Sustainability of Sustainable Palm Oil: a Market Integration Analysis

Crude Palm Oil (CPO) is the biggest consumed vegetable oil in the world. The increase in CPO production raises concern on the environmental impact even outside the producing countries. As a response to this matter, the EU has made a requirement to only import certified CPO (CSPO). India and China, the two biggest importers in the world, are less restrictive to the environmental issues, and their demands are more influenced by CPO price levels. These countries are the main export markets for Indonesia and Malaysia, the two biggest CPO exporters in the world. This research using monthly price data from the Netherlands, Germany, Italy, EU28, India, China, Indonesia and Malaysia. Market integrations are tested with Cointegration Test, Vector Error Correction Model and Seemingly Unrelated Regression. The results show that these markets are integrated, but European countries are unlikely to lead the price movement. Therefore, the concern on sustainable certification from the European countries still slowly spreads to other main importers, resulting in low absorption of CSPO.


INTRODUCTION
CPO is one of the important commodities in the international trade. However, its enormous growth raises fears about the impact on environment quality and sustainability in the long term. CPO is required by a large number of countries, as it becomes raw materials for various products with competitive prices. Addressing this issue, RSPO offers sustainable management through a number of principles and criteria so that these two interests can be balanced. Producers who have IXO¿OOHG WKHVH SULQFLSOHV DQG FULWHULD ZLOO REWDLQ 5632 FHUWL¿FDWLRQ 6LQFH LWV LQWURGXFWLRQ LQ SDOP oil mills and 64 growers producing 13.18 tons CSPO KDYH EHHQ FHUWL¿HG $OWKRXJK QRW ERXQG PDQ\ 5632 consumer members, especially those from European countries have committed to only import CSPO. RSPO FHUWL¿FDWLRQ EHFRPHV D GHPDQG VKRFN DQG LQÀXHQFHV volume imports in European market.
7KHUHIRUH DOWKRXJK 5632 FHUWL¿FDWLRQ QHHGV H[WUD costs and work, companies with European countries as their main export destination are more likely to obtain WKH FHUWL¿FDWHV WKDQ WKRVH ZLWK RWKHU H [SRUW PDUNHW destinations (Anderson SW HW DO 1999). European countries demand mostly in palm stearin form which is widely used in food products with RSPO label on their packages. However, palm stearin represents only 20% of the palm oil world import volume (SPOTT, 2016).
A number of big palm oil companies from Indonesia DQG 0DOD\VLD KDYH REWDLQHG 5632 FHUWL¿FDWLRQ D lot of Small companies and Smallholders have not EHHQ FHUWL¿HG 7KLV FRXOG QRW EH LJQRUHG EHFDXVH Smallholders' total palm oil areas are more than 40% of the total palm oil areas (Palm Oil Research, 2015). )RU PRVW 6PDOOKROGHUV WKH FHUWL¿FDWLRQ FRVWV DUH KLJK and some of the principles and criteria are still quite complicated (Stanton and Burkink, 2008;Parrish HW DO 2005;Hoebink HW DO 2014;Nelson HW DO 2010). However, the compensation is still unclear and there are no CSPO premium price references. Currently, CSPO price is determined through deals, and from 2011 to 2014 the premium price declined, remaining approximately US $ 1,97 per metric ton and there is a difference between the prices of CSPO and conventional-CPO . This can be a disincentive to produce CSPO, especially for main CPO producers from Indonesia and Malaysia.
European countries are not the main export market destinations for main CPO producers and exporters. In 2015, these countries' exports and production were accounted for 86% and 91% in the world respectively. Indonesia exports 48,43% of its total production to India, while Malaysia exports 16,41% to China. India is the biggest palm oil importer and consumer in the world with 20,5% and 15% share, respectively, and China is the third importer and fourth consumer with 12,5% and 9,7% share, respectively (USDA, 2015). In India and China, price is still a more important factor than environmental issues. Their consumers are not willing to pay more for cooking oil that use CSPO as its raw material. Physically CSPO and CPO are not different and produce the same quality of end product. Therefore, India and China consumers consider CSPO and CPO as homogeneous product. Higher CSPO price leads to a low demand the India present market. The government, likewise, has little incentive to absorb surplus costs. Currently, India and China tend to enforce less restrictive import requirements. When Indonesia and Malaysia decrease their export tax on processed &32 5H¿QHG %OHDFKHG DQG 'HRGRUL]HG 5%' ROHLQ ,QGLD VLJQL¿FDQWO\ LQFUHDVHG WKHLU LPSRUW +XFDO Arora HW DO 2014;Ningsih, 2016). Indonesia is also recorded as the second largest palm oil consumer, with 13% share of the total world consumption (USDA, 2015). In 2015 Indonesia has also issued a policy to increase the percentage of biofuel palm oil content in Indonesia, from 15% to 20%, which is equivalent to an increase in palm oil consumption by 3-4 million tons/ year. However, Indonesian consumers are also unlikely to pay a higher price for CSPO. Without support from India, China and Indonesia, sustainable palm oil will slowly progress. Since 2009 CSPO uptake has only reached 48,64% of its total production . However, if European markets are integrated with Indian, Chinese or Indonesian markets, demand shocks in one market will be transmitted to other markets.
This study was conducted to analyze whether the sustainable issue in European markets will be spread to India and China markets, by analyzing price movements in these markets. The hypothesis is palm oil prices in India, China, Indonesia and Malaysia are not integrated with those in European countries, thus price movement in European markets are not transmitted in Indian, Chinese, Indonesian and Malaysian markets. With such a condition the sustainability of sustainable palm oil (CSPO) in the long run can be threaten. Section 2 provides a brief review of previews studies, while Section 3 describes the methods. Section 4 analyzes the estimation results, and Section 5 concludes and presents a number of suggestions and policy implications.
0DUNHW LQWHJUDWLRQ LV GH¿QHG DV WKH UHODWLRQVKLS EHWZHHQ markets, which is indicated by the price transmission from a market to another, either horizontally or vertically, and their price share stable long run price equilibrium. The attainment of such equilibrium FDQ EH UHDFKHG LI HLWKHU WUDGH ÀRZV RFFXU EHWZHHQ spatially separated of a homogeneous commodity RU LI LQIRUPDWLRQ ÀRZV :LWKRXW WUDQVPLVVLRQ RI information, differentiated commodity might be treated as homogeneous. For example, without labels, organic commodities could be treated as conventional ones but not vice versa, thus such a substitutability is considered asymmetric (Wurriehausen, 2012). A similar condition occurs in the palm oil market, in which CSPO and CPO are not physically different. CSPO can be grouped into 3, namely segregated, mass balance and book and claim types. Among them, segregated palm oil is the only group that fully separates CSPO and conventional CPO. Therefore, mass balance and book and claim are often sold as conventional CPO.
In integrated markets each market employs information from the other when forcing its own price expectations, either in a uni directional or bi directional causality. A number of studies have been conducted in agricultural commodities in the domestic market (Wani HW DO 2015;Carew HW DO 2012;Hossain, and Verbeke, 2010;Zhou, ZY HW DO 2000;Edet HW DO 2014), international market (Rosa HW DO 2014;Jones and Kwiecinsk, 2010) or connection of both markets (Fossati HW DO 2007). In previous studies, the transmission is determined by their co-movement through the Cointegration Test. The short run price movement can drift apart or adjusts and converge towards an equilibrium, in which speed of adjustment is measured through the Error Correction Model (Rapsomanikis HW DO 2006). The direction of causality is measured by Granger Causality and used to determine the leader and follower role of each market (Arshad and Hameed, 2014). However, Granger Causality test is designed to handle pairs of variables and may be misleading when true relationship involves more variables, hence similar Vector Autoregressive (VAR). Model with full information method such as Seemingly Unrelated Regression (SUR) can be used to overcome the problem (Roy HW DO 1994).
7KH LQWHJUDWLRQ GHJUHH FRXOG EH LQÀXHQFHG E\ VWRFN availability, contracts and government interventions. Stocks and contracts are often used to stabilize the agricultural product prices in future palm oil markets, such as in Malaysia and Rotterdam (Rahman HW DO 2012). Rotterdam, which is located in the Netherlands, is an important port for importing palm oil to European countries. Biofuel plants of large companies such as 1HVWOH DQG $EHQJRD DV ZHOO DV YHJHWDEOH RLO UH¿QHULHV IURP &DUJLOO 5H¿QHG 2LOV (XURSH /RGJHUV &URNODDQ IOI Edible Oils and Wilmar are located in the port of Rotterdam (Port of Rotterdam, 2015). European countries cannot produce CPO, but have other edible RLOV VXFK DV VR\EHDQ UDSHVHHG DQG VXQÀRZHU DV D substitution. However, CPO is still the main imported edible oil to European countries. In 2009-2010, palm oil was recorded to be 60% of the total imported edible oil to European countries and 36% of the palm oil was imported through the Rotterdam port (Ridder HW DO 2014;Gerasimchuck and Koh, 2013). Bursa Malaysia Derivative Exchange uses two main ports in Malaysia, namely the port of Klang and Pasir Gudang. Similarly, D QXPEHU RI PDMRU FRPSDQLHV VXFK DV &DUJLOO UH¿QHU\ Sime Darby, Felda IFFCO International Mewah Group, and Shell are located at these ports.
The government intervenes the domestic market price through the determination of the output selling price or of input subsidy (Zhou HW DO 2000), and the international markets conduct this through tariff and non-tariff barriers (Hossain and Verbeke, 2010 Previous studies show that Less Developed Countries appear to be the most exposed to non-tariff barriers (Kooster, 2000).

METHODS
This study uses monthly CPO price data from 2011 to 2014, which covering 5 importing countries, namely China, India, Germany, The Netherlands, Italy, a group of importing countries EU28, and 2 main exporting countries, Indonesia and Malaysia. There is a lack of complete monthly China price data, therefore this study uses China Hongkong SAR as a proxy. Hongkong price appears to be higher than China, but the movement is VLPLODU 7KLV MXVWL¿HV WKH XVDJH RI +RQJNRQJ SULFH DV the proxy because the focus of this study is to analyze the price co-movement.
The CPO price data of the European countries used in the study were calculated from the data import value DQG YROXPH RI LPSRUWV &HUWL¿HG DQG QRQ FHUWL¿HG SDOP oil price data are not separated in European countries. However, Germany has committed to use only CSPO from 2014, while the Netherland from 2015 (RSPO, &HUWL¿HG &32 SULFHV KDYH D JUHDWHU UDQJH WKDQ WKRVH QRW FHUWL¿HG EHFDXVH WKH\ LQFOXGH WKUHH W\SHV RI FHUWL¿FDWLRQ QDPHO\ VHJUHJDWLRQ PDVV EDODQFHG ERRN and claim. Not all traded physically. Partly through virtual transactions Green Palm, with premium pricing through direct negotiations between buyers and sellers. Therefore, CSPO price could be listed only a conventional CPO price (Ningsih, 2015). In 2009-2014, an average 14,5% of total CPO production was recorded as CSPO physical uptakes, while 30% uptakes were recorded through the Green Palm program (RSPO, 2014 (Dicky and Fuller, 1970) for all series by using the following equation.
Yt is the price series from each country in question, ! LV WKH WHVW FRHI¿FLHQW DQG P LV WKH ODJ OHQJWK FKRVHQ from ADF test. The null hypothesis is Yt has a unit root or non-stationer. The aim is to avoid spurious regression, such as series with the same trend that can KDYH D VLJQL¿FDQW FRHI¿FLHQW DOWKRXJK QRW UHODWHG WR one another. The test starts by using price series in level data. If the test cannot reject the null hypothesis, the data is considered with null order I(0). Otherwise the test continues with the different data until they reject rejects the null hypothesis to determine the price series order. Price series with different orders show that they are not related one to another, which means there is no market integration between the countries in question. If the series have the same order, then they are further tested with Johansen Cointegration Test (1995) (Eagle and Granger, 1987). 7KH VLJQL¿FDQFH RI WKH VKRUW UXQ G\QDPLF VSHHG RI adjustment and convergence of the long run equilibrium are estimated with the VECM that can be written as:  Figure 2 shows a declining price movement in importer and exporter countries in 2011-2014 monthly CPO prices. In general CPO prices in exporting countries are lower than those in the importing countries. Between the exporting countries, the price in Malaysia is higher than that Indonesia. CPO price movements in both exporter countries are similar, with the lowest price around September and highest price in the months of February-March. In addition among importing countries, India appears to be the leader. An increase (decrease) in India price is followed by an increase (decrease) in other importing countries in the following months.

Cointegration Test and Vector Error Correction Model
The unit root test results on the price series show that the ZKROH VHULHV LV VWDWLRQDU\ LQ WKH ¿UVW GLIIHUHQFH 6LQFH all the series have the same order I(1), then the analysis can proceed to the cointegration test (Table 1). Trace and Maximum Eigenvalue in the cointegration test results for the importing country group and for the mixed of importing and exporting country group contain cointegration equations (Table 2 and Table 3). The addition of exporting country in each sub group increases the number of cointegrating equations, indicating market integration between Indonesia or Malaysia and other importing countries. Therefore, Indonesia and Malaysia are also included in both the VECM and SUR. The number of cointegrating equations shows that in comparison with China, CPO price in India is likely more integrated with those of other countries. Table 2 and Table 3 also show that none of Group I and Group III that separate the European countries have full integrations. For example, Group I in Table 2 only has 2 cointegrating equations, while the maximum integration is 3. However, this test is unable to indicate the integrated or unintegrated countries in a group. Unless the group only contain 1 dependent and independent series such as that in Group II and Group IV in Table 2 or full integration in the same groups in Table 3, which shows both India and China price are cointegrated with EU 28. Because of the cointegration, the estimation could be continued to VECM to explore the indicated integrations.

Seemingly Unrelated Regression
Similar with the cointegration test and VECM, SUR also consists of 2 scenarios that separate the 3 European countries (scenario 1) and join them in EU28 (scenario 2).   India itself is a producer of some vegetable oils and even started to develop oil palm plantations. Palm oil appears to be the main oil with 77% share of the total vegetable oil consumption in India. Because of the competitive price, ninety percent of the edible oil in India is sold in loss rather than branded form which means no premium is available to be paid as the "branded sustainable palm oil". Companies in India KDYH VWDUWHG WR REWDLQ 5632 FHUWL¿FDWHV LQ EXW until 2014 their CSPO consumption was still very low (WWF, 2013;Arora HW DO 2014). Similarly, Indonesia also uses most of the CPO for edible oil and most of WKH HQG FRQVXPHUV ZRXOG XVH WKH FHUWL¿HG SURGXFW only if the prices are the same or lower than those of WKHLU FXUUHQW XQFHUWL¿HG SURGXFW 'DHPHWHU This partly explains the slow improvement in RSPO FHUWL¿HG &32 &632 DEVRUSWLRQ DQG WKH GHFUHDVH in CSPO premium price. In fact, sustainable palm oil products are more demanded by European countries, but not transmitted to other main palm oil importers such as India and China. Without the participation of the two countries, the sustainability of sustainable palm oil is questionable.

Recommendations
Since European countries are not the main export destinations of Indonesia, it should proportionally respond to their demands, including the RSPO FHUWL¿FDWLRQ 7KH UHDVRQ LV LW UHTXLUHV D FRQVLGHUDEOH amount of money with a large amount of effort, but the uptake and premium price are still low. Indonesia VKRXOG LPSURYH HI¿FLHQFLHV DQG SURYLGH FRPSHWLWLYH selling prices to meet demands from India and China.