kupang. The two cities are provincial capitals which are the centre of trade transactions.

Spatial Market Integration Analysis Method. Data analysis method to see the level of spatial market integration in traditional and modern markets between the surplus and deficit areas of garlic, using the Vector Autoregression (VAR) - Error Correction Model (VECM) model. The formation of the VAR-VECM model includes stationary and cointegration problems. The stages of finding the VAR-VECM model, in the analysis of market integration are shown in Figure 1.

The data stationary test uses the Augmented Dickey-Fuller (ADF) approach at a price level or different to get stationary data (Nendissa et al, 2018; Kapioru 2020), that is, the variance has a tendency to approach its average value (Enders, 1995). ADF tested to see the trend of shallot price data movements, with a formula:

METHODS OF RESEARCH

m

I=1

Where:

Pt = Price of shallots in each market in period t (Rp / kg); Pt-1 = The price of shallots in each market in the previous t period (Rp / kg) ; APt = Pt - Pt-i; APt-i =Pt_i -AP(t_i)_i; m = amount of lag; a0 = intersep;

a, p, y = Parameter coefficient; £t = Error term.

Time Series data Monthly prices (July 2016-May 2020)

I

Data Stationary Test: Root Test Analysis

I

Stationary at Level

I

I

ADF Test

Not Stationary at Level

VAR in Level

Stationary Test at Differential level

z

Stationary Test at the same Differential Level

VAR in Differences

Not Cointegration

H-

I

Cointegration Test (Johansen's Cointegration)

Granger Causality Test and Analysis

^ There is cointegration

Vector Error Corection Model (VECM)

I

Policy Recommendations

Figure 1 - Stages of VAR-VECM Model Formation (modification from Widarjono 2018; Basuki and Prawoto, 2017)

Hypothesis test:

• H0: y = 0 time series data is not stationary;

• Hx: y <0 stationary time series data. Testing criteria:

1. If the ADF statistical ADF is critical, then reject Ho, meaning that the price data is not stationary;

2. If the ADF statistic < ADF is critical, then accept Ho, it means that the price data is stationary.

Determination of Optimal Lag. The optimal lag length is needed to see the effect of each variable on other variables in the VAR model, using the Akaike Information Criteria (AIC). The criterion that has the smallest AIC and SIC values is the lag used.

Cointegration Test. The cointegration test uses Johansen's Cointegration Test, to find out whether there is integration or not (Hjalmarsson & Osterholm, 2010; Mensah et al, 2017; Naidu et al, 2017; Tursoy, 2019). The long run equation is defined as follows:

Y = C + P1X1 + P2X2 + P3X3 + ... + PXn + £

Where: Y = dependent variable; C = constant; fi = estimate value; X = independent variable; £ = residual.

If in the test, there is no cointegration relationship, the analysis is carried out using the VAR difference method, and if there is a cointegration relationship, VECM analysis is carried out using the Johansen test.

Granger Causality Test. The Granger causality test is used to see the short-term causality of each variable that has root and is co-integration (Bhutto et al, 2020; Higgoda, R & Madurapperuma, 2020; Rizwanullah et al, 2020; Hu et al, 2020; Mohamed, 2020). The Granger causality test is a statistical hypothesis test to determine whether one time series is useful in predicting another (Hood et al, 2008; Fanchette et al, 2019; Ptackova et al, 2019; Plub-in, & Songsiri, 2019; Zhang et al, 2020).

VECM tests. The VECM model is used to overcome data instability, where this model will gradually correct the imbalance, deviation through short-term partial adjustment (Enders, 1995; and Gujarati, 2004; Nkalu et al, 2020; Ters, & Urban, 2020; Molin, 2020; Giudici & Pagnottoni, 2020). The general form of VECM (p) where p is the lag of the endogenous variable with cointegration rank r < k is as follows:

Ayt = nyt_! + If-1 r Ayt_! + Dt + Et

Where:

• A = differencing operators, with Ayt = yt - yt-1;

• yt-1 = endogenous variable vector with lag to-1;

• £t = error vector with size (k x 1);

• Dt = constant vector with size (k x 1);

• n = cointegration coefficient matrix with n - apf;

• a = vector adjustment, size matrix (k x r);

• p = cointegration vector with size matrix (k x r);

• ri = coefficient matrix (k x k) the coefficient of endogenous variables k-i (Lutkepohl, 2005).

Several stationary exogenous variables can be included as additional regressors along with some of the lags with the following equation:

Ayt = nyt_i + I-1 r Ayt_i + O Xt_i + Dt + Et

Where:

• A = Differencing operators, with Ayt = yt - yt_1;

• yt_1 = endogenous variable vector with lag to-1;

• Et = error vector with size (k x 1);

• Dt = constant vector with size (k x 1);

• n = cointegration coefficient matrix with n - apf;

• a = vector adjustment, size matrix (k x r);

• p = cointegration vector with size matrix (k x r);

• ri = coefficient matrix (k x k) endogenous variable coefficient to-i;

• O = coefficient vector (1 x k) exogenous variables to-i.

The VECM model used in this study is as follows:

p p

APKat = Go + ^ a,APKat_1 + ^ p,APKot _1 + E1t

1=1 1=1

pp

APKot = So + ^ S APKat_1 + ^ o, APKot _1 + E1t

1=1 1=1

pp

APJkt = 0o + ^ 0, APKat_1 + ^ W, APKot_1 + E1t

1=1 1=1

Where:

• APSat = The price of shallots in the modern market of Surabaya in the period to — t (Rp/Kg);

• APKat.1 = The price of shallots in the modern Kupang market in period to-t, before (Rp/kg);

• APSot =Price of shallots in Surabaya traditional markets period to-t (Rp/Kg);

• APKot—! = Price of shallots in Kupang traditional markets, period to — t, before (Rp/Kg);

• a, 5,6, p, y, z, w, 9 = regression coefficient;

• £it = error term ke — i, waktu ke — t.

RESULTS AND DISCUSSION

Shallot production of Indonesia is concentrated on the island of Java, because it generally has very good agro-climatic conditions and land for shallot cultivation. East Java Province is the second largest shallot producing province after Central Java, followed by West Java, then NTB and others outside Java (BPS 2017). From the aspect of total production, East Java experiences an overproduction (surplus) almost every year, so that East Java becomes one of the main suppliers of shallots to deficit areas, one of which is NTT.

Data Stationarity Test. The results of the stationary test based on the ADF test or the DF test of table 1, showed that data is not stationary at the level or integration level of zero, I(0), so the time-series economic model stationary requirements can be obtained by difference data at the 1st difference level.

Table 1 - Stationarity test results using the ADF Test approach

Unit Root Test

Variable Level st 1 Difference Information

ADF Prob. ADF Prob.

Modern_Kupang Traditional _Kupang Modern Surabaya -2.521976 -2.315400 -2.122807 0.1118 0.1682 0.2360 -8.859378 -12.07558 -15.89345 * * * o o o o o o 000 .0 .0 .0 0. 0. 0. Stationary Stationary Stationary

Traditional Surabaya -2.134123 0.2316 -14.31119 0.0000* Stationary

Source: Secondary data, 2020.

Determination of Lag Length. The variable data is stationary at the 1st Difference level, so the estimation is expected to produce a valid model output. VAR model estimation starts with determining the appropriate lag length of the VAR model.

Table 2 - Results of the Optimal Lag Determination Test

Lag LogL LR FPE AIC SC HQ

0 -6961.317 0.00000 8.17e+26 73.31912 73.38748 73.34682

1 -6829.552 256.5939 2.42e+26 72.10055 72.44234 72.23901

2 -6786.743 81.56336 1.82e+26 71.81835 72.43357* 72.06756

3 -6754.126 60.77037 1.53e+26* 71.64343* 72.53209 72.00341*

4 -6738.721 28.05239* 1.54e+26 71.64970 72.81179 72.12045

5 -6724.096 26.01755 1.57e+26 71.66417 73.09970 72.24568

6 -6714.544 16.59016 1.68e+26 71.73205 73.44101 72.42432

7 -6706.826 13.07976 1.84e+26 71.81923 73.80162 72.62226

8 -6696.428 17.18454 1.96e+26 71.87819 74.13402 72.79199

The result of the lag length test of VAR by entering AIC shows the optimal lag length is 3 with an AIC value of 71.64343*.

Cointegration tests. The results of the Johansen cointegration test of Table 3 show that the value of Trace Statistics> Critical values / valued with Prob <0.05 means that there is an integration of shallot prices in the long run for / of the Kupang modern and traditional Kupang markets.

Table 3 - Johansen Cointegration Test between Markets in Modern Kupang and

Traditional Kupang markets

Hypothesized No. of CE(s) Trace Statistic 0.05 Critical Value Prob.** Max-Eigen Statistic 0.05 Critical Value Prob.**

None * 21.70332 15.49471 0.0051 18.15112 14.26460 0.0116

At most 1 3.552196 3.841466 0.0595 3.552196 3.841466 0.0595

Source: Secondary data processed, 2020. Error level (a) = 0.05.

Changes in prices for modern Kupang and traditional Kupang have a relationship in the long term, but in the short term it is unlikely to happen.

Table 4 - Johansen Cointegration Test between modern Kupang Markets and Modern Surabaya

Hypothesized No. of CE(s) Trace Statistic 0.05 Critical Value Prob.** Max-Eigen Statistic 0.05 Critical Value Prob.**

None * 20.54135 15.49471 0.0080 16.70772 14.26460 0.0201

At most 1 3.833626 3.841466 0.0502 3.833626 3.841466 0.0502

Source: Secondary data processed, 2020. Error level (a) = 0.05.

The results of the Johansen cointegration test in Table 4 show that the value of Trace Statistics> Critical value with Prob <0.05 means that there is an integration of shallot prices in the long run between the modern Kupang market and modern Surabaya.

Table 5 - Johansen cointegration test between Kupang traditional markets and traditional Surabaya

Hypothesized No. of CE(s) Trace Statistic 0.05 Critical Value Prob.** Max-Eigen Statistic 0.05 Critical Value Prob.**

None * 44.82258 15.49471 0.0000 38.23987 14.26460 0.0000

At most 1 6.582708 3.841466 0.0103 6.582708 3.841466 0.0103

Source: Secondary data processed, 2020. Error level (a) = 0.05.

The results of the Johansen cointegration test of Table 5 show that the value of Trace Statistics> Critical values with Prob <0.05 means that there is an integration of shallot prices in the long term between traditional Kupang markets and traditional Surabaya. The results of the research by Kapioru, et al. (2020) between prices in traditional markets and prices at wholesalers in NTT in the long term that integration occurs.

Table 6 - Johansen cointegration test for modern Surabaya markets and traditional Surabaya

Hypothesized No. of CE(s) Trace Statistic 0.05 Critical Value Prob.** Max-Eigen Statistic 0.05 Critical Value Prob.**

None * 18.98671 15.49471 0.0143 14.50852 14.26460 0.0458

At most 1 4.478190 3.841466 0.0343 4.478190 3.841466 0.0343

Source: Secondary data processed, 2020. Error level (a) = 0.05.

The results of the Johansen cointegration test of Table 6 show that the Trace Statistics> Critical values with Prob <0.05 means that there is an integration of shallot prices in the long run for the modern Surabaya market and traditional Surabaya.

Granger Causality. Test The causality test is to see the reciprocal relationship between prices spatially between surplus and deficit areas in traditional and modern markets.

Table 6 - Granger Causality Test, between markets in surplus and deficit areas

|Null Hypothesis: Obs. F-Statistic Prob. |

MODERN SURABAYA doesn't Granger Cause MODERN KUPANG 197 2.43531 0.0903

MODERN KUPANG doesn't Granger Cause MODERN SURABAYA 102.203 6.E-31

TRADITIONAL KUPANG doesn't Granger Cause MODERN KUPANG MODERN KUPANG does not Granger Cause TRADITIONAL KUPANG 197 9.37923 0.01526 0.0001 0.9849

TRADITIONAL SURABAYA doesn't Granger Cause MODERN KUPANG 197 7.53351 0.0007

MODERN KUPANG does not Granger Cause TRADITIONAL SURABAYA 0.00267 0.9973

TRADITIONAL KUPANG doesn't Granger Cause MODERN SURABAYA MODERN SURABAYA doesn't Granger Cause TRADITIONAL KUPANG 197 7.38170 0.83674 0.0008 0.4347

TRADITIONAL SURABAYA does not Granger Cause MODERN SURABAYA 197 6.90459 0.0013

MODERN SURABAYA doesn't Granger Cause TRADITIONAL SURABAYA 1.31295 0.2714

TRADITIONAL SURABAYA doesn't Granger Cause TRADITIONAL KUPANG 197 0.46069 0.6315

TRADITIONAL KUPANG doesn't Granger Cause TRADITIONAL SURABAYA 157.160 4.E-41

Source: Secondary data processed, 2020. Error level (a) = 0.05.

The results of the Granger causality test, which are presented in table 6, show that spatially there is a two-way causality relationship that only occurs to modern markets in Kupang and modern markets in Surabaya. Meanwhile, the relationship between other markets is only a one-way relationship. This gives an indication that the relationship between markets does not influence each other, except in modern markets in areas of surplus and deficit. Figure 2 provides an illustration of the causality relationship.

Figure 2 - Illustration of Causality Relationship between Surplus (Surabaya) and Deficit (Kupang) Areas in Modern and Traditional Markets.

The existence of a causal relationship between modern markets in surplus and deficit areas indicates that the modern markets in Surabaya and Kupang are interrelated, because generally shallots sold in modern markets in Kupang are from their business partners in Surabaya. A modern market in the form of a Hypermart which has business partners with Surabaya. So that any price changes that occur to the Surabaya modern market are transmitted directly to the Kupang modern market, because the market information system is running fast. Whereas the other three markets relationships do not have a causal relationship, there is a possibility of information asymmetry, market infrastructure constraints and inefficient demand and supply mechanisms. This is supported by the results of the study (Goletti et al., 1995; Suryana et al., 2014), the absence of spatial integration suggests that price changes in one producer market are not reflected as price changes in geographically different producer markets. Other reasons for the absence of market integration are the distance between cities, infrastructure, road transport flows by Varela et al. (2012); and Hidayanto et al, 2014). The effect of commodity trading policies by Sexton et al, (1991) and Aryani (2009). There is no comprehensive integration in all markets in the long term (Adiyoga et al, (1999).

This means that the pressure on the demand for shallots in the traditional Kupang market as a deficit area becomes a stimulus for prices that are formed in the market surplus in the short term. This condition is often used by inter-island wholesalers to determine prices. The findings of a study by Kapioru et al, (2020) on the commodity of red chili in NTT show that price changes in traditional markets and prices in wholesalers do not have a causal relationship. Changing prices at wholesalers prompted red chili prices in traditional markets to change, but the reverse did not happen. The findings of Akhter's (2017) study show that the domestic rice prices of India, Bangladesh and Nepal are integrated in the short and long term, even though India has imposed export restriction policies. Akhter suspects that the price of rice is being transmitted effectively because of the informal cross-border trade that extends over the porous border between India, Bangladesh and Nepal.

Vector Autoregression (VAR) tests. The vector auto regression_(VAR) test is intended because the data is not stationary which has been derived based on the optimal lag, namely the optimal lag 4. The cointegration test variable data shows that there is no cointegration

between all price variables in the three markets, this indicates that there is no relationship or imbalance in modern Kupang market, traditional Kupang, modern Surabaya and traditional Surabaya in the long run. However, in the short term there is a possibility of integration. So it is necessary to apply a VAR / VARD test approach. The VAR / VARD test results are displayed in appendix 1. The estimation results based on the VARD model in the attachment to Table 1, show that in the short term changes in the price of shallots in the modern Kupang market are significantly influenced by prices in the market, modern Surabaya one month earlier was 1.321, modern Surabaya two months earlier was 0.800 and modern Kupang itself two months earlier at - 2,806. This value indicates that each increase in the price of modern market shallots in Surabaya one month earlier, modern Surabaya two months earlier and modern Kupang two months earlier by 1%, It will increase prices in the modern market in Kupang in the current period respectively by 1.321% and 0.800% and There was a decline in the price of shallots in the previous two months in the modern Kupang market by 0.2806. Meanwhile, the price of shallots in the Kupang traditional market was significantly influenced by the price of the traditional Surabaya market in the previous month of 1.084, two months earlier of 0836, the previous three months of 0.343. The previous month and the previous three months by 1%, It will reduce the market price in Surabaya traditional markets in the current period respectively by 1.084%, 0.836% and 0.343%.

The estimation results based on the VARD model in the appendix to Table 1, show that in the short term changes in the price of shallots in the modern Kupang market are significantly influenced by prices in the modern market of Surabaya one month earlier of 1.321, modern Surabaya two months earlier of 0.800 and modern Kupang itself two the previous month was -2,806. This value indicates that each increase in the price of modern market shallots in Surabaya one month earlier, modern Surabaya two months earlier and modern Kupang two months earlier by 1%, It will increase prices in the modern market in Kupang in the current period respectively by 1.321% and 0.800% and There was a decline in the price of shallots in the previous two months in the modern Kupang market by 0.2806. Meanwhile, the price of shallots in the Kupang traditional market was significantly influenced by the price of the traditional Surabaya market in the previous month of 1.084, two months earlier of 0836, the previous three months of 0.343. The previous month and the previous three months by 1%, It will reduce the market price in Surabaya traditional markets in the current period respectively by 1.084%, 0.836% and 0.343%.

Changes in the price of shallots in Surabaya traditional markets are significantly influenced by prices in Surabaya traditional market itself one month before of - 0,912, two months earlier of - 0,581 and three months earlier of - 0,226, In traditional markets of Surabaya one month earlier of 1,047, meaning that each the increase in prices in the traditional market in Surabaya in one month, two months earlier and three months earlier by 1% will reduce prices in the traditional market in Surabaya in the current period by 0.912%, 0.581% and 0.226%. Meanwhile, price changes in the modern market in Surabaya were significantly influenced by the price of the modern market in Surabaya itself one month earlier, which was - 0,989, In the previous two months it was - 0.671, meaning that each price increase in the modern Surabaya market in modern Surabaya in the current period was 0.989% and 0.671%.

Vector Error Correction Model (VECM) tests. The VECM test results from Appendix 2, show that ECT at traditional prices in Surabaya is significant at the 5% error level, namely -0.095 (-0.4752 <-.97196). Significant ECT values ''indicate that the importance of long-term cointegration relationships in the process of forming shallot prices among market players. Price changes are influenced by the long-term relationship between modern Kupang market and modern Surabaya and modern Surabaya and modern Surabaya itself one month before and two months before. In the short term, changes in the modern market prices in Kupang are only influenced by changes in prices in the modern market in Surabaya one month earlier at 1,309. This value indicates that each 1% increase in prices in the modern market in Surabaya in the previous month will increase the modern market price of Kupang in the current period by 1.309%. Meanwhile, the short-term change in the price of shallots in the modern market in Surabaya was influenced by the price of shallots in the modern Surabya

market itself in the previous month and in the previous two months of - 0.978 and - 0.661. This means that every 1% increase will reduce the price of shallots in the modern Surabaya market in the current period by 0.978% and 0.661%, respectively. Thus, the transmission mechanism and market integration do not always show a consistent pattern. Every market has a mechanism for responding to changes, because many factors determine the effectiveness of spatial market integration. Several studies on the determinants of market integration found that.

CONCLUSION

The price relationship to between traditional and modern markets between the surplus and deficit regions is not perfect. The price relationship between all the two markets is only a one-way relationship. This means that there is no integration between markets, except between modern markets in surplus areas (Surabaya) and deficits (Kupang) where there is a causal relationship. This is because the modern markets in the two regions are trading partners that have a perfect flow of information so that the law of one price (LOP) applies. Meanwhile, price relationships between other markets, namely between traditional markets, between traditional and modern markets, do not have a causal relationship. The role of traders in surplus areas is in taking advantage of the imbalance opportunity in supply and fever to regulate the price of shallots in deficit areas. Marketing infrastructure constraints, asymmetric distance information and transportation are the determinants of spatial market integration. Improvement on market infrastructure, information disclosure and compliance with prices as well as maintaining the security of shallot stock need to be a joint commitment, especially policy makers.

ACKNOWLEDGMENTS

Thanks to LPPM, University of Nusa Cendana for their support in presenting this paper.

APPENDIX

APPENDIX 1 - VAR tests / tested for the VARP test, in Kupang Modern Market, Kupang Traditional, Modern Surabaya and Surabaya Traditional Markets

D(MODERN D(TRADITIONAL D(TRADITIONAL D(MODERN

KUPANG) KUPANG) SURABAYA) SURABAYA)

D(MODERN_KUPANG(-1)) 0.516052 0.049767 0.020684 1.321149

(0.08061) (0.10012) (0.08252) (0.11488)

6.40160] 0.49706] 0.25067] ' 11.5007]

D(MODERN_KUPANG(-2)) -0.280567 -0.025789 0.002794 0.800895

(0.12185) (0.15135) (0.12473) (0.17365)

-2.30247] -0.17040] 0.02240] ' 4.61222]

D(MODERN_KUPANG(-3)) -0.133048 0.138983 0.075811 0.324754

(0.14402) (0.17887) (0.14741) (0.20523)

-0.92385] 0.77700] 0.51427] ' 1.58242]

D(TRADITIONAL _KUPANG(-1)) 0.112265 0.298725 1.084834 0.102750

(0.07656) (0.09509) (0.07837) (0.10911)

1.46629] 3.14134] 13.8423] ' 0.94174]

D(TRADITIONAL KUPANG(-2)) 0.075561 0.173531 0.836299 0.099735

(0.10900) (0.13538) (0.11157) (0.15532)

0.69325] 1.28185] 7.49591] ' 0.64212]

D(TRADITIONAL _KUPANG(-3)) 0.124265 -0.110968 0.343499 0.254974

(0.10390) (0.12904) (0.10635) (0.14805)

1.19605] -0.85994] 3.22996] ' 1.72216]

D(TRADITIONAL SURABAYA(-1)) -0.032454 -0.177284 -0.912336 -0.131568

(0.08997) (0.11175) (0.09209) (0.12821)

-0.36072] -1.58649] -9.90661] -1.02618]

D(TRADITIONAL SURABAYA(-2)) -0.097479 0.032403 -0.581632 -0.227858

(0.09993) (0.12412) (0.10229) (0.14240)

-0.97546] 0.26107] -5.68616] -1.60009]

D(TRADISIONAL_SURABAYA(-3)) 0.133185 -0.009370 -0.226967 -0.020606

(0.06879) (0.08544) (0.07041) (0.09803)

1.93613] -0.10967] -3.22338] -0.21020]

D(MODERN_SURABAYA(-1)) -0.026645 -0.067126 -0.065396 -0.989803

(0.05665) (0.07036) (0.05798) (0.08072)

-0.47036] -0.95406] -1.12782] -12.2615]

D(MODERN_SURABAYA(-2)) 0.231783 -0.121669 -0.092054 -0.671188

(0.11710) (0.14544) (0.11986) (0.16687)

1.97940] -0.83657] -0.76800] '-4.02229]

D(MODERN_SURABAYA(-3)) 0.026880 -0.023746 0.039481 -0.177327

(0.10111) (0.12558) (0.10350) (0.14409)

0.26585] -0.18909] 0.38147] -1.23070]

C -22.67957 61.54727 33.73261 -156.2568

(124.415) (154.527) (127.351) (177.294)

-0.18229] 0.39829] 0.26488] '-0.88134]

Source: Secondary data processed, 2020. T-table at 5% error level = 1.97196.

APPENDIX 2 - VECM Test Results on Kupang Modern, Traditional Kupang, Modern Surabaya and

Traditional Surabaya Markets

D(MODERN D(TRADITIONAL D(TTRADITIONAL D(MODERN

Error Correction: KUPANG) KUPANG) SURABAYA) SURABAYA)

CointEq1 0.023610 5.012091 ■0.095912 0.009901

(0.02084) 0.02595) 0.02018) 0.02979)

■ 1.13317] 0.46586] -4.75283] 0.33238]

D(MODERN KUPANG(-1)) 0.487117 0.034949 0.138233 1.309015

(0.08450) 0.10526) 0.08184) 0.12081)

■ 5.76464] 0.33202] 1.68898] 10.8357]

D(MODERN KUPANG(-2)) -0.303530 ■0.037549 0.096082 0.791265

(0.12343) 0.15376) 0.11955) 0.17647)

-2.45904] -0.24420] 0.80368] 4.48394]

D(MODERN KUPANG(-3)) -0.149351 0.130635 0.142038 0.317917

(0.14462) 0.18015) 0.14007) 0.20675)

-1.03271] 0.72514] 1.01403] 1.53766]

D(TRADITIONAL KUPANG(-1)) 0.339516 0.415103 0.161646 0.198046

(0.21464) 0.26738) 0.20789) 0.30686)

■ 1.58178] 1.55251] 0.77754] 0.64540]

D(TRADITIONAL KUPANG(-2)) 0.232929 0.254121 0.197008 0.165726

(0.17649) 0.21985) 0.17094) 0.25231)

■ 1.31981] 1.15591] 1.15252] 0.65683]

D(TRADITIONAL KUPANG(-3)) 0.201604 ■0.071362 0.029317 0.287406

(0.12424) 0.15476) 0.12033) 0.17762)

■ 1.62269] -0.46111] 0.24363] 1.61811]

D(TRADITIONAL SURABAYA(-1)) -0.234825 ■0.280920 ■0.090221 ■0.216430

(0.19994) 0.24906) 0.19365) 0.28584)

-1.17448] -1.12792] -0.46589] -0.75717]

D(TRADITIONAL SURABAYA(-2)) -0.223393 ■0.032079 ■0.070114 ■0.280660

(0.14939) 0.18609) 0.14469) 0.21357)

-1.49536] -0.17238] -0.48457] -1.31411]

D(TRADITIONAL SURABAYA(-3)) 0.086776 ■0.033137 ■0.038433 ■0.040067

(0.08001) 0.09967) 0.07750) 0.11439)

■ 1.08453] -0.33247] -0.49593] -0.35027]

D(MODERN SURABAYA(-1)) -0.000331 ■0.053650 ■0.172295 ■0.978768

(0.06118) 0.07621) 0.05926) 0.08747)

-0.00541] -0.70395] -2.90753] -11.1901]

D(MODERN SURABAYA(-2)) 0.254861 ■0.109851 ■0.185805 ■0.661511

(0.11877) 0.14794) 0.11503) 0.16979)

■ 2.14592] -0.74252] -1.61526] -3.89602]

D(MODERN SURABAYA(-3)) 0.039923 ■0.017067 ■0.013503 ■0.171858

(0.10169) 0.12667) 0.09849) 0.14537)

■ 0.39261] -0.13474] -0.13710] -1.18218]

C -25.61046 50.04633 45.63907 ■157.4859

(124.345) 154.894) 120.435) 177.768)

-0.20596] 0.38766] 0.37895] -0.88591]

R-squared 0.303739 0.102978 0.592736 0.566160

Adj. R-squared 0.253732 0.038551 0.563485 0.535000

Sum sq. resids 5.43E+08 8.42E+08 5.09E+08 1.11E+09

S.E. equation 1731.345 2156.700 1676.909 2475.191

F-statistic 6.073859 1.598364 20.26375 18.16959

Loq likelihood -1723.476 ■1766.314 ■1717.247 ■1793.173

Akaike AIC 17.82027 18.25963 17.75638 18.53511

Schwarz SC 18.05526 18.49462 17.99136 18.77009

Mean dependent -5.128205 33.07692 78.20513 ■6.410256

S.D. dependent 2004.177 2199.513 2538.106 3629.797

Determinant resid covariance (dof adj.) 9.78E+25

Determinant resid covariance 7.26E+25

Loq likelihood ■6912.557

Akaike information criterion (AIC) 71.51340

Schwarz criterion 72.52048

Source: Secondary data processed, 2020. T-table at 5% error level = 1.97196.

REFERENCES

1. Adiyoga W, Ameriana M, Hidayat A. 1999. Segmentasi dan Integrasi Pasar: Studi Kasus dalam Sistem Pemasaran Bawang Merah. Jurnal Hortikultura. 9(2):153-163.

2. Akhter, S. (2017). Market Integration Between Surplus and Deficit Rice Markets During Global Food Crisis Period. Australian Journal of Agricultural and resource economics, 61(1), 172-188.

3. Aryani D. 2009. Integrasi Pasar Beras dan Gula di Thailand, Filipina dalndonesia. Tesis. Sekolah Pascasarjana Institut Pertanian Bogor.

4. Basuki, A. T dan Prowoto, N. 2017. Analisis Regresi, dalam Penelitian Ekonomi dan Bisnis, (dilengkapi Aplikasi SPSS dan Eviews). Penerbit Tajawali Pers. Jakarta.

5. Baulch, B. (1997). Transfer costs, spatial arbitrage, and testing for food market integration. American Journal of Agricultural Economics, 79(2), 477-487.

6. Bhutto, S. A., Rajper, Z. A., & Kishan, J. (2020). The Essentials of Financial Policies and Interest Rate Shocks in Downturn and Upswing of Stock Market: A Cointegration and Causality Analysis. International Journal of Psychosocial Rehabilitation, 24(07).

BPS, 2017. Statistik Harga Komoditi Pertanian Tahun 2018. Badan Pusat Statistik RI. Jakarta

7. Cahyaningsih E., Nurmalina R., & Maulana A. 2012. Integrasi spasial dan vertikal pasar beras di Indonesia. PANGAN. 21(4):317-332.

8. Cermeño, A. L., & Santiago-Caballero, C. (2020). All roads lead to market integration. Lessons from a spatial analysis of the wheat market in 18th century Spain. Working Papers in Economic History, (2020-02).

9. Enders, W. 1995. Applied Applied Econommetric Time Series., New York: Jonh Wiley&Sons. Inc. Gokta§, 0.(2005) Teorik ve Uygulamali Zaman Serileri Analizi, istanbul: Be§ir Kitabevi

10. Faminow, M. D., & Benson, B. L. (1990). Integration of spatial markets. American Journal of Agricultural Economics, 72(1), 49-62.

11. Fanchette, Y., Ramenah, H., Casin, P., Benne, M., Tanougast, C., & Adjallah, K. (2019, September). Predictive Causality of Granger Test for Long Run Equilibrium to Photovoltaic System. In 2019 10th IEEE International Conference on Intelligent Data Acquisition and Advanced Computing Systems: Technology and Applications (IDAACS) (Vol. 2, pp. 942-946). IEEE.

12. Ghafoor, A & Aslam, M. (2012). Market Integration ad Price Transmission in Rice Markets of Pakistan. SANEI Working Paper Series No 12-08.

13. Giudici, P., & Pagnottoni, P. (2020). Vector error correction models to measure connectedness of Bitcoin exchange markets. Applied Stochastic Models in Business and Industry, 36(1), 95-109.

14. Gitau, R., & Meyer, F. (2018). Spatial market integration in the era of high food prices. A case of surplus and deficit maize markets in Kenya. Agrekon, 57(3-4), 251-265.

15. Goletti F dan Christina-Tsigas E. 1995. Analizing Market Integration. Scott, Gregory J. Editor. Price Product and People, Analizing Agricultural Market in Developing Countries. Lynne Rienner Publisher. London.

16. Goletti F, Raisuddin A dan Farid N. 1995. Structural Determinants of Market Integration: the Case of Rice Markets in Bangladesh. The Developing Economies, Vol 33(2); pp:185-202.

17. Goodwin, B. K., & Piggott, N. E. (2001). Spatial market integration in the presence of threshold effects. American Journal of Agricultural Economics, 83(2), 302-317.

18. Gujarati, D. 2004. Basics Econometrics, Fourth Edition. New York (US): The McGraw Hill Companies.

19. Hanani, A. A., Anindita, R., & Mutisari, R. 2020. Analysis of Market Integration Cayenne Pepper (Capsicum frutescens L.) in Malang District. Agricultural Socio-Economics Journal, 20(1), 23-30.

20. Hidayanto MW, Anggraeni L, Hakim DB. 2014. Faktor penentu integrasi pasar beras di Indonesia. Pangan. 23(1):1-16.

21. Higgoda, R., & Madurapperuma, W. (2020). Air passenger movements and economic growth in Sri Lanka: Co-integration and causality analysis. Journal of Transport and Supply Chain Management, 14, 13.

22. Hjalmarsson, E., & Osterholm, P. (2010). Testing for cointegration using the Johansen methodology when variables are near-integrated: size distortions and partial remedies. Empirical Economics, 39(1), 51-76

23. Hood III, M. V., Kidd, Q., & Morris, I. L. (2008). Two sides of the same coin? Employing Granger causality tests in a time series cross-section framework. Political Analysis, 324344.

24. Hu, Y., Hou, Y. G., & Oxley, L. (2020). What role do futures markets play in Bitcoin pricing? Causality, cointegration and price discovery from a time-varying perspective?. International Review of Financial Analysis, 101569.

25. Irawan, A., & Rosmayanti, D. (2016). Analisis integrasi pasar beras di Bengkulu. Jurnal Agro Ekonomi, 25(1), 37-54.

26. Ismet M, Barkley AP dan Llewelyn RV. 1998. Government Intervention and Market Integration in Indonesian Rice Markets. Agricultural Economics, Vol 19(3); pp:283-295

27. Kapioru, Ch., Bano, M., & Nendissa, D. R. (2020). Market Cointegration and Red Chili Price Behavior Between Wholesalers and Traditional Markets, Russian Journal of

Agricultural Soscial-Economic Sciences; RJOAS, 8(104), August 2020; DOI 10.18551/rjoas.2020-08.20.

28. Kementerian Perdagangan. 2010. Rencana Strategis Kementerian Perdagangan Periode 2010-2014. Jakarta (ID): Kementrian Perdagangan.

29. Kementerian Perdagangan, (2014). Outlook Pangan Tahun 2015-2019, Laporan Ringkas. Pusat Kebijakan Perdagangan Dalam Negeri, Badan Pengkajian dan Pengembangan Kebijakan, Kemendag.

30. Lutkepohl, H. 2005. New Introduction to Multiple Time Series Analysis. Springer Science & Business Media.

31. Mensah, L., Obi, P., & Bokpin, G. (2017). Cointegration test of oil price and us dollar exchange rates for some oil dependent economies. Research in International Business and Finance, 42, 304-311.

32. Mohamed, S. E. (2020). Energy Consumption, CO2 Emissions and Economic Growth Nexus in Oman: Evidence from ARDL Approach to Cointegration and Causality Analysis. European Journal of Social Sciences, 60(2), 67-78.

33. Molin, S. (2020). House Price Dynamics in Sweden: Vector error-correction model.

34. Naidu, S., Pandaram, A., & Chand, A. (2017). A Johansen cointegration test for the relationship between remittances and economic growth of Japan. Modern Applied Science, 11(10), 137-151.

35. Nendissa, D. R., Anindita, R., Nuhfil, H., & Wahib, M. A. (2018). Beef Market Integration in East Nusa Tenggara of Indonesia. Russian Journal Of Agricultural Socio-Economic Science (RJOAS), 8(80), August 2018. ISSN2226-1184 (online). DOI: https://doi.org/10.18551/rjoas.2018-08.51.

36. Nendissa, D. R., Olviana., T & Kapioru, Ch. (2020) The Impact of the Covid-19 Pandemic on Price Disparities and Fluctuations of Shallots in Traditional Markets; Russian Journal Of Agricultural Soscial-Economic Sciences; Rjoas, 7(103), July 2020. DOI 10.18551/rjoas.2020-07.14.

37. Nigatu, G., & Adjemian, M. (2020). A Wavelet Analysis of Price Integration in Major Agricultural Markets. Journal of Agricultural and Applied Economics, 52(1), 117-134

38. Nkalu, C. N., Ugwu, S. C., Asogwa, F. O., Kuma, M. P., & Onyeke, Q. O. (2020). Financial Development and Energy Consumption in Sub-Saharan Africa: Evidence From Panel Vector Error Correction Model. SAGE Open, 10(3), 2158244020935432.

39. Ozturk, O. 2020. Market Integration and Spatial Price Transmission in Grain Markets of Turkey. Applied Economics, 1-13.

40. Ptackova, V., Stepanek, L., & Hanzal, V. (2019, October). Is one time-series in the business tendency survey able to predict another one? Granger causality between time series. In 22nd International Scientific Conference on Applications of Mathematics and Statistics in Economics (AMSE 2019). Atlantis Press.

41. Plub-in, N., & Songsiri, J. (2019, October). Estimation of Granger Causality of StateSpace Models using a Clustering with Gaussian Mixture Model. In 2019 IEEE International Conference on Systems, Man and Cybernetics (SMC) (pp. 3853-3858). IEEE.

42. Rashid, S., & Minot, N. (2010). Are staple food markets in Africa efficient? Spatial price analyses and beyond (No. 1093-2016-87814).

43. Rizwanullah, M., Liang, L., Yu, X., Zhou, J., Nasrullah, M., & Ali, M. U. (2020). Exploring the Cointegration Relation among Top Eight Asian Stock Markets. Open Journal of Business and Management, 8(03), 1076.

44. Roman, M. (2020). Spatial Integration of the Milk Market in Poland. Sustainability, 12(4), 1471.

45. Sa'diyah, A. A., Nendissa, D. R., & Sinaga, A. M. IMPACTS OF RISING STRATEGIC FOOD PRICES ON POVERTY IN INDONESIA. In Proceeding of the International Conference on Food and Agriculture (Vol. 2, No. 1). ISBN: 978-602-14917-9-9

46. Salazar, C., Ayalew, H., & Fisker, P. (2019). Weather shocks and spatial market efficiency: evidence from Mozambique. The Journal of Development Studies, 55(9), 1967-1982.

47. Sexton, R. J., Kling, C. L., and Carman, H. F. 1991. Market Integration, Efficiency of Arbitrage, and Imperfect Competition: Methodology and Application to US Celery. American Journal of Agricultural Economics, 73(3), 568-580.

48. Suryana C, Asriani PS, Badrudin R. 2014. Perilaku harga dan integrasi pasar horizontal beras di provinsi Bengkulu. AGRISEP. 14(2): 131-146.

49. Svanidze, M., & Götz, L. (2019). Spatial market efficiency of grain markets in Russia: Implications of high trade costs for export potential. Global Food Security, 21, 60-68.

50. Ters, K., & Urban, J. (2020). Estimating unknown arbitrage costs: Evidence from a 3-regime threshold vector error correction model. Journal of Financial Markets, 47, 100503.

51. Traub, L. N; Myers R. J., & Meyer, F. H. (2014). Measuring Integration and efficiensy in Maize Grain Market: The Case of South Africa and Mozambique. Paper Presemtased at African Association of Agriculture Economics (AAAE) and Agricultural Economics of South Africa (AEASA), Conference.

52. Tursoy, T. (2019). The interaction between stock prices and interest rates in Turkey: empirical evidence from ARDL bounds test cointegration. Financial Innovation, 5(1), 7.

53. Varela, G., Aldaz-Carroll, E., & Iacovone, L. (2012). Determinants of market integration and price transmission in Indonesia. The World Bank.

54. Varela, G., Aldaz-Carroll, E., & Iacovone, L. (2012). Determinants of market integration and price transmission in Indonesia. The World Bank.

55. Widarjono, 2018. Ekonometrika: Pengantar dan Aplikasinya Disertai Panduan Eviews UPP STIM YKPN, Yogyakarta.

56. Zhang, Z., Hu, W., Tian, T., & Zhu, J. (2020). Dynamic Window-level Granger Causality of Multi-channel Time Series. arXiv preprint arXiv:2006.07788.