Rethinking the Asian Marker Debate

MASTER’S THESIS

Fabian Weber International Energy (M.A.)

Academic Year 2014/15 Supervisor: Giacomo Luciani, Scientific Advisor

The copyright of this Master's thesis remains the property of its author. No part of the content may be reproduced, published, distributed, copied or stored for public or private use without written permission of the author. All authorisation requests sh ould be sent to [email protected] Rethinking the Asian Marker Debate

Acknowledgements First and foremost, I would like to sincerely thank Giacomo Luciani for his continuous guid- ance, persistence and trust throughout the planning and writing of this paper. My deep grati- tude furthermore goes to Bassam Fattouh for enabling me to conduct research under his men- torship at the Oxford Institute for Energy Studies (OIES). Both his expertise and experience were of inestimable value and greatly influenced the work at hand. I would also like to thank the OIES at large for the uncomplicated administration of my stay. This paper has also ma- jorly benefited of the discussions with several experts in the field, whom I would like to thank for their willingness to share their views and expertise. Although these engaging discussions have deeply shaped this paper, any remaining shortcomings and errors are my own. Last but not least, I would like to thank Dominika for her continuous and absolute support throughout the entire process. It is in her arms that my mind finds the peace and strength for all challenges ahead.

-2- Rethinking the Asian Marker Debate

Abstract

The purpose of this paper is to firstly analytically structure the fundaments of crude price benchmarks before secondly addressing the issue of the Asian marker. In doing so, a particular focus shall be placed on the question whether the reference crude for the Asian basin should also originate from this market, given the eastward shift characterising international crude oil markets today and even more so in the future. Firstly, crude price benchmarks and their two main functions in the current crude oil pricing system are introduced: price discovery and price risk management. Thereafter, recent structural shifts in international oil markets in terms of both market fundamentals and regulatory environment are presented. Subse- quently, the Asian marker debate and the most common candidates are established, before critically evaluating the need for an Asian marker located in Asia. Last but not least, the DME Oman/Dubai complex is proposed as an innovative example of an Asian marker located outside of Asia, consequently rebuilding the current Middle Eastern benchmark system and strengthening its position in times of eastward shifting balance of power in international oil markets.

-3- Rethinking the Asian Marker Debate

List of Contents 1 Introduction ...... 6 2 Formula Pricing and Benchmarks...... 6 2.1 The Rationales for Benchmarks ...... 7 2.1.1 Price Discovery...... 7 2.1.2 Price Risk Management ...... 10 2.2 Two-Dimensionality of Benchmarks ...... 12 3 Oil Markets in Transition ...... 13 3.1 Market Fundamentals...... 13 3.2 Market Regulation...... 16 4 The Asian Marker Debate ...... 18 4.1 Candidates Located Outside of Asia ...... 19 4.1.1 The Platts Dubai/Oman Assessment...... 19 4.1.2 The DME Oman Crude Oil Futures Contract...... 23 4.1.3 The Platts Dated Brent Assessment ...... 26 4.2 Candidates Located in Asia...... 28 4.2.1 The East Siberia-Pacific Ocean Blend...... 28 4.2.2 The Shanghai Crude Oil Futures Contract...... 33 4.3 Evaluating the Arguments for a Marker Located in Asia ...... 35 5 An Alternative: Rebuilding the Middle Eastern Marker ...... 38 5.1 The Linkage...... 38 5.2 Fixing Dubai...... 40 5.3 Fixing DME Oman...... 41 5.4 Dealing with Shifting Market Trends ...... 42 6 Outlook and Conclusions...... 42 References...... 44

List of Figures Figure 1: East Siberia-Pacific Ocean Pipeline (Phase I/II) ...... 30

-4- Rethinking the Asian Marker Debate

List of Abbreviations API...... American Petroleum Institute APPI...... Asia Petroleum Price Index ASCI ...... Argus Sour Crude Index b/d ...... Barrels per day BALMO ...... Balance of the Month BFOE ...... Brent, Forties, Oseberg and Ekofisk BWAVE...... Brent Weighted Average CEO...... Chief Executive Officer CNPC ...... China National Petroleum Corporation DME...... Dubai Mercantile Exchange EFP...... Exchange of Futures for Physicals EFS...... Exchange of Futures for Swaps EIA...... U.S. Energy Information Administration ESMA ...... European Securities and Markets Authority ESPO...... East Siberia-Pacific Ocean FEPCO ...... Far East Petrochemical Complex FOB...... Free-on-board GMT...... Greenwich Mean Time GPW...... Gross Products Worth ICE ...... Intercontinental Exchange IEA...... International Energy Agency INE...... Shanghai International Energy Exchange IOSCO...... International Organization of Securities Commission IPE...... International Petroleum Exchange IPRO ...... Independent Price Reporting Organisations km ...... Kilometres Libor...... London Interbank Offered Rate Mogas...... Motor Gasoline NYMEX...... New York Mercantile Exchange OPEC ...... Organization of the Petroleum Exporting Countries OSP ...... Official Selling Price OTC...... Over-the-counter PDO...... Petroleum Development of Oman PEMEX...... Petróleos Mexicanos PRA...... Price Reporting Agency SHFE...... Shanghai Futures Exchange WTI...... West Texas Intermediate

-5- Rethinking the Asian Marker Debate

1 Introduction This paper shall examine the influence of recent changes in the dynamics of global oil markets on the international pricing system for crude oil. As the Asian market is about to become “[…] the new center of gravity for the oil markets”1 the debate concerning a crude price benchmark for the Asian basin has gained further momentum. 2 A new dimension to this debate has been added as Asian candidates have entered the competition for the Asian benchmark status and their necessity shall be examined in detail. This debate inevitably touches on fundamental questions about crude price benchmarks, such as, for example: Why do we benchmark and which services should a functional benchmark be able to deliver? Which conditions have to be fulfilled for successful benchmarks to evolve and where lies the added value of new benchmarks for Asia? These questions have already been posed by Fattouh but still remain startlingly under- researched in the current academic literature. 3 Although international crude oil markets – and in particular crude pricing systems – have been investigated thoroughly, currently two main topics stand out. Firstly, the historical development of the various crude oil pricing systems up to today’s market-related system. 4 Secondly, the issue of an increasing financialization of oil markets, speculation and concomitantly price volatility. 5 Hence, there is lack of a structural framework for the current debate on the Asian marker, as both pricing systems and price levels have been analysed but conceptual work on the underlying price benchmarks is still somewhat limited. Accordingly this paper has two purposes. Firstly, the fundaments of crude price benchmarks shall be analytically structured before secondly, the issue of the Asian marker shall be addressed. In doing so, a particular focus shall be placed on the question whether the reference crude for the Asian basin should also originate from this market, given the eastward shift characterising international crude oil markets today and even more so in the future. The paper is divided into six sections.Section 2 introduces crude price benchmarks and their two main functions in the current crude oil pricing system: price discovery and price risk management. Section 3 presents recent structural shifts in international oil markets, taking into account both changes in market fundamentals and the regulatory environment. Section 4subsequently establishes the Asian marker debate and the most common candidates, before critically evaluating the need for an Asian marker located in Asia. Last but not least, Section 5 proposes an innovative example of an Asian marker located outside of Asia, rebuilding the current Middle Eastern benchmark system and strengthening its position in times ofeastward shifting balance of power in international oil markets.

2 Formula Pricing and Benchmarks International crude oil trade is still mainly based on the so-called market-related pricing system, which was first used by the Mexican national oil company Petróleos Mexicanos (PE-

1Imsirovic, ‘Asian Oil Markets in Transition’. 2Fattouh, ‘Anatomy’, 78. 3Ibid. 4Fattouh, ‘Anatomy’; Mabro, ‘International Oil Price Regime’; Fattouh, ‘Origins and Evolution’; Fattouh and Mahadeva, ‘OPEC’; Mabro, ‘On Oil Price Concepts’. 5Fattouh, Kilian, and Mahadeva, ‘Role of Speculation’; Fattouh and Mahadeva, ‘Assessing Financialization Hypothesis’; Fattouh and Scaramozzino, ‘Uncertainty, Expectations, and Fundamentals’.

-6- Rethinking the Asian Marker Debate

MEX) in 1986, as an alternative to netback pricing, and established itself as the main pricing technique by the late 1980’s. This involves a method referred to as formula pricing, where volumes of oil in longterm contracts are linked to a market, i.e. spot, price. Because of the significant differences in quality among crudes 6the price for a given crude is set in relation to a reference or marker crude 7, often referred to as the benchmark crude. 8However, quality differences cannot be the sole reason for this relative pricing method as they stay fairly stable over time and thus cannot explain the flexible pricing function with adjusting differentials. It is the changes in Gross Products Worth (GPW) which different crudes yield after refining, depending on the relative margins of petroleum products, that make adjusting price differentials necessary as movements in GPW drive the demand and supply for different crude types. 9The method of pricing crudes as premiums/discounts to a chosen set of benchmarks is the reason for the naming formula pricing :“Specifically, for crude oil of variety X, the formula pricing can be written as P X=P R±D where P X is the price of crude X, P R is the bench- 10 mark crude price; and D is the value of the price differential.” The price P Xis commonly referred to as an ‘Official Formula Price’ and should be distinguished from an ‘Official Sell- ing Price’ (OSP), which price level is entirely set by the government.The differential can either be set by the respective oil-producing country or assessed by oil price reporting agencies (PRAs), which will be introduced in the subsequent section. If set by an oil-producing country, the differential is usually specified in the month before the loading month by each oil- producing country and is adjusted monthly or quarterly. The differential for May, for example, is announced in April and based on available data on GPW from March. 11 This gives this type of pricing the name ‘forward pricing system’ as opposed to ‘retroactive pricing’, where pricing takes place after the cargoes have been loaded. Thus, the setting of price differentials involves significant time lags so that differentials often do not reflect market conditions at time of loading and even less so at time of arrival. In longterm contracts these time lags can be dealt with by adjusting the differential downward for subsequent deliveries in order to compensate the buyer for previous mispricing. 12 2.1 The Rationales for Benchmarks 2.1.1 Price Discovery The respective benchmark crudes used in formula pricing fulfil two main objectives: price discovery and price risk management. Firstly, benchmark crudes deliver a price discovery service which is relied on by oil companies and traders to price oil exports, by financial markets for the settlement of derivative instruments (e.g. futures, option or swap contracts) and by governments for taxation. 13 Therefore, price signals need to be immediate and correct, whereas specific market participants can have particular requirements for the price signal

6The most commonly used metric is two dimensional and comprises the sulphur content (sweet/sour) as well as the American Petroleum Institute (API) gravity (light/heavy). 7 Or a set of marker crudes, e.g. eastbound exports from most Middle Eastern producers are priced against an average of Dubai and Omani crude. 8 Following the general terminology the terms reference/marker crude, benchmark (crude) or simply ‘marker’ shall be used synonymously throughout this paper. 9Fattouh, ‘Anatomy’, 20 ff. 10 Ibid., 21. 11 Note, however, that the setting of the differential is also influenced by the pricing of close competitors. It is thus also a means of competition between oil exporters (Fattouh, Anatomy , 22). 12 Ibid., 21–22; 68. 13 Ibid., 24.

-7- Rethinking the Asian Marker Debate

(e.g. fairly stable marker prices for the predictability of government tax revenues). One important characteristic of these reference prices is that they – in theory at least – stem from a spot market, which in turn should result in price allocation at the margin of the corresponding physical oil market. In reality, however, there are well-known and widely discussed shortcomings to this method of price discovery through a set of physically limited reference crudes. Their declining production may lead to concerns regarding adequate price discoveryand risk of price manipulation in increasingly illiquid markets with low quantities of irregularly concluded deals. This gave rise to so-called price reporting agencies (e.g. Platts or Argus), assessing the reference prices according to specific methodologies 14 – e.g. using information on bids and offers –in the absence of sufficient amounts of actual transactions.A prominent example of such an assessment procedure – that will frequently be referred to throughout this paper – is the so-called ‘Platts window’. This is an electronic platform set up by Platts on which traders can buy and sell physical oil. This allows Platts to obtain information on the players, volume and price involved in a specific transaction as well as on bids and offers. This is important as oil markets are opaque, meaning that physical transactions do not have to be reported, thus hampering price discovery without institutionalized procedures such as the Platts window. 15 In some cases, as e.g. for Dated Brent, the status of a marker crude was maintained by widening the set of grades assessed under the label Brent, thus (at least tempo- rarily) mitigating the problem of insufficient physical liquidity. 16 The physical production of other reference crudes, however, declined to the point that they were abandoned as markers (as in the case of Alaska North Slope for North America or Minas and Tapis for the Asian market). 17 To some degree there has also been a shift from using physical benchmarks to financial benchmarks, which evolved around these physical markers, for pricing crude exports in order to deal with these issues of physical illiquidity.A prominent example for such a financial benchmark is the Brent Weighted Average (BWAVE)18 , an index derived from Brent futures prices, which has been adopted by Saudi Arabia, Kuwait and Iran for their crude exports to Europe in 2000/2001. 19 Such financial benchmarks have the advantage of a significantly higher volume of transaction on financial markets, thus decreasing the risk of price manipulation. Additionally, the financial layers around physical benchmarks play a crucial role in the price discovery process. The prices generated on the financial markets are influenced by other market fundamentals than the physical prices (e.g. regarding the timing or location) but are nonetheless essential for generating benchmark prices. It has even been argued that “[…] without these financial layers it would not be possible to ‘discover’ or ‘iden- tify’ oil prices in the current oil pricing system.” 20 Some of these price generation processes on the financial markets and their linkages to the physical markets will be discussed later in this paper, for instance, using the example of the Dubai marker and the Brent complex. Addi- tionally, the prices posted by exchanges are those of actual deals and not assessed by price

14 Hence, “[T]he prices generated are assessed prices and not market prices.” (Fattouh, ‘Origins and Evolution’, 65). 15 Fattouh, ‘Dubai Benchmark and Its Role’, 3; Fattouh, ‘Anatomy’, 26 ff. 16 However, this gradual widening from Brent to BFOE (Brent, Forties, Oseberg and Ekofisk) has itself led to additional problems. For a detailed description see Fattouh, ‘Anatomy’, 39. 17 Mabro, ‘International Oil Price Regime’, 9 ff. 18 “Specifically, the BWAVE isthe weighted average of all futures price quotations that arise for a given contract of the futures exchangeduring a trading day, with the weights being the shares of the relevant volume of transactions on that day.” (Fattouh, Anatomy , 25). 19 Ibid., 25. 20 Ibid., 78.

-8- Rethinking the Asian Marker Debate reporting agencies’ methodologies that inherently remain somewhat subjective. Last but not least, market information provided on both prices and market liquidity (e.g. on daily transaction volumes or open interest) is transparent and real-time information. 21 All in all we can conclude, thatthe benchmark system is not stable over time, although its fundamental structure has not been significantly modified since its introduction in the second half of the 1980’s. In regard to the price discovery function two categories of benchmarks can be distinguished. Firstly, regional benchmarks which reflect the supply and demand conditions of a specific market. Theoretically a regional marker shall reflect the supply and demand conditions of the market the crude, which it is pricing, is exported to. An example for this are the two crudes Minas and Tapis which traditionally where used as regional markers for Asia (mostly by Asian producers). 22 By linking the price of the exported crude to the export market it shall be ensured that the export price varies depending on the local supply and demand conditions of this particular region. Secondly, so-called international benchmarks exist, such as Dated Brentand formerlyWest Texas Intermediate (WTI). These reference crudes not only reflect one particular regional market, but additionally mirror the global supply and demand characteristics of their particular crude quality. This allows these international markers to be used not only to price crude flowsinto their region, as it is the case for a regional marker, but enables producers to price their crude exports to various markets based on these benchmarks. In this sense, international markers can be thought of as a subset of regional benchmarks, i.e. all global benchmarks are also regional benchmarks but not all regional benchmarks are also global benchmarks. This global status is attained due to their embeddedness in the global trade of crude, as exemplified by Dated Brent which functions as a global marker for sweet crude. Dated Brent can freely flow in and out the North Sea, allowing it to play a central role for global arbitrage of sweet crude. The price of Dated Brent is both influenced by the market fundamentals in the Atlantic basin as well as in the Asian basin: If Dated Brent prices are low relative to WTI, trans-Atlantic arbitrage is possible allowing Dated Brent to flow to North America.This leads to higher prices for Dated Brent, as demand is increased, until the trans- Atlantic window closes. If prices of Dated Brent are high compared to Dubai/Oman (the widely used benchmark for Middle Eastern exports to Asia), crude oil from West Africa flows into the North Sea basin. This results in a reduction of Dated Brent prices due to in- coming supply in the North Sea. Moreover, as the Brent-Dubai differential narrows it becomes economic to export European or African crude to Asia, until increasing prices for Dated Brent close the arbitrage opportunity. 23 Because of this reflection of global sweet crude fundamentals Dated Brent is used not only to price oil in the North Sea, but also in “[…] West Africa, the Mediterranean, South and Latin America, Canada and North America, Cen- tral Asia and in Russia. More than 60% of the world’s internationally traded crude oil is priced against Dated Brent.”24 This widespread usage of Dated Brent to price crude exports to various regions is the reason why we refer to Dated Brent as an ‘international benchmark’.

21 Mabro, ‘International Oil Price Regime’, 9 ff.; Fattouh, ‘Anatomy’, 25 ff.; 78. 22 They were then replaced by Dated Brent. This replacement of regional markers by more and more consoli- dated global marker(s) is a general phenomenon which will be touched on in section 2.1.2. 23 Platts, ‘Dated Brent’, 2; Fattouh, ‘Origins and Evolution’, 62. 24 Platts, ‘Dated Brent’, 2.

-9- Rethinking the Asian Marker Debate

2.1.2 Price Risk Management It is important that the benchmark figure is perceived as ‘correct’, i.e. as a meaningful repre- sentation of the underlying market fundamentals, by the market participants. This concrete figure then also – and arguably primarily – has its importance for the management of price risk. In general, physical delivery of crude oil can be contractually organized in two ways: either through the spot market or term contracts, whereas most transactions are managed through the latter. In 2010 Platts estimated the share of crude oil and oil products to be sold under term contracts to be 90-95%. 25 Although this includes petroleum products, it gives us an idea of the prevalent distribution. However, it is important to note that even spot transactions are not immediately delivered but have a certain element of forwardness – due to the logistics of oil transporting – which can be up to 45-60 days. This begs the question of when spot pricing takes place, as price levels can obviously vary significantly over the time period of two months. Combined with the potentially large quantities physically delivered and the resulting monetary value, substantial price risk has to be managed by both the buyer and seller. Usu- ally the spot price is determined at the time of loading, thus reducing the period vulnerable to price movements compared to the sometimes chosen pricing at the time of agreement (trans- forming the transaction more to a forward contract). 26 In this context formula pricing, which may be applied to spot, forward or longterm contracts, is crucial. It allows the buyer and seller to share the price risk, as the price of the shipment can be linked to the market conditions prevailing at the time of delivery. Crude exports from Saudi Arabia to the USA, for example, are priced against the trade month (20 day minimum) average of the Argus Sour Crude Index (ASCI) for the trade month of the time of delivery. By using an average the price volatility (and hence risk) is reduced in comparison to using the benchmark price of one particular day. In addition, pricing the crude at delivery technically moves the point of sale closer to the destination rather than the origin. In terms of price risk, this reduces the risk borne by the buyer and as a consequence increases the amount of risk the seller has to cover (in comparison to pricing at time of loading, where the buyer bears all the price risk). For- mula pricing enables this transfer of risk as it (theoretically at least) allows access to a spot price at time of delivery. This is attempted by means of using the price signal of the benchmark crude. This theoretical ideal of pricing at the margin is approximated at best, because of the above-discussed element of forwardness inherent to the so-called spot market for crude as well as the fact that benchmark prices are assessed prices as discussed previously. 27 This does, however, not change the fact that the price risk is shared between seller and buyer by shifting the moment of pricing from time of loading to time of delivery. This shift, in turn, is made possible by drawing on a marker price, with the underlying intention to reflect a price at the margin at time of delivery. 28 Hence, using a system of formula pricing based on a reference price allows two dimensions of risk management: risk sharing (as has just been explained) and risk management using financial instruments (i.e. hedging) as will be illustrated in the following.

25 Platts, ‘Structure of Global Oil Markets’, 2. 26 Fattouh, ‘Anatomy’, 20–21. 27 For a detailed example using the Dated Brent market see Fattouh, ‘Origins and Evolution’, 66–67. 28 Fattouh, ‘Anatomy’, 20 ff.

-10- Rethinking the Asian Marker Debate

From a perspective of price risk management this means that buyers/sellers exposed to price movements of the imported/exported crude between loading and delivery should focus on the benchmark price component of formula pricing. The price risk is commonly hedged by un- dertaking a purely financial transaction equivalent to the physical deal committed to. Usually this is done so with the help of futures contracts or swaps, less so through options contracts. Let us assume a producer wants to sell a specific quantity of his crude X, which is priced off 29 of the reference crude R. At the time of loading (t0) the price of X is P X0. However, the producer expects the price to fall to P X1 (i.e. P X1

Basis risk = ∆PX-∆PH Because P X = P R±D we can derive:

↔ (P R0 +D 0) - (P R1 +D 1) - (P H1 -PH0 ) For reasons of simplification we write ‘+D’. We can thus solve this to:

↔ P R0 +D 0-PR1 -D1-PH1 +P H0 Which in turn can be simplified to:

↔ ∆PR-∆PH+∆D. This firstly shows that misleading price signals are not of concern for risk management, which is purely about spread risk. Thus, marker crudes not accurately reflecting the supply and demand conditions of the respective import market may of course lead to misleading price signals, but if we are purely concerned with risk management this is not of inter-

29 Remember that the price structure in formula pricing is P X = P R ±D. 30 Selling short means he does not own the futures contract yet in t0. 31 Equivalent considerations can be done for the case of hedging with other financial instruments such as contracts for differences. 32 Pirrong, ‘Economics of Commodity Trading Firms’, sec. III.

-11- Rethinking the Asian Marker Debate est.Secondly, the marker crude is of special importance as it requires a functioning financial market around it 33 , which would ideally lead to an exclusion of the first component of basis 34 risk (i.e. ∆PR=∆PH ).Usually, however, some basis risk is accepted because of liquidity considerations, meaning that a hedging instrument is chosen that is not perfectly linked to the marker crude because the liquidity of this financial instrument – and resulting lower transaction costs – offset the incremental basis risk of this mismatch.This means that finding a hedging instrument that is better suited in terms of price correlation is either too time consuming or too expensive to buy/sell, which in the end would render this hedging technique unprofit- able.Hence, the advantage of trading standardized and liquid derivative contracts, such asIn- tercontinental Exchange (ICE) Brent futures, is the comparatively low transaction costs resulting from exactly this easiness of finding and entering/exiting a contract at relatively low costs in a heavily traded market.Evidently, this entails a positive feedback mechanism leading to the dominance of a small number of heavily traded standardized hedging instruments for a given commodity. The more market participants trade a given contract – i.e. the more liquidity is attracted – the more the transaction costs of trading this contract are reduced. This, in turn, incentivises even more trading of this contract. 35 2.2 Two-Dimensionality of Benchmarks As mentioned in the introduction tosection2.1.2, it can be argued that the price risk management function of benchmarks is of paramount importance compared to the price discovery service. This position, which can be frequently found throughout the Asian marker debate, is based on the idea that market participants will always be exposed to flat price risk no matter how accurate the price signal of the benchmark is. Hedging – i.e. the transformation of flat price risk into spread/basis risk – constitutes a well-established way to manage this risk. For the process of hedging the benchmark plays a crucial role as shown above. Additionally, the use of formula pricing with a marker price (theoretically) derived from a spot market enables to share the risk between buyer and seller which is a second dimension of risk management enabled by means of a benchmark. This line of thought culminates in the position that the benchmark figure is only derived in order to hedge, as this process requires a concrete number which can easily be linked to a financial instrument. According to this interpretation, the performance of a particular benchmark can be exclusively measured by its ability to allow successful hedging. As shown above, this is in the end comes down to liquidity as the sole criterion for a functioning benchmark. Hence, this interpretation, in fact, concludes that a benchmark without hedging is meaningless and thus entirely neglects the price discovery function. It is, however, not convincing that the existence of flat price risk, and the consequent importance of a benchmark for hedging, should make the function of price discovery unnecessary. This, in fact, only allows us to derive that risk management will always be a required objective of benchmarks, but not that the other objective – price discovery – can be neglected. The relevance of the price discovery function lies in its ability to divert crude oil trade flows from

33 This financial layer does not necessarily have to stem from the benchmark itself. In case of the Argus Sour Crude Index, that Saudi Arabia started using as a substitute for WTI in 2010, hedging is still done via the WTI contract to which ASCI is linked (Fattouh, Anatomy , 60). 34 This holds if we use a financial instrument H linked to the crude R to hedge the price of crude X which is priced off the marker R, as then P H0 =P R0 when short selling the futures contract in t 0 and at closure of the financial position in t 1the relation PH1 =P R1 holds. 35 Pirrong, ‘Economics of Commodity Trading Firms’, sec. II.

-12- Rethinking the Asian Marker Debate one market to another, depending on the respective prices reflecting the underlying market fundamentals. In this sense, functioning price signals are of vital importance for the correct functioning of arbitrage mechanisms as discussed earlier using the example of Dated Brent. Additionally, crude marker prices are used by governments for taxation purposes and by financial players for the settlement of derivatives. 36 If one tries to debate the relevance of the price discovery function by questioning its importance for risk management, as the above presented position does, one sets up an artificial conflict between two categories that cannot be set into relation with each other. Price discovery and risk management are two separate objectives of benchmark crudes and should be treated as such. In this sense, the debate about benchmarks in general, and the evolvement of an Asian marker in particular, should ac- knowledge this two-dimensionality of marker crude’s objectives. In the following analysis on whether an Asian marker located in Asia is required, I will draw on this structure to clarify the existing arguments and critically evaluate them.

3 Oil Markets in Transition 3.1 Market Fundamentals Over the last years structural changes in international crude oil markets have attracted a lot of attention to the international crude pricing system and oil price benchmarks in particular. Ac- cording to Imsirovic Asia has become “[…] the new center of gravity for the oil markets” 37 , not only altering the trade flows of crude but also potentially having significant impact on the global landscape of crude oil price benchmarks. For a long time Asian consumers where se- verely dependent of Gulf crude exports. Even in 2011 over 14 million barrels per day (b/d) of crude imports to the Asian market came from the Middle East, equalling 57% of the region’s total oil imports. 38 Compared to the share of Gulf crude imports for the USA (17%) or Europe (21%) this extreme dependence becomes even more evident. It is this market power of Gulf producers that is perceived to be the origin of the so-called ‘Asian Premium’, the claim that Asian customers are paying more for their crude imports than those in the USA or Europe. Several studies quantify the Asian premium to about $1.00-$1.50 per barrel for the time period 1990 to the early 2000’s. However, recently it has also been argued for the occasional appearance of the opposite phenomenon, i.e. an ‘Asian discount’, as Gulf producers struggle for Asian market share. 39 This is mainly due to two structural changes in international oil markets. Firstly, the production of unconventional hydrocarbons in North America adds more crude supply to the market. Between December 2011 and December 2012 US crude oil production increased by about 900,000 b/d 40 and has, according to the International Energy Agency (IEA)41 ,now reached output levels above 10 million b/d. 42 This new domestic supply effects global oil markets as North America becomes less reliant on crude imports. According to a report by IHS CERA, US net crude imports peaked at 12.5 million b/d in 2005 and by 2012

36 Fattouh, ‘Anatomy’, 24. 37 Imsirovic, ‘Asian Oil Markets in Transition’. 38 Note that the specific dependence can vary significantly for different countries within the Asian market (cf. to Imsirovic, ‘Asian Oil Markets in Transition’). 39 Imsirovic and Doshi, ‘Exploring Asian Premium’. 40 Burkhard, Houlton, and Nguyen, ‘America’s Great Oil Revival’, 1. 41 Note that according to the US Energy Information Administration (EIA) the US crude oil production is still slightly below the 10 million b/d mark (U.S. Energy Information Administration, ‘U.S. Field Production’). 42 International Energy Agency, ‘World Energy Outlook 2014’, 96.

-13- Rethinking the Asian Marker Debate they had fallen by 40% to 7.5 million b/d. This figure is expected to fall by at least another 2 million b/d by 2020. 43 According to the IEA, the remaining imports to Canada, Chile, Mexico and the USA by end of the decade will be mostly heavy, sour grades from Latin America and the Middle East (mainly Saudi Arabia). 44 Fattouh and Sen specifically investigated the influence of this phenomenon on West African crude exports to North America and showed that these could, in fact, entirely come to an end by the first quarter of 2015. 45 This crude will have to be relocated and will increasingly flow into the Asian market as WTI- and Brent-linked crudes are relatively cheap in an oversupplied Atlantic basin. In September 2014, for example, 2.1 million b/d of West African crude alone where scheduled to flow into Asia. 46 And IHS CERA expects crude trade flows from Latin America to Asia to rise from 700,000 b/d in 2011 up to 1.6 million b/d by 2025. 47 Venezuela has been supplying China and India with oil on favourable pre-payment terms, Russia is supplying Asia with more oil through its East Siberia-PacificOcean (ESPO) pipeline and even North Sea crude regularly heads eastward given weak European demand. It is this additional eastbound supply from North and West Africa, South America, Russia and the North Sea that Gulf producers now have to compete with. 48 Additionally,US crude has been directly shipped to the Asian market: In October 2014 the first cargo of medium, sour Alaska North Slope 49 in more than a decade arrived in South Korea. If the US legislation on crude oil exports is softened in the future, US crude heading eastward may become a regular market feature, intensifying the supply diversification and competition for market share in Asia even more. 50 Secondly, demand for crude oil isshifting towards Asia and will increasingly do so in the future,despite the recent slowdown of the Chinese economy. China’s oil demand (including Hong Kong) has increased from 7.2 million b/d in 2005 to currently 10.5 million b/d (12% of global oil demand). 51 Furthermore, China’s crude imports have increased from less than 2.5 million b/d to 6 million b/d over the past decade. 52 In 2014, however, Chinese oil demand growth has slowed down due to weaker economic activity. Various forecasts for 2015 expect growth in Chinese oil demand to pick up again, reaching rates between 2.8% and 3.6% (which is still lower than previous growth rates around 4%). 53 Despite these lower growth rates Asia/Oceania is expected to overtake the Americas as the world’s largest oil consuming region in 2015. 54 Furthermore, over the forecast period 2013-2019 the IEA expects the strong- est growth in oil demand to be in the Asia/Pacific region. During this time period Asian/Pacific oil demand is forecasted to annually grow by 655,000 b/d on average, as opposed to 307,000 b/d for the Middle East, 170,000 b/d for Africa, 100,000 b/d for the Former

43 Burkhard, Houlton, and Nguyen, ‘America’s Great Oil Revival’, 6. 44 International Energy Agency, ‘Oil Medium-Term Market Report 2014’, 121. 45 Fattouh and Sen, ‘New Swings’, 6–7. 46 Argus Media, ‘Race to the Bottom’. 47 Burkhard, Houlton, and Nguyen, ‘America’s Great Oil Revival’, 6. 48 Argus Media, ‘Race to the Bottom’; Imsirovic, ‘Oil Markets in Transition and Dubai’, 3; Imsirovic, ‘Asian Oil Markets in Transition’. 49 Alaska North Slope is currently one of the exemptions to the US ban on crude oil exports (U.S. Energy In- formation Administration, ‘Exports’). 50 Imsirovic, ‘Oil Markets in Transition and Dubai’, 3; Energy Intelligence, ‘Mideast Rethinks Marketing’. 51 Deloitte, ‘Oil and Gas Reality Check 2014’, 7. 52 Argus Media, ‘Out Dated’. 53 Energy Intelligence, ‘Key Pillar’. 54 International Energy Agency, ‘Oil Medium-Term Market Report 2014’, 24.

-14- Rethinking the Asian Marker Debate

Soviet Union, 70,000 b/d for the Americas and a decline of 30,000 b/d for Europe. 55 The eastward shift in oil demand also mirrors itself in the refining industry. The IEA expects net refining capacity to increase by 16 million b/d between 2013 and 2040. Hereof China accounts for one-third, closely followed by the Middle East and India. Other regions of significant net refining capacity growth are Africa, Brazil and Southeast Asia. Furthermore, the excess of refining capacity over refinery runs will continue to rise and thus intensify pressure on refining margins. This will most likely lead to further closure of refineries in areas of weak demand and small margins, notably the already struggling refining industry in Europe. 56 To sum up, this eastward shift of crude oil demand as well as additional supply from North America result in international crude oil trade flows shifting towards Asia. According to the IEA, two out of three barrels of crude oil traded internationally by 2040 will be destined for Asia (up from less than 50% today). Furthermore, the Atlantic basin’s net crude inflows have fallen to 2.25 million b/d in September 2014. 57 According to Argus this is the lowest figure since at least two decades and down from 4 million b/d in 2004. 58 All in all, we see the Asian region gaining market power because of both the diversification of its supply as well as the demand shift for crude oil. In fact, we see a reversal of the traditional situation of dependence between Middle Eastern producers and Asian customers – with Asia in a position of strength. This led Imsirovic to proclaim the formation of a ‘buyer’s market’ (i.e. a market in which buyers have a stronger position than sellers), with Asian customers occasionally buying Atlantic ‘swing barrels’ 59 (e.g. West African crude) to meet their demand. 60 In order to manage these arbitrage flows on their own, rather than depending on trading companies or oil majors, Asian oil companies are increasingly integrating their trading operations. 61 With this more active position in crude trading comes greater influence on the price setting process of global oil prices as the incremental barrel of oil – the so-called ‘marginal barrel’ of oil which sets the price – increasingly heads towards Asia. 62 In this sense, Asian market participants are becoming more and more involved in the price setting process. The Chinese Unipec and China Oil – trading arms of the national oil companies Sinopec and China National Petroleum Corporation (CNPC) – as well as the South Korean refiner SK Energy have become the major players in the Dubai Platts window. 63 Asian influence on the price assessment of Dated Brent has also risen by occasional large buying of the Forties grade by Chinese and South Korean market players. As this grade – that usually sets the price for the Dated Brent assessment – becomes more expensive with rising demand, the overall price signal for Dated Brent may be significantly influenced. 64 According to Imsirovic as well as Energy Intelligencethis constitutes a transformation from ‘price makers’ to price takers’, meaning that Asian market participants start influencing the price setting process themselves

55 Ibid., 27. 56 International Energy Agency, ‘World Energy Outlook 2014’, 131–132. 57 Ibid., 95. 58 Argus Media, ‘Out Dated’. 59 These are barrels that are bought on the spot market depending on their relative price and the specific demand situation at a point in time. In this sense, these barrels may swing between one customer and another (Imsirovic, ‘Asian Oil Markets in Transition’). 60 Imsirovic, ‘Oil Markets in Transition and Dubai’, 3. 61 Imsirovic, ‘Asian Oil Markets in Transition’. 62 Imsirovic, ‘Oil Markets in Transition and Dubai’, 9. 63 Imsirovic, ‘Asian Oil Markets in Transition’. 64 ‘Asian Buyers Influence Key Oil Benchmarks’.

-15- Rethinking the Asian Marker Debate rather than solely using a price generated by other players in the window. 65 This increased Asian interest in a more active role in the trade of crude is also visible in the large investments in storage capacity. Currently South Korea is Asia’s major oil storage hub and will continue to grow, e.g. by means of the Ulsan storage facility project (28.4 million barrels capacity added by 2020). In addition, projects like Sinopec’s planned storage terminal in Batam, Indonesia (16.5 million barrels capacity) intend to establish a Singapore oil hub re- sembling Europe’s Amsterdam-Rotterdam-Antwerp trading hub. For Middle Eastern crude producers the consequence is that they face more severe competition and have, in fact, started fighting for Asian market share. The Iraqi oil corporation State Oil Marketing Organization, for example, has started discounting its eastbound exports relative to the Saudi grade Arab Light. Term contracts have been offered on extended credit terms by Iraq, Iran and Kuwait in order to secure large contracts. The Abu Dhabi National Oil Company dropped the traditional destination restrictions on their crude in order to make it more attractive for customers. Last but not least, storage facilities in the Asian region are increasingly being leased in order to have more direct access to Asian customers. 66 A prominent example is the renewal of the storage deal between Saudi Aramco and Japan Oil, Gas and Metals National Corporation in 2014. The storage facility on the Japanese island Okinawa will allow Saudi Aramco to store up to 6.3 million barrels for another three years (up from 1.3 million barrels in the preceding contract). This makes Saudi Aramco’s storage capacity on Okinawa Island significantly big- ger than their facility in Rotterdam (3.9 million barrels of capacity). Together with the fact that Saudi Aramco gave up its storage lease for 5 million barrels of crude in the Caribbean in 2009 (which is now utilized by China), this has been seen as a clear sign that Saudi Aramco is shifting its focus to Asia as its key export market. 67 3.2 Market Regulation Another major trend for international crude oil markets, that could intensify the above- described eastward shift of oil pricing, is the potential regulation of crude oil price benchmarks in both the USA and Europe. In July 2010 a financial reform was signed in the US mainly attempting to prevent a repetition of the 2008 financial crisis. The so-called ‘Dodd Frank Wall Street Reform and Consumer Protection Act’, that is not yet fully implemented, consists of three key components and also has consequences for speculation on energy markets, including oil markets. Firstly, the introduction of position limits in order to prevent traders from exerting too much influence on prices. This is supposed to prevent price manipulation but may also adversely influence financial liquidity. Secondly, the ban ofbanks’ direct involvement in proprietary trading for their own accounts. The legislation solely indirectly allows this via investments of up to 3% of their tangible equity in hedge and private equity funds. Hence, the reform supports the separation of servicing clients and speculative transactions, but again may negatively influence liquidity as banks increasingly exit commodity trading. Thirdly, most standardized over-the-counter (OTC) derivatives will be required to be traded on regulated exchanges in order to increase transparency and reduce default risk by means of a clearinghouse. Some market players will still be allowed to manage their price risk on OTC markets (such as petroleum end-users, producers, refiners or airlines). However, these exemptions (i.e. all players not categorized as ‘swap dealers’) lacks conceptual clarity

65 Imsirovic, ‘Asian Oil Markets in Transition’; Energy Intelligence, ‘Asia From Price Taker to Maker’. 66 Imsirovic, ‘Asian Oil Markets in Transition’. 67 ‘Saudi Arabia Increases Focus’.

-16- Rethinking the Asian Marker Debate and may lead players to exit the market due to the additional transaction costs of trading on an exchange. 68 In the European Union a discussion on regulating financial benchmarks is also taking place. In light of high oil price volatility the G20leaders asked the International Organization of Securities Commission (IOSCO) to examine the role of price reporting agencies for international oil markets in November 2011. One year later IOSCO’s report withdrew its initial plans to regulate price reporting agencies and instead suggested a set of principles to be followed voluntarily. 69 These guidelines are, among others, improvements of methodologies and transparency and the introduction of procedures against price manipulation as well as formal complaints processes. These principles are thus mostly in line with the so-called ‘Price Re- porting Code for Independent Price Reporting Organisations’ (IPRO Code), a list of self- regulatory guidelines, that the price reporting agencies Argus, Platts and ICIS formulated on 30 th April 2012 in anticipation of potential regulation. In September 2014, IOSCO published a report monitoring compliance with its principles. It states that the price reporting agencies have made good progress in implementing the suggested guidelines and speaks against a further alignment of the guidelines for price reporting agencies with those for financial market benchmarks. 70 This is a clear statement against regulation of price reporting agencies and crude oil price benchmarks in times of increased scrutiny by EU regulators. After the financial crisis in 2008 – and even more so in the aftermath of the London Interbank Offered Rate (Libor) manipulation scandal in 2012 – discussions on regulating financial benchmarks intensified in the EU. Additionally, the European Commission raided offices of major oil companies and price reporting agencies in April 2013 and is currently investigating potential price manipulation (mainly concerning Platts’ Dated Brent assessment). 71 The European Commis- sion’s regulatory proposal intends to move regulation of benchmarks to the Paris based Euro- pean Securities and Markets Authority (ESMA). Especially the idea of making the manipulation of benchmarks a criminal offense has led to severe concerns. This means that oil companies could sue a price reporting agency if its price assessments are not accurate. Obviously this would be a major source of risk for price reporting agencies and could seriously challenge Brent’s benchmark status. 72 The proposed legislation is expected to become effective in 2015, so that regulatory details and their impact on price reporting agencies as well as their benchmark assessments remain yet to be seen. 73 All in all, there are serious concerns that the planned regulation of crude oil price benchmarks in both the USA and Europe could incentivise market players to move to less regulated jurisdictions. Such regulatory arbitrage could significantly undermine the current importance of WTI and Brent by diverting financial liquidity to alternative, emerging benchmarks. 74 Hence,

68 Energy Intelligence, ‘US Financial Reform’; Energy Intelligence, ‘Will Tougher Rules Spur New Bench- marks?’. 69 One major reason for this decision is that IOSCO does not see systemic risks for the financial system, as it is the case for other financial benchmarks under regulatory scrutiny (Energy Intelligence, ‘Price Reporting Agen- cies Won’t Be Regulated’). 70 Ibid.; Energy Intelligence, ‘PRAs Have Made “Good Progress”’; Argus Media, ‘Energy Price Reporting Or- ganisations’. 71 Energy Intelligence, ‘PRAs Have Made “Good Progress”’; ‘Price Reporting Probe’. 72 Energy Intelligence, ‘EU Brings Oil Into Libor Regulation’; Energy Intelligence, ‘EU to Bring Out Bench- mark Legislation’. 73 European Commission, ‘Citizens’ Summary’. 74 Energy Intelligence, ‘Will Tougher Rules Spur New Benchmarks?’.

-17- Rethinking the Asian Marker Debate one can conclude that Asian buyers are not only about to exit their position of dependence from the Middle East but may also play an increasing role in crude pricing as Western regulation diverts financial liquidity, and thus crude price benchmarks, towards Asia.

4 The Asian Marker Debate TheAsian marker debate deals with the question which crude will position itself as the predominant regional reference crude for exports to the Asian market.As Asia is and will increasingly be a major import market for crude oil, exporters should use a specific marker crude in order to price their exports into the Pacific basin.Keeping the two objectives of benchmarks in mind this marker should ideally reflect the fundamentals of the market the crude, it is pricing, is exported to (price discovery) and should be linked to a liquid financial market in order to allow effective hedging (risk management). In terms of price discovery the regional marker for Asia (the so-called ‘Asian marker’) should be a crude reflecting the supply and demand conditions of the Asian market. This in turn means that ideally the Asian marker would involve price discovery taking place in Asia, i.e. prices being generated by Asian buyers and sellers either at point of loading (for physical benchmarks) or at an Asian exchange (for financial benchmarks). Further this will be referred to as a regional marker for Asia ‘located in Asia’ , thus reflecting Asian fundamentals.However, currently used markers for Asia do not fulfil this theoretical condition necessary for accurate price signals. In order to cope with this imperfection the market came up with two solutions:Firstly, the utilization of markers ‘located outside of Asia’ and the application of price spreads to this benchmark figure in order to derive the price for the Asian market. An example of such a spread is the Du- bai/ESPO spread, linking the Middle East and the Sea of Japan. Secondly, a focus on the other objective of crude price benchmarks, i.e. risk management through hedging, as exemplified by the usage of Dated Brent or Platts’ Dubai which both benefit from the deep liquidity of the financial layers around the physical Brent market. The current debate on the Asian marker very much focuses on the question which of the candidates located in Asia will establish itself as the predominant marker for the Asian basin. Generally speaking, there is the consensus underlying the debate that the reference crude should also originate from Asia, given the previously described structural changes in international oil markets. In the following I will argue against this position, which is commonly taken for granted. In so doing, I shall present the existing benchmark crudes for the Asian basin as well as prominent candidates. Firstly, the currentoptions located outside of the Asian market will be presented before introducing the two candidates located in the Asian market. Note that the focus will be put on the most prominent candidates. Marginal suggestions such as Canadian grades (Syncrude Sweet Blend, Western Canadian Select) or ICE’s Brent futures price index ‘Brent 1 Minute Singapore Marker’ will not be evaluated. 75 Throughout the discussion it is useful to keep some general criteria for benchmark crudes in mind – although not all existing reference crudes entirely fulfil these criteria at every point in time. Firstly, a physical benchmark should have a production level of about 500,000 b/d in order to reduce risk of price manipulation. In addition to this, volumes available for spot trade should be sufficient to guarantee transparent and efficient price discovery. Secondly, the benchmark crude should be freely tradable. This criteria entails the absence of destination restrictions, no or few barriers to entry for market players and functioning infrastructure without bottlenecks.

75 Energy Intelligence, ‘Battle of Asian Crude Benchmarks’; Imsirovic, ‘Oil Markets in Transition and Dubai’, 9.

-18- Rethinking the Asian Marker Debate

Thirdly, the crude should be of a quality utilized by a variety of buyers and similar to other grades, thus allowing it to be used as benchmark for these. Fourthly, production should take place under diversified equity ownership and political stability, thus reducing risk of manipulation of production volumes. Fifthly, a liquid spot/forward market with a variety of buyers/sellers is required in order to ensure financial liquidity, consequently reducing risk of price manipulation. Last but not least, the reference price should be generated by a variety of market players and the price signal should be transparent and accessible to the market at large. 76 4.1 Candidates Located Outside of Asia 4.1.1 The Platts Dubai/Oman Assessment Currently, Platts’ assessment of the Dubai price is the predominant benchmark for global crude exports to Asia, basis for pricing of almost 30 million b/d. 77 This figure encompasses pricing based exclusively on Dubai as well as on combinations of different markers, including the Dubai price. Most crude exports from the Gulf region to Asia, for example, are priced against Platts’ price assessments of Dubai or Oman, respectively a combination of the two. Currently, for instance, Saudi Arabia, Iran, Iraq and Kuwait use the arithmetic average of Platts’ Dubai and Oman price assessment as the benchmark for their combined 13 million b/d of crude exports to Asia. 7879 In the following both price assessments, their origins and development over time shall be presented. The medium, sour Dubai crude (API gravity 30.4°, sulphur content 2.13%) 80 evolved as the main marker for crude exports from the Middle East to Asia in the mid 1980’s. This is mainly because it was one of the few Gulf crudes available on spot market, as the Saudi Arabian crude Arabian Light stopped being traded on the spot market around 1984.Additionally, equity ownership in the production of Dubai crude was accessible for international oil companies at that time, which was unusual for the region and avoided concerns about concentration of power in the production process. Until 2007 the concession for the major production fields offshore of Dubai (Falah, Fateh, South West Fateh and Rashid) belonged to an international consortium inter alia including ConocoPhillips, Total and Repsol. This, however, changed in April 2007 as the concession was passed on to the entirely government owned company Du- bai Petroleum Establishment. Hence, Dubai does no longer satisfy the ownership diversification criterion, which is however of subordinate importance as the Dubai price assessment is nowadays heavily dependent on Omani crude, as will be explained in the following. In the beginning the Dubai market was rather illiquid, with only few market participants being involved in small volumes of trade. This changed as Japanese trading houses and Wall Street refiners entered the market between 1985 and 1987. However, the Dubai market only really took off in 1988 when important members of the Organization of the Petroleum Exporting Countries (OPEC) started pricing their crude exports to Asia with formula pricing based on

76 Montepeque and Stewart, ‘Sour Crude Pricing’; Fattouh, ‘Origins and Evolution’, 60–61; Horsnell and Mabro, Oil Markets and Prices . 77 Imsirovic, ‘Oil Markets in Transition and Dubai’, 2. 78 Energy Intelligence, ‘Dubai Looks for Liquidity’; Fattouh, ‘Anatomy’, 61. 79 Throughout this paper I will often solely refer to the “Dubai benchmark”, following Platts’ focus on the Dubai price. It shall, however, be kept in mind that the Dubai marker is often used in combination with Platts’ Oman assessment (Imsirovic, Oil Markets in Transition and Dubai , 7–8). 80 Davis, ‘Crude Pricing’, 29.

-19- Rethinking the Asian Marker Debate

Dubai as the benchmark crude. 81 This quickly made Dubai the predominant marker for crude exports to Asia, so that already in 1993 Horsnell and Mabro referred to it as the “Brent of the East” 82 . Since then the Dubai benchmark has been amended several times and has developed to a brand name, as it does not anymore only consist of crude from Dubai fields. This has to do with the declining production in Dubai, which dropped from a peak of 400,000 b/d between 1990 and 1995 to less than 50,000 b/d today.Clearly, these low production levels and resulting low trading activity can make the price discovery process on the spot market difficult. With less than four cargoes offered per month, of which some are bound to longterm contracts, a functioning spot market is virtually non-existent. 83 Accordingly, the price of Dubai is assessed by oil price reporting agencies. Platts, for instance, assesses the price of Dubai based on concluded transactions in the ‘Platts window’. If the concluded deals are not sufficient for price discovery, Platts uses information from bids and offers in order to generate a price signal. If this information is not sufficiently available either, Platts calculates the required price from the financial layers surrounding Dubai (i.e. the Dubai swap market). This means that national oil companies in the Gulf do not actively set the price of oil, although their large volumes of physical production would allow them to do so. The pricing power rather lies in the hands of a small group of traders participating in the Platts window(notably Shell and Vitol). The oil exporting countries do not participate in this price setting process, but simply use the price assessments generated in the Platts window for their pricing formula. Although their production behaviour does have certain effects on price levels, this allows the oil exporters to somewhat free themselves from concerns of price manipulation. 84 Because of Du- bai’s declining production Platts had to amend their price assessment several times in order to maintain the viability of the price signal. In this sense, Dubai is now a brand or an index, representing a basket of mid sour grades. For instance, in 2001, Platts allowed the delivery of Omani crude against Dubai contracts in the Platts window through the so-called ‘alternative delivery mechanism’. 85 Note that this means that Oman has a twofold role in the pricing of most Gulf exports to Asia: Firstly, through its involvement in the assessment of the Dubai price and secondly because of its 50% influence in the averaged Dubai/Oman benchmark. Oman’s production of crude oil and condensates reached 950,000 b/d in the third quarter of 2013, of which typically 750,000 b/d are exported. Hence, Oman’s output has reached levels similar to those of the combined BFOE-system (i.e. Brent, Forties, Oseberg and Ekofisk) and does fulfil the physical liquidity requirements of a benchmark as stated by Montepeque and Stewart. 8687 As Oman is not an OPEC-member its production is not subject to their quotas, thus decreasing risk of abrupt production declines. Additionally, Oman allows reselling of cargoes which is important for the development of a liquid and freely-traded secondary mar-

81 Fattouh, ‘Anatomy’, 61. 82 Horsnell and Mabro, Oil Markets and Prices , 210. 83 Fattouh, ‘Anatomy’, 61–62; Dubai Mercantile Exchange, ‘Brent Oman Spread’; Energy Intelligence, ‘Dubai Looks for Liquidity’. 84 Fattouh, ‘Anatomy’, 64–65; Energy Intelligence, ‘Dubai Looks for Liquidity’. 85 Fattouh, ‘Anatomy’, 61 ff. 86 Dubai Mercantile Exchange, ‘Oman Crude Oil & DME’; Montepeque and Stewart, ‘Sour Crude Pricing’; Platts, ‘Dubai Crude Oil Price’. 87 Note, however, that expansions of Omani refining capacity may increase domestic usage, thus reducing export volumes by 327,000 b/d (‘Price Reporting Probe’).

-20- Rethinking the Asian Marker Debate ket 88 . However, the Omani oil production is dominated by the national oil company Petro- leum Development of Oman (PDO), which is responsible for 70% of Oman’s oil production. 89 PDO is 60% owned by the Omani government. The remaining shares are held by Shell (34%), Total (4%) and the Portuguese oil and gas company Partex (2%). The Oman Blend is exported from the Mina al Fahal terminal, which benefits from its location outside of the chokepoint Strait of Hormuz and thus significantly decreases the vulnerability to geopolitical tensions. 90 The involvement of Oman has, however, created new problems. Omani crude has lower sulphur content (1.14%) and higher API gravity (32.95°) than Dubai crude and is thus of higher quality. 91 Occasionally, the price differential between Dubai and Oman can reach a dollar per barrel or even more, which is difficult to combine with the fact that these two crudes form ‘the Dubai’ price. 92 As a consequence another crude was included in the price assessment process, which is closer to the price of Dubai than to Oman. 93 This is the Abu Dhabi crude Upper Zakum (API gravity 33.9°, sulphur content 1.84%),which was added to the Du- bai brand in 2006. The field is jointly operated by Abu Dhabi National Oil Company (60%), ExxonMobil (28%) and Japan Oil Development Company (12%) and production is about 650,000 b/d. 94 The widening of the Dubai benchmark 95 effectively sets a cap for the Dubai price. It thus removes possibility forprice manipulation leading to exorbitant prices as Upper Zakum should only be delivered if the Dubai price reaches the market price of Upper Zakum (analogously for Omani crude, setting a second price cap on a higher price level). 96 All in all, Platts now allows for alternative delivery of Omani or Upper Zakum crude against a Dubai contract in the Platts window, thus establishing a physical foundation of over 1.5 million b/d of which normally around 1.0 million b/d are freely tradable (i.e. not bound in longterm contracts). 97 Additionally, in 2004 Platts introduced the so-called ‘partials mechanism’ in order to increase trading volumes and thus the efficiency of price discovery. This mechanism allows partition- ing a cargo into smaller bundles that can then be traded, so-called ‘partial contracts’. The smallest trading unit is set to 25,000 barrels. Physical delivery, however, only occurs if buyers trade a minimum of 20 partials (i.e. 500,000 barrels) with a single counterparty. Quanti- ties traded below this barrier will not be physically delivered but cash settled. 98 This mechanism results in a significant reduction of price risk involved in trading partial contracts as compared to the standard single Dubai cargo of 500,000 barrels. This has attracted new players to trade in the Platts window. Whereas equity producers and trading companies where the dominant market participants before the introduction of the partials mechanism (due to the

88 A ‘secondary market’ is a financial market on which previously issued financial contracts, such as options and futures, are traded. Thus, an exchange or an OTC market are examples for secondary markets. 89 There are several international oil companies operating besides PDO, of which Occidental Petroleum is the largest single producer (Dubai Mercantile Exchange, Oman Crude Oil & DME) . 90 Ibid. 91 Davis, ‘Crude Pricing’, 29. 92 Imsirovic, ‘Oil Markets in Transition and Dubai’, 7–8. 93 Fattouh, ‘Anatomy’, 61 ff. 94 Platts, ‘Dubai Crude Oil Price’; ExxonMobil, ‘About Upper Zakum’. 95 Note that further extensions of the benchmark are not unthinkable. In early 2014 Platts said that it is discuss- ing the addition of Qatari and Iraqi crudes to the Dubai brand, in order to widen the pool of deliverable crudes given the political instability in the Gulf region (Energy Intelligence, ‘Dubai Looks for Liquidity’). 96 Imsirovic, ‘Oil Markets in Transition and Dubai’, 5. 97 Platts, ‘Dubai Crude Oil Price’. 98 Binks, ‘Middle East Crude Pricing’, 29; Argus Media, ‘Shell Buying Supports Dubai’; Platts, ‘Review of Partial Mechanism’, 2.

-21- Rethinking the Asian Marker Debate high risk associated with full cargo trading), now also Asian refiners such as China’s Unipec and China Oil as well as South Korea’s SK Energy are involved in the price discovery process. This, in turn, has led to an increase of trading volumes in the Platts window and an improved efficiency of the price signal, as implied by smaller bid/offer spreads. 99 This initial success, however, was not maintained in the period between 2008 and 2010 when partial trades where concluded on only 50% of the trading days. 100 As Imsirovic points out this has only recently changed with a record level of partials traded in the beginning of October 2014. This matches the development of liquidity in the Platts Dubai window, which has been constantly increasing since 2012. According to Imsirovic this is mainly due to an oversupply of Dubai swaps and the desire of their owners to ‘fix’ the Dubai price through more active involvement in the Platts window. The excess supply of Dubai swaps, in turn, results from an intensified conver- sion of Brent-linkedswing barrels to Asia into Dubai pricesby means of buying Brent futures and selling Dubai swaps. These two trades can becombined in one transaction by means of buying a so-called Brent/Dubai Exchange of Futures for Swaps (EFS). 101102 However, it remains to be seen whether this phenomenon can establish itself as a common market feature, boosting the window’s liquidity beyond the shortterm. For the time being Platts relies on information from bids and offers as well as from Dubai-linked derivatives in order to generate a price signal for Dubai in case of insufficient liquidity in the window. The financial layers surrounding the physical forward Dubai market mainly consist of the above-mentioned Brent/Dubai EFS and the Dubai inter-month swaps markets. These are both OTC markets, thus information on trade volumes or market participants is limited. An own futures market around Dubai has so far not emerged, despite attempts to launch such a contract in London and Singapore in the early 1990’s. Buying Brent/Dubai EFS allows traders to transform a Brent futures position into a forward month Dubai swap plus a quality premium spread (be- causeBrent is lighter and sweeter).This, as described above,is done by combining two trades (buying Brent futures and selling Dubai swaps) in one transaction (buying Brent/Dubai EFS). Alternatively, market players can sell Brent/Dubai EFS in order to convert their Dubai price exposureinto Brent price exposure (i.e. selling Brent futures and buying Dubai swaps). This can be beneficial from a price risk management perspective, as Brent price exposure should be easier to hedge given the high liquidity of the Brent futures market. The Dubai inter-month swap represents the differential between the differential of a Dubai swap for one month and another. It allows hedging of Dubai price risk from one month to another, by locking in a given inter-month differential in anticipation of changes in this spread. 103 How the price generation via the financial layers around Dubai exactly works can be best examined when look- ing at the example of Argus’ price discovery methodology. This is because Argus uses the financial layers as the standard method for price discovery and not solely as a backup-option as in the case of Platts. 104 TheBrent/Dubai EFS price is listed as a differential to the ICE Brent futures contract. This enables Argus to calculate an outright price level for Dubai for a particular month by subtracting the EFS value from the ICE Brent futures price 4.30pm Singa-

99 Platts, ‘Review of Partial Mechanism’, 2. 100 Fattouh, ‘Anatomy’, 65. 101 Imsirovic, ‘Oil Markets in Transition and Dubai’. 102 For a detailed description see: Ibid., 3–4. 103 Fattouh, ‘Anatomy’, 62–63. 104 Binks, ‘Middle East Crude Pricing’, 28.

-22- Rethinking the Asian Marker Debate pore time. The resulting price is referred to as the Dubai swap price.This is, however, a forward price as Dubai is loaded two months in advance. This means that the price calculated from the EFS in December, for instance, is the price for Dubai delivered in February. In order to derive the price in December the inter-month Dubai spreads are used. By first subtracting the January-February Dubai swap price differential from the December price and then from this value subtracting the December-January Dubai swap price differential, one ends up with the forward price of Dubai in December. 105106 How the pricing signal functions in practice has been described by Fattouh: “For instance, strong Asian demand relative to Europe reduces Brent’s premium to Dubai, causing the Brent/Dubai EFS to fall and encouraging traders to send crude from the Atlantic Basin to Asia. The adjustment in the price differential is re- flected in a higher Dubai price level.” 107 For Oman, price derivation is relatively straightforward once the Dubai price has been generated. Both Platts and Argus compare Dubai and Oman in order to derive the Oman price. This means that the assessed Oman price is simply an extension of the Dubai market. Platts uses Oman/Dubai swaps in case of a lack of Oman partials traded in the Platts Dubai window and insufficient information on bids and offers. These contracts trade the differential between Oman’s OSP and the Dubai price for the respective month and thus can be simply used to calculate the outright Oman price. Argus uses a similar methodology, calculating the differential of Oman to Dubai swaps and then adding/subtracting this figure from the outright Dubai price in order to generate the forward Oman price. 108 Hence, the price of both Dubai and Oman can be assessed without relying on a physical foundation, but purely from the financial layers that have evolved around Brent, Dubai and Oman. 109 4.1.2 The DME Oman Crude Oil Futures Contract The Dubai Mercantile Exchange (DME), which is located in the free zone Dubai Interna- tional Financial Center and regulated by the Dubai Financial Services Authority, was established in 2006. 110 Its flagship contract, the Oman Crude Oil Futures contract was launched in 2007, with the goal to establish it as a crude price benchmark for Middle Eastern exports to Asia. As a financial contract therisk management function plays a crucial role for this objective – especially because no own financial layers have previously evolved in the region. 111 The contract trades on a month plus two basis, e.g. in January the front month contract is March, and one contract accounts for 1,000 barrels. The contract allows for financial settlement against physical delivery and the settlement price for a given month is calculated as the volume-weighted average price of trades for the respective month between 4.15pm and 4.30pm (Singapore time). 112 Note that due to its location in Dubai the DME allows to fill the gap between the closing time of European exchanges and the opening time in Singapore. 113 The DME is to 50% owned by the CME Group, the world’s largest futures exchange owner. This is a strong backup for the DME, not only because of the financial 105 Argus Media, ‘Methodology and Specifications Guide’, 16; Fattouh, ‘Anatomy’, 65–66. 106 For a numeric example see: Argus Media, ‘Methodology and Specifications Guide’, 16–17. 107 Fattouh, ‘Dubai Benchmark and Its Role’, 6. 108 Argus Media, ‘Methodology and Specifications Guide’, 17; Fattouh, ‘Anatomy’, 66. 109 Fattouh, ‘Dubai Benchmark and Its Role’, 6. 110 Saidi, Scacciavillani, and Ali, ‘The Dubai Mercantile Exchange’, 5; Dubai Mercantile Exchange, ‘Oman Crude Oil & DME’, 2. 111 Fattouh, ‘Anatomy’, 66 ff. 112 Saidi, Scacciavillani, and Ali, ‘The Dubai Mercantile Exchange’, 7–8; Dubai Mercantile Exchange, ‘Oman Crude Oil & DME’, 3. 113 IHS, ‘Dubai Mercantile Exchange Begins Trading’.

-23- Rethinking the Asian Marker Debate owner. This is a strong backup for the DME, not only because of the financial support, but most importantly due to the reputation and knowledge of CME Group. The CME Group, for instance, handles the clearing process. It is thus subject to US regulation by the Commodity Futures Trading Commission, which allows market participants to trade in a well-known and relatively stable regulatory environment. Additionally, the increase of CME Group’s equity share in 2012 to 50% from previously 25% was perceived as a clearsignal of confidence for the DME. Remaining stakes are held by the Oman Investment Fund (29%), Dubai Holding (9%) and other strategic investors (12%) such as Vitol, Shell, JPMorgan, Morgan Stanley or Goldman Sachs. 114 Throughout 2014 average trading volumes of the DME Oman futures contract rose by 33%, reaching an average of 8,431 lots/day 115 .116 Additionally, the barrier of 10,000 contracts traded daily on a monthly average was broken for the first time in February 2014 (10,764 lots/day).The quantity of 10,000 contracts used to be the daily trading goal of the DME, but has been abandoned as the contract struggled to attract sufficient trading activity. Its relevance stems from the fact that oil-producing countries, such as Saudi Arabia, indicated that this level was sustainably required for them to consider using the DME Oman contract as a marker for their Asian exports. 117 The average daily trading volumes are, however, still fairly volatile: After having reached almost 9,000 lots/day in September 2014 trading volumes bounced back to just above 6,000 lots/day in October 2014. 118 An interesting feature of the DME Oman futures contract, which allows settlement against physical delivery of Omani crude, is that a significant share of contracts actually results in physical delivery per month. 119 According to the DME, currently 10-17 million barrels per month are physically delivered through the DME Oman futures contract. This typically represents between 50 and 75% of the total Oman export program. These figures make the DME Oman contract the crude oil futures contract with the largest physical delivery in the world. 120 The Light Sweet Crude Oil Futures contract on the New York Mercantile Exchange (NYMEX), better known as WTI, for instance exceeded physical delivery of four million barrels per month only once in January 1995. That the DME Oman contract is strongly used as a mechanism for physical delivery of Omani crude also mirrors itself in the development of open interest of a given contract over a month. Open interest refers to the amount of contracts that have not been fi- nancially closed or delivered at a given point in time. For a contract primarily used as a hedging instrument it should decrease as expiry date of the contract approaches as traders aim to close their positions to avoid physical delivery. In the case of the DME Oman contract, however, open interest tends to increase as the expiry date of the contract approaches which is an anomaly and indicates that the contract is simply used to access physical Omani crude. 121 This is supported by the distribution of trading activity over a given day. Although latest fig-

114 Dubai Mercantile Exchange, ‘Oman Crude Oil & DME’; Dipaola, ‘CME Will Boost Oil Bourse Investment’; Energy Intelligence, ‘DME Hopeful of Futures Contract Breakthrough’. 115 One lot usually transfers to 1,000 barrels (for an exception see the planned Shanghai crude oil futures contract with one lot equalling only 100 barrels). 116 Platts, ‘Dubai Mercantile Exchange 2014 Traded Volume’. 117 Platts, ‘Interview’; Energy Intelligence, ‘Mitsubishi Helps DME Advance Asia Strategy’; Blas, ‘Rise of Oman’. 118 Dubai Mercantile Exchange, ‘Historical Data’. 119 Fattouh, ‘Anatomy’, 67. 120 Dubai Mercantile Exchange, ‘Oman Crude Oil & DME’, 4. 121 Fattouh, ‘Anatomy’, 67.

-24- Rethinking the Asian Marker Debate ures are not available, Energy Intelligence stated in 2011 that trading of the contract is most active during the five-minute window in which the daily settlement price is fixed. 122 All in all, this all allows the conclusion that the contract is used more as a cargo pricing mechanism than a risk management tool. This is not necessarily beneficial for the contract to attract liquidity, as concerns about the capability of the physical delivery mechanism as well as physical bottlenecks around the delivery point may discourage market participants to trade the contract out of fear of volatile price signals. 123 But what impact has the contract had on the pricing of Gulf crude? With its introduction the Omani government, which up till then used a retroactive official selling price, switched to a forward pricing system based on the DME contract to price its crude. Note, however, that the OSP for Omani crude for physical delivery does not have a physical foundation. It is calculated as the average of the daily settlement prices of the DME contract over the respective month. The OSP for June, for example, is the average of the daily settlement prices (for June) of contracts for delivery in two months (i.e. August). Besides Oman, Dubai is the only other oil-producing country that has (in 2009) 124 decided to price its exports off of the DME Oman futures contract up till today. This reluctance to adopt the DME contract as a benchmark for exports to Asia is mainly due to lack of liquidity, necessary both for price discovery and risk management. As discussed above, large amounts of contracts expiring to physical delivery may be detrimental for the contract. This factor is only reinforced as physical deliverability increases while trading liquidity remains low, as the transaction costs of offsetting one’s financial position before expiry date rise. This poses a barrier for trading the contract, in particular, for market participants not primarily interested in physical delivery of Omani crude but rather in the purely financial utilization of the contract as a hedging instrument. 125 Hence, the DME Oman futures contract faces a chicken-and-egg problem, as market participants are not attracted to a financial market with low liquidity and it is precisely this lack of interest that leads to insufficient liquidity.In 2008, Fattouh warned of this resulting in a vicious circle, as “[i]n the same way that liquidity attracts further liquidity, illiquidity can result in more illiquidity.”126 This is a typical problem for the successful rollout of a futures contract and the evolvement of crude price benchmarks (analogue to the positive feedback mechanism described in section 2.1.2). It is in this sense that the CME Group’s launching of new swap and option contracts 127 linked to DME intends to attract liquidity by providing new risk management tools. 128 Regarding the risk management service the Asian market is of particular importance, especially big Asian refineries which have been reluctant to trade the contract. 129 The increase of the CME Group’s equity share to 50% in 2012 as well as the appointment of new Chief Executive Officer (CEO) for the DME in the same year (Christopher Fix, succeeding Thomas Leaver) is part of the strategic shift towards Asia. Instead of primarily trying to con-

122 Energy Intelligence, ‘Oman’s Physical Limits’. 123 Fattouh, ‘Anatomy’, 68. 124 ‘Dubai to Base OSP on DME’. 125 Fattouh, ‘Anatomy’, 68 ff. 126 Fattouh, ‘Prospects of DME Oman’, 5. 127 In 2010/11, for instance, alone six new contracts have been launched: DMEOman Crude OilSwap Futures, DME Oman Crude Oil versus ICE Brent Swap Futures, DME Oman Crude Oil Average Price Option, Singa- pore Mogas (Motor Gasoline) 92 Unleaded versus DME Oman Crude Oil Swap Futures, DME Oman Crude OilBALMO (Balance of the Month) Swap Future and Singapore Gasoil versus DME Oman Crude Oil Swap Futures(Energy Intelligence, ‘DME Introduces Swaps’). 128 Ibid. 129 Fattouh, ‘Anatomy’, 68 ff.

-25- Rethinking the Asian Marker Debate vince Middle Eastern producers to adopt their Oman contract as a marker, the DME now focuses on Asian market participants to trade the contract in hope that Gulf producers will eventually follow as liquidity reaches sufficient levels. The strategic realignment also ac- knowledges that low liquidity and a conservative wait-and-see approach have up till now prevented a shift of Middle Eastern oil countries from the predominant Platts Dubai marker to the DME Oman contract. This is despite the fact that regional producers, who traditionally do not hedge their oil production, should have an interest in the contract from a perspective of price discovery. Especially Saudi Arabia is seen to play a crucial role as their shift would provide the required liquidity to incentivise other regional producers to follow. 130 In this sense, Asian traders and refiners shall now attract the liquidity required to incentivise Gulf producers to adopt the DME Oman contract. As the new CEO Fix put it in 2013: “If you ask where the next 5,000-10,000lots are going to come from, it's going to be the consumers.” 131 This new strategy is both linked to CME Group’s technological support and marketing net- work in Asia as well as Christopher Fix’s experience and reputation in the region. 132 Some hopeful results have been achieved so far, e.g. that the Japanese Marubeni Corporation and Idemitsu Kosan – as the first Japanese refiner – started trading the contract in 2014. Addi- tionally, a memorandum of understanding has been signed in March 2014 in order to strengthen the cooperation between DME and Japan’s Tokyo Commodity Exchange as to increase trading activity in both markets by means of joint services and marketing activi- ties. 133 To what extent this strategic realignment will prove to be successful in the long term, however, still remains to be seen. 4.1.3 The Platts Dated Brent Assessment Besides Platts’ Dubai/Oman marker and the DME Oman futures contract for sour Gulf crudes another crude price benchmark for the Asian market has evolved. Over the last years Dated Brent has increasingly been used as a marker for sweet crude exports from Asian producers to consumers in Asia. Sometimes Dated Brent is also referred to as ‘Dated BFOE’. This is because the Brent benchmark has repeatedly been widened in order to deal with dwindling production and now has turned into a brand name. It consists of four different North Sea grades (BFOE: Brent, Forties, Oseberg and Ekofisk) of which the cheapest at a time sets the price and is loaded at the Sullom Voe terminal in the Shetland Islands, UK. 134 According to Platts, the physical foundation of BFOE is currently 1 million b/d. 135 Energy Intelligence quantifies the average production for 2014 at around 800,000 b/d with BFOE volumes expected to sink to about 600,000 b/d by 2017, making a further widening of the benchmark required. New fields will not be able to prevent this, as the Norwegian field Johan Sverdrup will, for instance, only add a maximum of 100,000 b/d by late 2017. Hence, Platts will eventually have to include further grades into the BFOE basket. Regional grades as the Norwegian Gullfaks and Statfjord Blend are discussed as well as more geographically distant crudes such as the Russian Urals Blend, grades from West and North Africa, the Caspian and even Bra- zil. 136 This constantly widening of the Brent benchmark is not without criticism and has been

130 Ibid.; Energy Intelligence, ‘Dubai Exchange Names New CEO’; Energy Intelligence, ‘Oman’s DME Looks East’. 131 Energy Intelligence, ‘Oman’s DME Looks East’. 132 Energy Intelligence, ‘Dubai Exchange Names New CEO’; Energy Intelligence, ‘Oman’s DME Looks East’. 133 Fix, Setting a Benchmark; Platts, ‘Japan’s Tocom’. 134 For more detailed information on the history of the physical Brent market see: Fattouh, ‘Anatomy’, 37 ff. 135 Platts, ‘Dubai Crude Oil Price’. 136 Energy Intelligence, ‘Shrinking North Sea Output’; Energy Intelligence, ‘Boosting Benchmarks’.

-26- Rethinking the Asian Marker Debate argued to only be able to artificially prolong the lifetime of the benchmark in the short term. 137 Hence, whether Dated Brent will be a long term solution as a marker for sweet crude flows in the Asian market or rather a transitory occurrence is yet to be seen. However, for the time being this physical benchmark based North Sea grades has in large parts substituted the traditional regional markers for sweet crude in Asia: the Malaysian Tapis 138 for light sweet crude as well as the Indonesian grades Minas 139 and Duri 140 for medium, respectively heavy, sweet grades. 141 This is mainly because production from these regional grades is in decline, resulting in less crude sold on the spot market, which in turn casts doubts on the reliability of the generated price signals. Production of Tapis is currently around 300,000 b/d 142 , Minas production is around 420,000 b/d and Duri production at approximately 240,000-250,000 b/d. 143 New regional sweet crude benchmarks as well as Asian sweet grades suitable for widening the existing markers are not available. The existing alter- natives are either too different in quality, do not have sufficient spot volume or are located too far away. Hence, in absence of a reliable spot market price signal for Asian sweet grades the global marker for sweet crude, Dated Brent, has been utilized. Additionally, as regional Asian sweet grades decline in production volumes more sweet crude from West and North Africa as well as Central Asia has to be imported. These grades are often priced against Dated Brent, so that utilizing this reference price is beneficial for Asian producers and consumers in terms of price comparability. This results in more and more Asian-Pacific producers changing their reference crudes, so that Asian-bound grades fromAustralia, Malaysia, Papua New Guinea and Vietnam are currently priced against Dated Brent. 144 Malaysia, for instance, used to price its exports against a price assessment referred to as the ‘Asia Petro- leum Price Index’ (APPI) from the Hong Kong based company Seapac Services. This way of pricing crude is referred to as ‘panel pricing’ as the prices for the 19 crudes assessed – one of which is the Malaysian Tapis – are derived based on data submitted by 40-50 market participants. The Malaysian Petronas used to set a price differential (known as the ‘p-factor’) to the APPI Tapis price, reflecting what they perceive as the actual price for Tapis. 145 In 2011 Ma- laysia abandoned APPI Tapis as their reference price and changed to Dated Brent. The price signal of Dated Brent is generated in North Europe and may be influenced by factors other than Asian fundamentals. However, it is seen to have the advantage of lower price volatility than traditional regional Asian markers which struggle with a lack of physical liquidity. Addi- tionally, Dated Brent benefits of its deep financial layers which enable effective risk management by means of hedging. 146

137 Argus Media, ‘No Future for Dated Benchmark’. 138 API gravity 42.7°, sulphur content 0.04% (ExxonMobil, ‘About Tapis’). 139 Also referred to as ‘Sumatran Light’. API gravity 33.94°, sulphur content 0.09% (Chevron, ‘Minas’). 140 Also referred to as ‘Sumatran Heavy’. API gravity 20.29°, sulphur content 0.21% (Chevron, ‘Duri’). 141 Platts, ‘Dated Brent’, 3. 142 Note that in 2013 Platts estimated the production of Tapis at around 150,000 b/d with a spot availability of only 300,000-600,000 barrels per month (Platts, ‘Malaysian Tapis Oil Field EOR Project’). 143 ExxonMobil, ‘About Tapis’; Chevron, ‘Minas’; Chevron, ‘Duri’. 144 Platts, ‘Dated Brent’, 3–4; Energy Intelligence, ‘Minas Touch’. 145 Energy Intelligence, ‘Death Knell’. 146 Energy Intelligence, ‘Petronas Joins Producers Using Brent-Linked Pricing’; Energy Intelligence, ‘Death Knell’.

-27- Rethinking the Asian Marker Debate

We shall now have a closer look at Dated Brent, which is at the heart of the North Sea Brent market and its complicated interplay of physical and financial layers 147 . The Dated Brent market refers to cargoes that have been allocated a specific forward loading date and vessel. This can be exemplified by comparing it to the paper or forward market, the so-called ‘cash BFOE’. A cash BFOE cargo has a designated delivery month but without a vessel or specific date being defined yet. Such a cash BFOE cargo must be ‘dated’, which means that a specific vessel is nominated for delivery, until the 25 th day before the end of its respective delivery month. 148 Dated Brent is commonly referred to as the spot market for Brent, although there is some notion of forwardness to crude oil spot transactions as previously explained. Usually Brent cargoes are sold/bought for delivery at least 10 days ahead. Hence, the prices for Dated Brent are quoted 10-25 days ahead. The price for Dated Brent on the 01 st January, for instance, reflects the price for delivery between the 11 th and 26 th January. The Dated Brent price on the 2 nd January then covers the period between the 12 th and 27 th January, and so on. 149 Platts publishes two different price assessments for Dated Brent: the original London assessment at 4.30 pm Greenwich Mean Time (GMT) as well as an Asian Dated Brent assessment at Asian market close (9.30 am GMT). This is to take into account for the time and thus price difference between the two regions. Hence, Asian producers and customers are able to use a Dated Brent price signal with an Asian timestamp to price the flows of sweet crude in the Asian market. Additionally, Platts also publishes its assessments of Asian sweet crudes as a differential to its Asian Dated Brent assessment for better price comparability. 150 4.2 Candidates Located in Asia Changing market trends have, however,arguably reinforced the issue of an Asian marker located in Asia. The eastward demand shift, the diversification and flexibilisation of Asian supply, an increased trading activity of Asian buyers as well as potential regulatory pressure on existing Western benchmarks pose the question whether an Asian marker located in Asia might be necessary in the near future. The underlying arguments will be presented in detail and scrutinised in section 4.3. The most widely discussed candidates located in Asia are the Russian ESPO Blend named after the East Siberia-Pacific Ocean Pipeline and the announced but yet to be launched Shanghai crude oil futures contract. The first has only become an option as the completion of the pipeline to Kozmino Baycreates a loading point in the Asian basin, whereas the second strives to establish trade at the Shanghai International Energy Ex- change (INE). Hence, these two candidates would enable price discovery taking place in the Asian market itself. 4.2.1 The East Siberia-Pacific Ocean Blend The East Siberia-Pacific Ocean(ESPO) Pipeline can be seen as a manifestation of the eastward trend in Russian oil strategy, as Russia aims to reduce its reliance on western markets and diversify eastwards. 151 This has several reasons such as declining oil output in West Sibe- ria that needs to be balanced, the maintenance of political power in its eastern territories, channelling funds in an underinvested region (and thus supporting the costly development of

147 These other layers are not of direct interest for this paper. For a detailed description of the entire Brent complex see: Fattouh, ‘Anatomy’, 36 ff. 148 Intercontinental Exchange, ‘ICE Brent Crude Oil’. 149 Fattouh, ‘Anatomy’, 45; Intercontinental Exchange, ‘ICE Brent Crude Oil’. 150 Platts, ‘Dated Brent’, 2. 151 Henderson, ‘Strategic Implications’.

-28- Rethinking the Asian Marker Debate

Russia’s Eastern Siberia resources)152 or simply the “[…] belief that thecentre of global economic power is shifting east and that Russia has an opportunity tobecome a leading player in the region […]”153 . It is in this line of thought that the promotion of the ESPO Blend as a new benchmark for the Asian market has to be seen. TheEast Siberia-Pacific Ocean Pipeline (see Figure 1)connects 4,188 kilometres (km) between Taishet and Kozmino on the coast of the Sea of Japan (near Vladivostok).154 The idea for a pipeline in Russia’s Far East was first conceived in the mid-1990s, but the planning process got delayed until mid-2000 because of disputes over the exact route as well as the bankruptcy of the Yukos oil company, which was a major supporter of the project. 155 In the end, the first phase of the pipeline was launched at the end of 2009, connecting 2,757 km between Taishet and Skovorodino and establishing 600,000 b/d of throughput.From Skovo- rodino, near the Chinese boarder, a 64 km branch to Mohe, China was built which, since No- vember 2010, is connected to the 1,000 km distant Daqing, China and transports 300,000 b/d.156 It is through this spur that the Russian state-owned oil company Rosneft will supply China National Petroleum Corporation (CNPC) with 300,000 b/d for 20 years. In turn, Ros- neft received a $25 billion loan from CNPC to help finance the first phase of the ESPO pipeline. In 2013, Rosneft and CNPC agreed to double the crude flows to 600,000 b/d by 2018. Rosneft will receive $270 billion for this 25 year deal, partially in prepayments. Accordingly, the capacity of the pipeline spur Skovorodino-Mohe will be expanded to 400,000 b/d in 2015 and 600,000 b/d in 2018. 157 Note that the start of the first phase of the ESPO pipeline coin- cided with the launch of the oil export terminal in Kozmino Bay, in order to ensure that the diversification strategy behind the ESPO project is fulfilled rather than solely relying on China as a customer. 158 Before the second phase of the ESPO pipeline – reaching from Skovo- rodino to Kozmino – was launched in November 2012 the remaining 300,000 b/d were trans- ported to the Kozmino terminal by rail. With the pipeline connection between Skovorodino and Kozmino the overall capacity of the pipeline to Skovorodino was increased to 1 million b/d, of which 600,000 b/d are directed to the terminal in Kozmino, which also expanded its export capacities 159 (from formerly 300,000 b/d, matching the initial throughput) 160 . The Fi- nancial Times estimates the overall costs for the first two phases of the ESPO pipeline to be $23 billion. Additionally, plans by the Russian oil pipeline monopoly Transneft to extend the capacity of the ESPO pipeline have been approved by the Russian government in February 2014. This third phase shall lift the pipeline’s total capacity to 1.6 million b/d by 2020 and cost approximately $4.9 billion.161

152 Radyuhin, ‘Changing Rules of Energy Game’. 153 Henderson, ‘Strategic Implications’, 8. 154 Radyuhin, ‘Changing Rules of Energy Game’. 155 Konończuk, ‘The ESPO Oil Pipeline’, 1. 156 Radyuhin, ‘Changing Rules of Energy Game’. 157 Gronholt-Pedersen, ‘Factbox’. 158 IHS, ‘Transneft Launches First Leg’. 159 Henderson, ‘Russia’s ESPO Crude’, 25. 160 Henderson, ‘Strategic Implications’, 13. 161 Gorst, ‘Russia – ESPO’.

-29- Rethinking the Asian Marker Debate

Figure 1: East Siberia-Pacific Ocean Pipeline (Phase I/II)

Source: Davis, ‘Crude Pricing’, 30. For the ESPO Blend to become a new benchmark for the Asian basin several conditions have to be fulfilled. Firstly, sufficient crude oil supply has to be available for transport through the ESPO pipeline in the longterm, in order to establish confidence in a sustainable benchmark. Hendersontends to see this condition as fulfilled, as the construction of the ESPO pipeline provides incentives to produce Eastern Siberian fields (as seen e.g. for the fields Vankor, Verkhnechonsk and Talakan). Additionally, the two large fields Yurubcheno-Tokhomskoye and Kuyumba are planned to be linked with the pipeline in 2016 and Rosneft has made several discoveries, which could be added to the pipeline in the future. All in all, this lets Hend- ersonconclude that the production of Eastern Siberia alone could reach 1 million b/d before 2020. Additionally, the ESPO pipeline can be supplied by fields in West Siberia, which have been linked to the ESPOsystem by means of a new pipeline from the Yamal region. Given this additional supply it seems possible that the planned third phase of the ESPO pipeline with a total capacity of 1.6 million b/d can be fully supplied. 162 In 2015, total ESPO flows could reach 1,030,000-1,080,000 b/d (both to Kozmino and through the spur to China). 163 The exact volumes of ESPO Blend delivered are, however, contingent on production increases in Eastern Siberia. In order to support crude production in the Far East the Russian government exempted new fields in Eastern Siberia from export duties in December 2009 (coinciding with the launch of the first phase of the ESPO pipeline). This effectively subsidised eastbound exports on the back of Western exports and applied for a total of 22 fields in East Siberia, of which all are planned to supply the ESPO pipeline at some point in time. In July 2010, the tax exemption was turned into a reduced export duty (relative to westbound exports) and has since then been revised on a monthly basis. 164 In November 2014 the costs of delivering crude through the ESPO pipeline to Kozmino Bay were $2.34 per barrel, according to Argus. This made subsidised exports from Kozmino $26.53 per barrel more profit- able than exporting Urals crude from the Black Sea port Novorossiysk. 165 It is thus not sur- prising that the export duty rates have constantly been a source of dispute between oil pro-

162 Henderson, ‘Russia’s ESPO Crude’. 163 Argus Media, ‘Russia Crude Output Flat’. 164 Platts, ‘Russian Crude Oil Exports to Pacific Basin’, 6–7. 165 Argus Media, ‘Pipelines and Rail Tariffs Rise’.

-30- Rethinking the Asian Marker Debate ducers and Transneft, whose profits are solely generated by these tariffs. 166 Transneft intends to eliminate the subsidies for eastbound exports by 2020 and aims to increase its tariffs for the ESPO pipeline by an average 7.5% in 2015. 167 For the ambition to achieve benchmark status for ESPO Blend the further development of the eastbound export duty rates plays a key role. Hence, producers’ and Transneft’s interest have to be well balanced: Transneft relies on the export duty rates in order to finance further expansions of the ESPO pipeline, whereas producers prefer low rates allowing them to further increase oil production in expensive East- ern Siberian fields. All in all, Argus expects Russian crude production in 2015 to “[…] stabi- lise at best, with the possibility of a drop in output, as sanctions and lower oil prices cap the upstream investment needed to offset field decline.”168 However, available oil supplies for the ESPO pipeline is only one part of the story. For the ESPO Blend to become a marker crude the creation of a liquid spot market in Asia is imperative. In other words, a sufficient amount of ESPO Blend has to be available for spot trading at Kozmino Bay, where price discovery takes place.Initially crude flows to Kozmino were planned to increase from 300,000 b/d in 2010 to 600,000 b/d in 2014. 169 That these volumes were not achieved is, inter alia, due to several diversions of ESPO throughput away from Kozmino. Besides the volumes sent from Skovorodino to China (up from 300,000 b/d to 600,000 b/d by 2018), Rosneft requires 482,000-600,000 b/dfor its planned Far East Petrochemical Complex (FEPCO)170 and additional 160,650 b/d, respectively 120,500 b/d, for the Komsomolsk and Khabarovsk refineries. This ultimately leaves only 118,850-236,850 b/d of ESPO Blend to be traded at Kozmino Bay, once all these diversions fully materialise. 171 It does thus not surprise that loadings at Kozmino Bay only reached 498,000 b/din 2014.172 In 2015, Argusexpects this figure to increase and potentially reach 540,000 b/d. 173 Keeping in mind that 500,000 b/d of production are generally perceived as a condition for a robust benchmark crude, ESPO Blend at Koz- mino Bay should be of sufficient volume. 174 These volumes, however, include term contracts. Energy Intelligence quantifies the share of Kozmino volumes sold under term contracts to be less than 5%. 175 This results in expected Kozmino spot volumes of more than 513,000 b/d in 2015. However, it is expected that even as total shipments to Kozmino Bay are set to rise in the future, spot market volumes at Kozmino are expected to fall due to longterm supply commitments. An example for this is Rosneft’s ambition to supply the planned Tianjin refinery in North China, a joint investment with CNPC, with 182,000 b/d. The crude for the refinery, that is expected to be operational in 2020, will be shipped from Kozmino. 176 But for ESPO Blend to become a benchmark crude the development of a liquid spot market for the price discovery process is crucial – too many longterm contracts and diversions of ESPO

166 Energy Intelligence, ‘Try to Freeze Transneft Tariffs’. 167 Argus Media, ‘Pipelines and Rail Tariffs Rise’. 168 Argus Media, ‘Russia Crude Output Flat’. 169 Gronholt-Pedersen, ‘Factbox’. 170 This plan has been criticized by Transneft and may well be subject to changes. Firstly, Transneft opposes the significant cost burdens FEPCO would impose on Transneft as pipelines have to be expanded. It thus perceives the project as uneconomic. Secondly, Transneft fears that Rosneft will use too much of the ESPO blend for its own purposes, reducing spot market volumes at Kozmino Bay to a point that achieving benchmark status becomes illusory (Energy Intelligence, ‘Tensions Rise Over ESPO Capacity’). 171 Energy Intelligence, ‘Moscow Moves Toward ESPO Benchmark’. 172 Argus Media, ‘Crude Export Outlook’. 173 Argus Media, ‘Moscow Forecasts Lower Output’. 174 Montepeque and Stewart, ‘Sour Crude Pricing’. 175 Energy Intelligence, ‘ESPO Benchmark Blues’. 176 Gronholt-Pedersen, ‘Factbox’.

-31- Rethinking the Asian Marker Debate throughput may subvert this ambition.This situation is further complicated as oil exports through the ESPO spur to China are priced off Kozmino Bay FOB-prices (Free-on-board). 177 This establishes an adverse incentive structure, as it encourages the creation of a supply shortage at Kozmino to create higher revenues from volumes heading to China. This, however, may seriously discourage the evolvement of a liquid spot market at Kozmino Bay. At this point it is also useful to look at the market structure at Kozmino Bay. Although new players entered the market the main suppliers of the blend are still Rosneft and the Russian oil company Surgutneftegaz with disturbingly high market shares of 40%, respectively 33.6%, in 2014. 178 Rosneft was able to expand its position in the market through the acquisition of TNK-BP in March 2013, which initially was one of the key sellers at Kozmino Bay. This, however, ultimately means that Rosneft, and thus potentially the Russian state, controls all pipeline sales to China and more than one third of all exports from Kozmino Bay. This has given rise to concerns about political instrumentalisation of the ESPO Blend and its potential benchmark status for Russia’s strategic ambitions. 179 Furthermore, the concentration of market shares on the buyers’ side also further increased. In 2013, more than half of the blend was bought by Japan’s JX Nippon Oil and Energy (20.6%), Shell (18.6%) and Unipec (17.8%). 180 In terms of destination countries most sales stay in Asia, with Japan, China and South Korea combined accounting for over 70% of all 2014 ESPO Blend sales at Kozmino Bay.181 Cargoes of the blend do also head towards the USA, which was the fourth biggest buyer in 2013. 182 Concerns regarding the potential impact of sanctions against Russia on Japanese buying behaviour (e.g. Japan did not bid for Rosneft’s September tenders), where ultimately unfounded as Japan resumed buying and maintained its market share of about one third of all seaborne ESPO Blend offered in 2014. 183 The large Japanese share is mainly due to two reasons: Firstly, the short delivery times of 3-5 days, compared to about 20 days for crude from the Middle East. This makes price risk management significantly easier and cheaper. Secondly, the convenient cargo size of 100,000 tonnes (so-called ‘Aframax tankers’) which better suits the size limitations of most Japanese ports (very large crude carriers from the Middle East, for instance, have to be reloaded on to smaller tankers). 184 Another issue of concern is the quality of the ESPO Blend. The blend was initially reported to have an API gravity of 34.7° (light) and a sulphur content of 0.6% (sour). 185 Energy Intel- ligence,however, recently reported a sulphur content of only 0.49%, thus categorizing the blend as sweet. Asian buyers, however, still perceive ESPO Blend as sour as its sulphur content is still notably higher than most Asian grades. 186 This poses problems for ESPO’s com- petitive advantage as Asian buyers tend to use either regional light/medium, sweet crudes or Middle Eastern medium, sour crudes. 187 Furthermore, doubts regarding the longterm stability of the blend’s quality pose a challenge for ESPO’s success. As westbound crude exports are

177 Platts, ‘Russian Crude Oil Exports to Pacific Basin’, 2. 178 Argus Media, ‘Asia-Pacific Appetite Grows’. 179 Henderson, ‘Russia’s ESPO Crude’; Shiryaevskaya, ‘Rosneft Completes TNK-BP Acquisition’. 180 Argus Media, ‘Asia-Pacific Appetite Grows’. 181 Argus Media, ‘ESPO Exports Rising’. 182 Argus Media, ‘Asia-Pacific Appetite Grows’. 183 Argus Media, ‘ESPO Exports Rising’. 184 Energy Intelligence, ‘Japan’s Importance Grows’. 185 Reed, ‘Russia’s ESPO Blend Has Bright Future’; Davis, ‘Crude Pricing’, 29. 186 Energy Intelligence, ‘ESPO Benchmark Blues’. 187 Energy Intelligence, ‘ESPO’s Aspirations’.

-32- Rethinking the Asian Marker Debate declining in quality Transneft might be required to transfer more sour crude eastwards as part of the measures to reduce the sulphur content for exports to the West. Furthermore, the introduction of new fields to the ESPO system in 2016 could be associated with changes in quality of the blend.188 Already today the blend’s quality is said to have changed over time as the supplied volumes from fields of different characteristics vary (e.g. light crude from Eastern Siberia or heavier crude from the Vankorskoye field and Western Siberia). 189 Also in terms of pricing, changes will have to be seen before ESPOBlend can become a crude price benchmark. The blend is currently priced as a differential to Dubai. Hence, without ESPOBlend being traded on a flat price basis there is no reason for crude exporters to use it as a reference price rather than Dubai. One way to achieve this would be to trade ESPOBlend on an exchange, given the fact that a functioning futures market is anyway needed for the benchmark status. However, there is currently no paper market for the ESPO Blend and the successful development of a financial market is anything but easy. 190 It has been discussed to launch an ESPO futures contract on the St. Petersburg International Mercantile Exchange or an affiliate in Vladivostok, but no concrete steps have been undertaken for the time being. 191 Moreover, all exports from Kozmino are under control of Russian companies. Al- though less than 5% of Kozmino exports are sold under term contracts (the rest being sold spot without destination restrictions), most of the spot volume is sold through invitation-only tenders. This allows suppliers to effectively control access to the market and raises concerns of Russia using the ESPOBlend as an instrument for its political goals. 192 It is this political risk as well as lack of a financial market around the ESPO Blend that arguably pose the biggest obstacles for the blend to achieve benchmark status. Whether these challenges can be overcome will remain to be seen. For the time being, ESPO Blend – although not yet a reference crude – does play a certain role inthe pricing of exports to Asia. Especially light/medium, sour grades from the Middle East (e.g. Abu Dhabi’s Murban, Umm Shaif or Upper Zakkum) track ESPO’s premium over Dubai while pricing their Asian exports. ESPO’s influence as a pricing guide is, however, limited for crudes priced off Brent, for which the Brent/Dubai EFS is a more important price indicator. 193 4.2.2 The Shanghai Crude Oil Futures Contract China’s crude imports rose by 10.2% year-on-year in the first half of 2014 and reached 6.15 million b/d. As a consequence, there has been a call for China to exert more influence on their crude import prices by means of an own marker crude (the validity of this argument will be discussed in section4.3). To achieve this, a crude oil futures contract shall be launched, which was initially planned to start trading in 2012 on the Shanghai Futures Exchange (SHFE). Additionally, the futures contract is seen as a vehicle to fulfil several purposes of strategic interest to China. Firstly, to support theYuan to become the international reserve and petro-currency. Secondly, to develop China’s financial markets and thirdly, to establish Shanghai as a global financial centre. 194 However, the launch of the contract has been con-

188 Henderson, ‘Russia’s ESPO Crude’. 189 Energy Intelligence, ‘Companies Urged to Support ESPO’; Reed, ‘Russia’s ESPO Blend Has Bright Future’. 190 Energy Intelligence, ‘ESPO Benchmark Blues’. 191 Energy Intelligence, ‘ESPO’s Aspirations’. 192 Energy Intelligence, ‘ESPO Benchmark Blues’. 193 Energy Intelligence, ‘Traders Warn’. 194 Energy Intelligence, ‘Beijing Harbors Big Ambitions’; Energy Intelligence, ‘Yuan Being Positioned to Chal- lenge Dollar’.

-33- Rethinking the Asian Marker Debate stantly delayed and in 2013 the plan for the futures contract was moved to a new exchange, the Shanghai International Energy Exchange (INE). The INE is a joint venture between SHFE and the Shanghai Futures Information Technology Co., a company focusing on development of software required for futures exchanges.INE is located in Shanghai’s new free trade zone where economic liberalization reforms are being piloted. This is supposed to attract international participants by means of tax incentives and promise of full convertibility of the Yuan. 195 After further delays of the trading start, the futures contract received regulatory approval from the China Securities Regulatory Commission early December 2014 and is now said to be launched March 2015. 196 Details on the Shanghai crude oil futures contract are still scarce and prone to changes. How- ever, it seems certain that the contract will settle against physical delivery, with a minimum delivery volume of 200,000 barrels. 197 Eight grades will be deliverable, of which six originate from the Middle East: Basra Light198 (Iraq), Dubai, Masila (Yemen), Oman, Qatar Marine and Upper Zakum. The other two grades will be the Chinese Shengli grade as well as the Russian Urals Blend 199 , in order to avoid overreliance on one supply region. 200 Additionally, the large amount of grades available for physical delivery should help minimize the risk of price manipulation, to which physically deliverable contracts are particularly prone. The INE will use a theoretical benchmark for pricing purposes, rather than one of the eight real crude grades. This means that the delivered crudes will be priced at premiums/discounts to an artificial grade with the following quality properties: 32° API gravity and 1.5% sulphurcontent (medium, sour). 201 For example, Urals (API gravity 31°, sulphur content 1.5%) will be priced at $1.00 discount to the theoretical benchmark. 202 One contract will account for 100 barrels, as opposed to the standard lot size of 1,000 barrels. This should, on the on hand, facilitate trade as it involves smaller risk and requires less financial strength, thus making the contract especially attractive for Chinese retailers as a risk management tool. On the other hand, however, the rather unusual lot size of 100 barrels will also impose additional transaction costs which may hamper trading activity. The minimum price fluctuation will be the standard value of $0.01 per barrel (respectively 0.1 Yuan per barrel). Expiry date will be the last trading day of the month preceding the month of delivery, with the settlement price being an average of the daily closing prices on the last five days of trading. 203 Under which currency the contract will eventually trade remains to be seen, but most likely it will be denominated in Yuan with US-dollar accepted as an alternative settlement currency. International companies without Chinese subsidiaries will be able to open a US-dollar bank account in China, specifically for

195 Energy Intelligence, ‘Shanghai Yet to Deliver on Crude Futures’; Energy Intelligence, ‘Shanghai Pressed to Launch China Futures’; Shanghai Futures Information Technology Co., Ltd., ‘Company’s Briefly Information’. 196 Energy Intelligence, ‘China Approves Crude Futures’. 197 Energy Intelligence, ‘China Moves Crude Futures Contract’. 198 Note that there have been rumours that Basra Light will be removed from the basket of physical deliverable grades due to quality concerns and the recent decision to split Basra Light into two separate, light and heavy, grades (Atanasova, ‘Iraq to Split Basra Light’). 199 According to Platts, Urals was removed from the basket of physically deliverable crudes without substitution, leaving the other seven grades unchanged. Whether this materialises or not remains to be seen (Platts, ‘Shanghai Exchange Eyes Bunker’). 200 Energy Intelligence, ‘Urals to Shanghai Exchange’; Energy Intelligence, ‘Brokers Given Shanghai Crude Futures Demo’. 201 Energy Intelligence, ‘Don’t Mock It’. 202 Energy Intelligence, ‘Urals to Shanghai Exchange’. 203 The daily closing price itself will be a weighted average price of the respective day’s trading volume (Energy Intelligence, ‘Don’t Mock It’).

-34- Rethinking the Asian Marker Debate trading of the futures contract. The trade in US-dollars, however, will be limited to a maximum of $5 billion per day. 204 However, there are several barriers to the success of the contract. For example, the restrictions on crude imports to China allow only a small group of Chinese (mostly state owned) companies to import crude oil to China. In order to cope with this regulations physical delivery of crude against the contract will be made to bonded storage, e.g. at Dalian, Yangshan, Zhousan, or Beihai. Because of this, the crude is not considered to have entered China under tax laws. This firstly avoids Chinese taxes and secondly allows the crude to be re-exported. 205 However, the crude is only able to enter China for refining or resale if it is imported by a company holding a crude import licence. 206 Hence, even though the use of bonded storage will make physical delivery more attractive for international companies, the question still remains whether sufficient liquidity can be attracted without free competition among import- ers. The import licence over 200,000 barrels for the Chinese private company Guanghui En- ergy in 2014 has been perceived as a small step in the right direction. However, it remains to be seen whether the planned crude futures contract will be able to significantly promote the liberalization of crude import quota in the near future. 207 Other regulationsalso might have to be weakened, for instance the regulation of petrochemical product prices which limits hedging strategies and thus may hamper trading volumes in the crude oil futures contract. 208 In this sense, the Shanghai crude oil futures contract is to large extent a bet on economic de- regulation in China. Previous experiences have shown that successfully launching a crude oil futures contract is everything but easy and the ongoing struggle of the DME Oman contract only confirms this. The SHFE has launched a mock trading system early 2014 and has been holding workshops in order to introduce the trading system to market participants. These ses- sions, however, are said to not have been able to attract interest of oil companies despite strong attempts from the SHFE. 209 On the other side, the DME has already communicated interest in cooperating with the Shanghai contract which may help boost liquidity. This is a feasible option given the relatively short time difference between Dubai and Shanghai (4 hours) which creates opportunities for hedging and spread trade. 210 All in all, only time will tell whether the contract will be able to gain sufficient liquidity to be successful and eventually even gain benchmark status. 4.3 Evaluating the Arguments for a Marker Located in Asia The four main arguments for the necessity of an Asian marker located in Asia cover the issues price discovery, price control, price risk management and regulatory risk. They are directly derived from the changing oil market dynamics presented in section 3. Their validity, which is commonly taken for granted in the Asian marker debate, will be examined in order to assess whether it is actually required that the Asian marker is also located in Asia.

204 Energy Intelligence, ‘China Approves Crude Futures’; Energy Intelligence, ‘Shanghai Pressed to Launch China Futures’; Energy Intelligence, ‘Don’t Mock It’. 205 Energy Intelligence, ‘China Wants Its Own Crude Price’; Argus Media, ‘Exchange to Launch Crude Futures’; Energy Intelligence, ‘Don’t Mock It’. 206 Argus Media, ‘Exchange to Launch Crude Futures’. 207 Wong and Aizhu, ‘Update 2’. 208 Energy Intelligence, ‘Beijing Harbors Big Ambitions’. 209 Energy Intelligence, ‘Don’t Mock It’. 210 Energy Intelligence, ‘Shanghai Yet to Deliver on Crude Futures’.

-35- Rethinking the Asian Marker Debate

Firstly, there is the claim that the lack of an Asian marker truly reflecting Asian supply and demand conditions may lead to price anomalies causing confusion in the Asian market. As crude trade flows shift towards Asia and Asian buyers become more actively involved in the trading process and management of arbitrage flows, immediate and accurate price signals for the Asian basin will only be of increasing importance. This criticism has been prominently put forward by Horsnell and Mabro who stated that the Dubai marker does not serve “[…] as an adequate marker against which to price term sales to the Far East.” 211 However, this may have changed over time as the Asian basin has grown and Asian market players have become more actively involved in the trade of crude as Fattouhpoints out.212 More importantly, however, the usage of price differentials allows market players to overcome this local component of the Dubai benchmark price. In this sense, the problem of misleading price signals can be bypassed by adapting the Dubai/ESPO spread to the Dubai price in order to take into account Asian market fundamentals. Nevertheless, Imsirovic has recently used the example of Platts’ Dubai/Oman assessment to reconsider the problem of misleading price signals by means of a case example. In the event of a major supply disruption in the Middle East traders would more actively trade Brent futures in order to hedge their risk. This in turn would lead to price adjustments mainly taking place in the Brent market. To be precise, the Brent/Dubai spread would widen as Brent prices rise due to increased demand, whereas the Dubai price would remain stable (‘sticky’, as called in trader jargon). This behaviour can be explained through the so-called ‘Portfolio Adjustment’ phenomenon, which means that market players with exposure to significant price volatility in their portfolio will trade the most liquid assets in order to hedge this price risk. Given the interlinkage of the Dubai market and the Brentcomplex by means of EFS (Exchange for Swaps, allowing the exchange of Brent futures contracts for Dubai swaps) traders will use the more liquid Brent futures market for risk management purposes. As mentioned above, this would result in the Brent/Dubai spread widening, which is not what one would expect as Middle Eastern crude becomes scarcer and should thus be priced with a premium. The result of these incorrect price signals would be that buyers would try to purchase scarce Middle Eastern crude, which is under-priced, instead of bidding for available substitutes such as Asian sweet crudes, North Sea or West African oil, which are too expensive as they are priced off Dated Brent. According to Imsirovic, this confusion could be avoided by one or several regional benchmarks for Asia reflecting the region’s fundamentals and thus generating immediate and correct price signals. 213 This argument may on the first view be categorized as addressing the price discovery function of crude price benchmarks. Thus, a necessity for an Asian marker located in Asia, reflecting the market’s supply and demand conditions, is concluded. Both the ESPOBlend and the Shanghai futures contract should theoretically be able to fulfil this as their pricing points are located in the Asian market (Kozmino Bay and Shanghai respectively). However, it is not convincing to see this as purely an issue of price discovery. If we look into Imsirovic’s example again,from a risk management perspective, the main problem isin reality lack of financial liquidity around the physical Dubai benchmark itself: Dubai is closely linked to the Brent market, which gives market participants access to the deep liquidity required for hedging and fulfils an important function in calculating the Dubai price if sufficient trades are not available – but financial layers did not evolve around Dubai itself. This, in turn, leads to misleading price signals as

211 Horsnell and Mabro, Oil Markets and Prices , 222. 212 Fattouh, ‘Anatomy’, 65–66. 213 Imsirovic, ‘Asian Oil Markets in Transition’; Imsirovic, ‘Oil Markets in Transition and Dubai’.

-36- Rethinking the Asian Marker Debate traders use the liquid Brent market to hedge the price risk. The correct conclusion would, however, be the need to develop an own financial market around Platts’ Dubai benchmark rather than criticizing its ability to generate correct price signals, which is not the root of the problem. This idea shall be considered in detail in the next chapter. The second argument is linked to the increasing volumes of crude imported into the Asian region and the thereof resulting growing financial burden for Asian buyers. In this sense, an own regional marker for the Asian basin has also been stated to allow Asian buyers to exert more control over prices for their imported crude oil. This argument, which aims at the price discovery function of benchmarks, has been closely tied to the announced Shanghai crude futures contract. It is, however, not clear how such control over prices should be exercised in detail. In addition, it is a central condition for reference crudes to not be politically influenced and whichever candidate might develop as an Asian marker will be under particular observa- tion, especially because of the respective governments backing the ESPOBlend and the Shanghai contract. Apart from that, it is not convincing why a benchmark located in Asia would be required in order to exert some kind of influence over the price generation process. Especially throughout 2014 Asian buyers have increasingly participated in the trading of crude and to some extent have shifted from price takers to price makers, as shown by Imsi- rovicwith the example of the Dubai Platts window. 214 Thirdly, the argument can be made that a financial benchmark launched at an Asian exchange could offer better risk management for Asian buyers. This is of increasing importance as crude oil trade flows shift to Asia and Asian buyers become more actively involved in the trading process, which in turn might drive demand for new risk management tools in the Asian basin. However, the ESPOBlend is far away from developing a liquid financial market to enable effective risk management and the details of the Shanghai futures contract are yet to be announced. Apart from the potential for risk management with the benchmark candidates located in Asia it is not clear why market participants should shift away from the deeply liquid Brent futures market underlying the Dubai marker. There has, for instance, not been a lot of interest in the Oman crude futures contract launched at the DME for risk management purposes – the contract has mainly been used as an instrument for physical delivery. Last but not least, regulatory pressure on Western crude price benchmarks has been building up and could possibly incentivise a shift to more laxly regulated jurisdictions. This would potentially affect Platts’ Dated Brent assessment, but may as well be extended to its Oman/Dubai assessment. 215 Either way this has the potential to significantly influence both the price discovery and risk management service of both Dated Brent and the widely used Dubai benchmark. Even if the planned European regulation of benchmarks does not directly influence the Dubai marker it will most likely indirectly do so through the adverse impact on Brent’s financial liquidity, as Dubai/Oman currently heavily relies on the Brent complex for both price discovery and financial liquidity. In this light, the two candidates presented earlier (ESPO Blend and the Shanghai contract) are proposed as being able to divert liquidity away from more regulated benchmarks. It is, however, not clear why a marker located in Asia would be necessarily required for this. The next section shall introduce a solution utilizing a

214 Imsirovic, ‘Oil Markets in Transition and Dubai’. 215 ‘Price Reporting Probe’.

-37- Rethinking the Asian Marker Debate crude price benchmark outside of Asia, for which the DME Oman futures contract may play a crucial role as it is not influenced by intensified regulatory pressure in the Western world. All in all, the four presented arguments do not confirm the necessity of an Asian marker originating from Asia. This is because all underlying issues can also be solved by means of a reference crude originating from outside of Asia. Additionally, the inertia of the benchmarking system poses a barrier to a shift from existing markers outside of Asia.The system of crude price benchmarks is inherently inertial, because of the financial liquidity required for adequate risk management. Due to the positive feedback mechanism discussed in section 2.1.2 we see a consolidation around a few, highly liquid financial instruments (e.g. WTI or Brent futures contracts). Because of this any new benchmark evolving will have to offer sufficient added value in terms of either price discovery and/or risk management in order to incentivise market participants to shift away from the established markers. In case of the Asian marker the predominant position of the Brent complex would have to be broken through, as it is its liquidity that maintains the Platts’ Dubai benchmark as the dominant marker for the Asian basin. The difficulty of establishing a new marker can be exemplified by the DME Oman futures contract which is still struggling to attract sufficient liquidity and also the now heavily traded ICE Brent futures contract which was not successful at the first go. Any Asian marker located in Asia faces the same challenges. In this sense, the evolution of new benchmarks does not take place in a vacuum. As existing benchmarks outside of Asia can also cope with these issues the need for an own Asian marker is not only proven wrong, but the inertia of the crude price benchmarking system makes the adoption of an Asian reference crude located in Asia rather unlikely.

5 An Alternative: Rebuilding the Middle Eastern Marker After having shown that the problems derived from changing dynamics in oil markets can also be dealt with a marker located outside of Asia, I will now support this by means of a promising example: a link between Platts’ Dubai marker and the DME Oman futures contract. This example is especially appealing as it involves the predominant Asian marker (Du- bai), thus making it easier to overcome the previously mentioned inertia of the benchmarking system. The idea of such an interplay has been mentioned in passing in a recent paper written by Imsirovic. 216 Imsirovic asks whether “[…] perhaps DME can do for Dubai what the Inter- national Petroleum Exchange (IPE) did for Brent?” 217 , without going into details what this would mean in particular. Hence, I shall firstly substantiate this idea, before secondly illus- trating howmany of the current issues around both Platts’ Dubai assessment and DME’s Oman futures contract could be solved in this way. Subsequently, it will be shown how this could deal with the previously mentioned arguments supposedly making an Asian marker located in Asia required. 5.1 The Linkage In order to better understand what such a link means in detail it is useful to look at the example of IPE and Brent, as mentioned by Imsirovic. 218 The third attempt to launch theIPE Brent futures contract was successful on the 23 rd June 1998, establishing both a financial instrument for price risk management as well as an additional price signal complementing that of the

216 Imsirovic, ‘Oil Markets in Transition and Dubai’, 6. 217 Ibid. 218 Ibid.

-38- Rethinking the Asian Marker Debate physical Brent (forward) market. 219 Note that in the following solely the notation ‘ICE Brent futures contract’ will be utilized for reasons of simplicity, as the IPE was acquired by the ICE in 2001. From the price risk management perspective the rationale to launch a futures contract for Brent is straightforward. As shown in section 2.1.2 the basis risk ∆PX-∆PH can be reorganised as ∆PR-∆PH+∆D. This means that the more the price of the reference crude (Dated Brent) and the price of the hedging instrument (ICE Brent futures contract) move in lockstep, the smaller the resulting basis risk is. In this sense, ICE Brent futures contracts are the salient choice to hedge Dated Brent price exposure and the deep liquidity of the financial layers around Brent has only reinforced this and further stabilised Dated Brent’s status as a global benchmark. Regarding price discovery, the ICE Brent futures contract generates an additional price signal formed by different market fundamentals than the physical Brent forward or spot market. This has even lead to the development of a separate financial benchmark (BWAVE, an index based on ICE Brent futures), used by Saudi Arabia, Kuwait and Iran to price their crude exports to Europe. The linkage between this price signal generated on the ICE and the physical Brent (forward) market works in both directions: Firstly, the ICE Brent futures contract fi- nancially settles against the price of the ICE Brent Index on the day following the last trading day of the futures contract. This index is based on the weighted average of first-month and second month trades in the 25-day cash BFOE market, taking into account the average of spread trades between the first and second month. Hence, the ICE Brent futures price con- verges to a forward price at expiry, not to a spot price as is usually the case. Additionally, the futures market is linked to the physical Brent forward market by means of an Exchange of Futures for Physicals (EFP) mechanism. This allows traders to convert a futures position into forward Brent (25-day cash BFOE), which will then eventually convert to Dated Brent (spot market). 220 Secondly, price reporting agencies often use information from ICE Brent futures contracts to calculate the forward Brent price (e.g. if trading liquidity is thin). This is done by means of the EFP-mechanism: As EFPs are usually priced as a differential to Brent futures prices, one can derive the forward Brent price for a given month by adding the EFP for the respective month to the Brent futures price of this month. 221 Thus one can firstly conclude, that because of the bidirectional link between Brent futures market and the physical forward market there is a notion of circularity involved in the price discovery process: Price reporting agencies may use information from the futures market to derive the Brent forward price – against which, in turn, the ICE Brent futures contract settles. Secondly, the link between ICE Brent futures and forward Brent market in terms of price discovery exemplifies that not only the physical Brent market has turned into a brand namebut also the ICE Brent futures contract is not exclusively linked to the Brentgrade but to the entire BFOE basket. When transferring this to the idea of linkingPlatts’ Dubai assessment and DME’s Oman futures contract, the linkage is established by the same two components: the possibility of hedging the physical Dubai benchmark with the DME Oman futures contract and generating an additional price signal for the Dubai benchmark on the DME. This is possible because the Dubai benchmark, despite its name, is a brand consisting of the grades Dubai, Oman and Up- per Zakum – of which, in practice,Omani crude dominates physical delivery. According to

219 Horsnell and Mabro, Oil Markets and Prices , 193. 220 Fattouh, ‘Anatomy’, 43–44; Intercontinental Exchange, ‘ICE Brent Crude Oil’. 221 Fattouh, ‘Anatomy’, 50.

-39- Rethinking the Asian Marker Debate the DME, “[a]s of mid-March 2014, all but one of the cargo deliveries into the Dubai pricing mechanism in 2014 have been Oman, representing more than 95% of deliveries.” 222 This is a relatively short period of time giving us a rather extreme distribution, but Imsirovic shows that between 2012 and July 2014 Oman and Upper Zakum have been the grades dominating physical delivery in the Platts Dubai window (Dubai crude was rarely delivered). 223 In other words, Omani crude is the link between the physical Dubai market and the financial Oman market. This is exemplified by the potential for arbitrage taking place between the two markets: Depending on the price differentials, physical Omani crude can be bought over the physical delivery mechanism of the DME Oman futures contract and then profitably sold in the Dubai Platts window under the alternative delivery mechanism (vice versa). 224 In terms of the first component, price risk management, the DME Oman futures contract should involve less basis risk for hedging exposure to Dubai prices than the currently common used ICE Brent futures. 225 This is due to the fact that the market fundamentals influencing the DME Oman futures price should be closer to those impacting the Dubai price than the market conditions of ICE Brent futures. Additionally, the fact that many Middle Eastern crude exports are priced off of an average of Platts Dubai and Oman price assessments should encourage the usage of the DME contract as a hedging instrument for these transactions (dis- regarding liquidity considerations). This is because Oman is to 50% part of the Dubai/Oman benchmark which should further increase the correlation of the reference price (P R) and the DME Oman futures price (P H). This, in turn, decreases the basis risk ∆PR-∆PH+∆D, in comparison to hedging solely Dubai price exposure. The second component of the link, price discovery, should be bidirectional as above- mentioned for ICE and forward Brent. This means that the Oman price signal generated on the DME would both origin from the Platts Dubai window as well as occasionally feed back into the physical Dubai market (entailing Dubai, Oman and Upper Zakum).Firstly, the DME Oman futures contract could, for instance, converge to a price signal generated on the physical Dubai market by means of an index (cf. to above-mentioned ICE Brent index). Secondly, the DME Oman price could be used to derive the Dubai price in case of insufficient concluded deals and bid/offer information in the Platts Dubai window. One possible method would be utilising the EFP-mechanism of the DME, which if added to the price of the Oman futures contract for a given month would yield the Oman forward price for the respective month. Then the Dubai forward price for this month could be derived using information from Dubai/Oman swaps. 5.2 Fixing Dubai The above-described linkage would be an improvement for the Dubai benchmark both in terms of price discovery and price risk management relative to the current situation. Cur- rently, Platts’ Dubai marker heavily relies on the financial layers around the Brent complex for both of these services. This has led to criticism of the Dubai marker, as it does not inde- pendently fulfil both functions identified as crucial for crude price benchmarks.Fattouh has shown that since 1989 most trading of Dubai has taken place as spread deals (i.e. mainly as Brent/Dubai EFS or Dubai inter-month swaps) instead of deals with an outright price for Du-

222 Dubai Mercantile Exchange, ‘Brent Oman Spread’, 2. 223 Imsirovic, ‘Oil Markets in Transition and Dubai’, 6. 224 Ibid. 225 Ibid., 9.

-40- Rethinking the Asian Marker Debate bai. In 1991, around 95% of all Dubai transactions where spread deals of which Brent/Dubai EFS constituted the biggest share. In fact, nowadays Dubai only trades on a fixed price basis in the Platts window. 226 This is also what has led Horsnell and Mabro to conclude that “Dubai has become close to being little more than another Brent add-on market” 227 , as ultimately the liquidity of the Dubai market mainly stems from the Brent complex. In this sense, Fattouh concludes that it is these financial layers linking the Brent and Dubai market that allow the Dubai benchmark to survive, even though its physical basis is thin and trading in the Platts window tends to be low. 228 Imsirovic recently argued along the same lines, stating that “[…] Dubai is not a real oil price Marker or benchmark, as its price is actually derived from Brent futures price.”229 By capping the link to the Brent complex and establishing a connection between Dubai and the DME Oman contract this criticism could be solved. In terms of price risk management Dubai would not have to rely on ICE Brent futures anymore but on the DME Oman contract, which would involve lower basis risk. Regarding price discovery in case of an illiquid Platts window, the Brent/Dubai EFS mechanism would be substituted by utilising a price signal generated on the DME. Additionally, the Platts window is often criticized as it allows a small group of participants to act as price makers. In case of the Dubai window, for instance, Shell and Vitol account for about half of all trades concluded in 2013. The Brent window is equally dominated by a small group of traders, but benefits from the additional price signal generated on the ICE. In this sense, the Platts Dubai window is more prone to this criticism as it currently lacks its own financial layers acting as an additional source of price discovery. This could be solved by linking the Dubai benchmark with the DME. 230 5.3 Fixing DME Oman Firstly, linking the DME Oman futures contract to the physical Dubai market would give the contract a physical foundation it currently lacks. At the moment financial settlement takes place against the average of daily settlement prices over the respective month, which would be substituted by financial settlement prices derived from the physical Dubai market. Sec- ondly, and more importantly, the DME and the Oman futures contract would benefit from the increasing use of the contract as a hedging instrument for Dubai price exposure, thus further attracting liquidity. In this context it is important to note that although the DME officially promotes a change in benchmarking from Platts’ Dubai assessment to its Oman contract, it should be in the interest of the exchange to help Dubai maintain its benchmark status if this involves increased trading of the DMEOman futures contract. This is because, in the end, trade volumes and liquidity are the important metrics for an exchange. Whether this involves reaching benchmark status for a specific contract is not the key issue from the exchange’s point of view. In this sense, there is a difference between successfully launching a futures contract – which means attracting sufficient trade volumes and liquidity – and this futures contract becoming a price benchmark. Accordingly, ICE would be the victim of such a link between DME and Dubai, as it currently benefits from the liquidity stemming from the EFS- mechanism through which the price of Dubai is often generated as well as the fact that Brent

226 Fattouh, ‘Anatomy’, 62–63. 227 Horsnell and Mabro, Oil Markets and Prices , 212–213. 228 Fattouh, ‘Dubai Benchmark and Its Role’, 4–5. 229 Imsirovic, ‘Asian Oil Markets in Transition’. 230 Imsirovic, ‘Oil Markets in Transition and Dubai’, 8.

-41- Rethinking the Asian Marker Debate futures are often used to hedge exposure to Dubai flat price risk. This liquidity would be cut if Dubai would rely on the DME Oman contract for price discovery and risk management, thus shifting liquidity towards the DME.Additionally, if liquidity increases on the DME there might be the possibility for a financial benchmark to develop based on the Oman futures contract, as has happened with ICE Brent futures and BWAVE. 5.4 Dealing with Shifting Market Trends I shall now show how the suggested link would be able to deal with the previously mentioned arguments for an Asian marker located in Asia. The four arguments, derived from changing market dynamics, that supposedly make this required were presented in section 4.2: price discovery, price control, price risk management and regulatory risk. Firstly, Imsirovic’s cri- tique of the price signal of the Platts Dubai benchmark fundamentally stems from a lack of financial liquidity around the physical benchmark Dubai. 231 This would be solved by linking the DME Oman futures contract to the physical Dubai market, enabling market players to hedge their price exposure with DME Oman contracts. This would lead to an accurate price signal: an increase in the price of Omani crude (relative to, for instance, North Sea or West African oil), which would then in turn lead to a higher Dubai price both through its influence in the price setting of Dubai as well as arbitrage between the Dubai partials market and the DME Oman market. Secondly,the price control argument is not very convincing. But even if we grant it some form of validity, establishing an own financial market around the physical Dubai benchmark by means of the DME Oman contract would offer market players a distinct platform to exert influence on Dubai prices. Thirdly, and as previously argued, hedging Du- bai price exposure with the DME Oman contract would involve less basis risk compared to Brent futures and should, in fact, be the salient hedging instrument for this purpose. Last but not least, regarding potential regulation of Western benchmarks (e.g. WTI and Brent) the linkage between the Dubai market and the DME Oman contract would be beneficial because this would effectively cap the existing link to the Brent complex. Both in terms of price discovery and risk management the physical Dubai benchmark would become independent of the Brent market and thus unaffected of regulatory risk potentially weakening the status of Dated Brent as a benchmark.

6 Outlook and Conclusions But what could facilitate the development of the above-described DME Oman/Dubai complex? The key factor for this link to be established is – once again – financial liquidity. It is the vast liquidity of the financial layers around Brent that incentivises market players to manage their Dubai price risk via the Brent complex. Accordingly, regulatory risk could play a crucial role in diverting liquidity away from the financial markets around Brent. In this case market players would have to look for alternative marker crudes as well as financial markets to hedge their price exposure. It is this gap that could be filled by the proposed DME Oman/Dubai complex. This is especially appealing as it is precisely the regulatory risk, which puts pressure on the Brent/Dubai complex,whichcould pavethe way for the DME Oman/Dubai linkage to develop. In other words, the demise of onebenchmarking complex is the beginning of the other, thus enabling a smooth transition. Moreover, linking the DME Oman to Dubai should be the salient choice for a restructuring of the benchmark system because of two reasons. Firstly, it offers extensive benefits both regarding price discovery and

231 Imsirovic, ‘Oil Markets in Transition and Dubai’.

-42- Rethinking the Asian Marker Debate price risk management. Secondly, and most likely more importantly, it allows to continue using the Dubai marker for the Asian basin, thus overcoming the inertia of the benchmarking system. Hence, maintaining the Asian marker outside of Asia seems to be the most feasible and frictionless option in case of major structural changes due to regulatory risk. All in all, we can conclude that there is no convincing case for an Asian marker located in Asia. There are serious drawbacks to each of the four arguments derived from changing oil market dynamics. Moreover, the DME Oman/Dubai complex constitutes an innovative proposal for an Asian marker located outside of Asia. The suggested DME Oman/Dubai link would not onlybe able to cope with the issues brought upon by eastward shifting oil markets. It would also solve many of the individual problems that the Dubai benchmark and the DME Oman futures contract are currently facing, arguably making them less resistant. Already in 2012, Fattouh pointed out that “[…] structural transformations could occur, and if this hap- pens, Dubai is likely to be the least immune to radical changes in the international pricing system.” 232 Accordingly, the DME Oman/Dubai complex is also a proposal to strengthen the benchmark status of Gulf crudes: It is by this linkage that the Middle Eastern benchmarking system can be prepared for the future challenges posed by an increasingly eastward shifting balance of power in international oil markets.

232 Fattouh, ‘Dubai Benchmark and Its Role’, 10.

-43- Rethinking the Asian Marker Debate

References Argus Media. ‘Asia-Pacific Appetite Grows’. Argus FSU Energy , 16 January 2014. ———. ‘Crude Export Outlook Hard to Call’. Argus FSU Energy , 15 January 2015. ———. ‘Energy Price Reporting Organisations Unveil Draft Self-Regulatory Code’. Press Releases , 30 April 2012. http://www.argusmedia.com/About-Argus/Press- Releases/2012/IPRO-Code. ———. ‘ESPO Exports Rising’. Argus Nefte Transport , 19 December 2014. ———. ‘Exchange to Launch Crude Futures Contract’. Argus China Petroleum , 5 December 2013. ———. ‘Methodology and Specifications Guide: Argus Crude’. Argus Media, February 2015. http://www.argusmedia.com/~/media/Files/PDFs/Meth/argus_crude.pdf?la=en. ———. ‘Moscow Forecasts Lower Output This Year’. Argus Global Markets , 16 January 2015. ———. ‘No Future for Dated Benchmark’. Argus Global Markets , 2 May 2014. ———. ‘Out Dated’. Argus Global Markets , 26 September 2014. ———. ‘Outlook: Russia Crude Output Flat, Exports Rising’, 22 December 2014. ———. ‘Pipelines and Rail Tariffs Rise’. Argus Nefte Transport , 19 December 2014. ———. ‘Race to the Bottom’. Argus Global Markets , 12 September 2014. ———. ‘Shell Buying Supports Dubai’. Argus Global Markets , 6 December 2013. ‘Asian Buyers Influence Key Oil Benchmarks’. Middle East Economic Survey 57, no. 09 (28 February 2014): 14–15. Atanasova, Slavka. ‘Iraq to Split Basra Light Crude into Two Grades’. ArabianOiland- Gas.com , 13 November 2014. http://www.arabianoilandgas.com/article-13160-iraq- to-split-basra-light-crude-into-two-grades/. Binks, Adrian. ‘Middle East Crude Pricing: The Dubai Debate’. Middle East Economic Sur- vey 48, no. 07 (14 February 2005): 27–30. Blas, Javier. ‘Rise of the Oman Crude Benchmark’. Financial Times . 9 May 2012. http://www.ft.com/cms/s/0/fcb484e4-99a9-11e1-8fce-00144feabdc0.html. Burkhard, James, Geoff Houlton, and Ha Nguyen. ‘America’s Great Oil Revival: Reshaping the Global Market’. Private Report. Cambridge, Massachusetts: IHS CERA, 2013. Chevron. ‘Duri’. Chevron Crude Oil Marketing . Accessed 6 February 2015. http://crudemarketing.chevron.com/crude/far_eastern/duri.aspx. ———. ‘Minas’. Chevron Crude Oil Marketing . Accessed 6 February 2015. http://crudemarketing.chevron.com/crude/far_eastern/minas.aspx. Davis, Mike. ‘Crude Pricing for the New Oil Fundamentals: Implications & Opportunities’. May 2013. http://www.platts.com/im.platts.content/productsservices/conferenceandevents/2013/p c366/presentations/mike_davis_ice_future_enterprises.pdf. Deloitte. ‘Oil and Gas Reality Check 2014: A Look at the Top Issues Facing the Oil and Gas Sector’. London: Deloitte, 2014. http://www2.deloitte.com/content/dam/Deloitte/global/Documents/Energy-and- Resources/gx-er-oilandgas-reality-check2014.pdf. Dipaola, Anthony. ‘CME Will Boost Dubai Oil Bourse Investment With Oman Fund’. Bloomberg , 21 February 2012. http://www.bloomberg.com/news/articles/2012-02- 21/cme-oman-fund-to-increase-stakes-in-dubai-mercantile-exchange. Dubai Mercantile Exchange. ‘Brent Oman Spread’. Futures Insights. Dubai: DME, March 2014. http://www.dubaimerc.com/pdf/fi/Futures%20Insights%20-%20Issue%202.pdf.

-44- Rethinking the Asian Marker Debate

———. ‘Historical Data - OQD’. Historical DME Settlement Price . Accessed 5 February 2015. http://www.dubaimerc.com/dmedata/historical-oqd.aspx. ———. ‘Oman Crude Oil & DME’. Futures Insights. Dubai: DME. Accessed 5 February 2015. http://www.dubaimerc.com/pdf/fi/Futures%20Insights%20-%20Issue%201.pdf. ‘Dubai to Base OSP on DME Oman Futures Contract’. Middle East Economic Survey 52, no. 26 (29 June 2009): 11. Energy Intelligence. ‘Asia Moves From Price Taker to Price Maker’. Petroleum Intelligence Weekly , 19 May 2014. ———. ‘Boosting Benchmarks’. Petroleum Intelligence Weekly , 24 February 2014. ———. ‘Brokers Given Shanghai Crude Futures Demo’. International Oil Daily , 9 January 2013. ———. ‘China Approves Crude Futures Contract’. International Oil Daily , 16 December 2014. ———. ‘China: Beijing Harbors Big Ambitions for Crude Futures Contract’. Energy Com- pass , 27 April 2012. ———. ‘China: Key Pillar of Global Oil Prices Weakens’. Energy Compass , 28 November 2014. ———. ‘China Moves Crude Futures Contract Into Shanghai Free Trade Zone’. Interna- tional Oil Daily , 18 November 2013. ———. ‘China Wants Its Own Crude Price’. EI Finance , 25 April 2012. ———. ‘Companies Urged to Support ESPO as Benchmark’. Nefte Compass , 26 June 2014. ———. ‘DME Hopeful of Futures Contract Breakthrough’. Petroleum Intelligence Weekly , 2 April 2012. ———. ‘DME Introduces Swaps, But Still Needs Saudis’. EI Finance , 15 December 2010. ———. ‘Dubai Exchange Names New CEO’. International Oil Daily , 22 June 2012. ———. ‘ESPO’s Aspirations’. Petroleum Intelligence Weekly , 19 November 2012. ———. ‘EU Brings Oil Pricing Into Libor Regulation Efforts’. International Oil Daily , 26 July 2012. ———. ‘EU to Bring Out Benchmark Legislation’. Oil Daily , 7 June 2013. ———. ‘IOSCO: PRAs Have Made “Good Progress”.’ International Oil Daily , 10 Septem- ber 2014. ———. ‘Japan’s Importance Grows in Major ESPO Crude Market’. International Oil Daily , 31 December 2013. ———. ‘Market Comment: Don’t Mock It: Fantasy Crude Trading Gets Under Way in Shanghai’. EI Finance , 23 January 2013. ———. ‘Market Comment: Shanghai Yet to Deliver on Crude Futures Contract’. EI Finance , 16 July 2014. ———. ‘Market Forces: Battle of the Asian Crude Benchmarks’. Energy Compass , 23 No- vember 2012. ———. ‘Market Forces: Death Knell’. Energy Compass , 15 April 2011. ———. ‘Market Forces: Dubai Looks for Liquidity’. Energy Compass , 28 February 2014. ———. ‘Market Forces: ESPO Benchmark Blues’. Energy Compass , 28 May 2014. ———. ‘Market Forces: Oman’s DME Looks East in Search for Credibility’. Energy Com- pass , 19 April 2013. ———. ‘Market Forces: Oman’s Physical Limits’. Energy Compass , 16 December 2011. ———. ‘Market Forces: US Financial Reform: A Guide’. Energy Compass , 23 July 2010. ———. ‘Market Forces: Will Tougher Rules Spur New Benchmarks?’. Energy Compass , 29 October 2010. ———. ‘Mideast Rethinks Marketing as Oil Flows Shift East’. Petroleum Intelligence Weekly , 13 October 2014.

-45- Rethinking the Asian Marker Debate

———. ‘Minas Touch’. Petroleum Intelligence Weekly , 21 May 2012. ———. ‘Mitsubishi Helps DME Advance Asia Strategy’. International Oil Daily , 14 June 2013. ———. ‘Moscow Moves Toward ESPO Crude Benchmark’. Nefte Compass , 98 2013. ———. ‘Oil Firms Try to Freeze Transneft Tariffs’. International Oil Daily , 7 November 2014. ———. ‘Perspective: Chinese Yuan Being Positioned to Challenge Dollar’. Energy Com- pass , 10 January 2014. ———. ‘Petronas Joins Ranks of Producers Using Brent-Linked Pricing’. International Oil Daily , 5 April 2011. ———. ‘Price Reporting Agencies Won’t Be Regulated’. International Oil Daily , 8 October 2012. ———. ‘Shanghai Pressed to Launch China Futures Contract’. Petroleum Intelligence Weekly , 23 June 2014. ———. ‘Tensions Rise Over ESPO Capacity’. Nefte Compass , 13 June 2013. ———. ‘Traders Warn ESPO Far From Real Benchmark’. Nefte Compass , 27 March 2014. ———. ‘Urals Makes It to Shanghai Exchange’. Nefte Compass , 24 January 2013. ———. ‘Will Shrinking North Sea Output Diminish Brent?’. Petroleum Intelligence Weekly , 7 April 2014. European Commission. ‘Citizens’ Summary: EU Financial Benchmarks Proposal’. European Commission, 18 September 2013. http://ec.europa.eu/finance/securities/docs/benchmarks/130918_citizens- summary_en.pdf. ExxonMobil. ‘About Tapis’. Crude Oils . Accessed 6 February 2015. http://www.exxonmobil.com/crudeoil/about_crudes_tapis.aspx. ———. ‘About Upper Zakum’. Crude Oils . Accessed 5 February 2015. http://www.exxonmobil.com/crudeoil/about_crudes_upperzakum.aspx. Fattouh, Bassam. ‘An Anatomy of the Crude Oil Pricing System’. WPM 40. Oxford: Oxford Institute for Energy Studies, January 2011. ———. ‘Prospects of the DME Oman Crude Oil Futures Contract’. Oxford Energy Com- ment. Oxford: Oxford Institute for Energy Studies, March 2008. ———. ‘The Dubai Benchmark and Its Role in the International Oil Pricing System’. Oxford Energy Comment. Oxford: Oxford Institute for Energy Studies, March 2012. ———. ‘The Origins and Evolution of the Current International Oil Pricing Regime: A Criti- cal Assessment’. In Oil in the Twenty-First Century , edited by Robert Mabro, 41–100. Oxford University Press, 2006. Fattouh, Bassam, Lutz Kilian, and Lavan Mahadeva. ‘The Role of Speculation in Oil Mar- kets: What Have We Learned So Far?’. WPM 45. Oxford: Oxford Institute for Energy Studies, March 2012. Fattouh, Bassam, and Lavan Mahadeva. ‘Assessing the Financialization Hypothesis’. WPM 49. Oxford: Oxford Institute for Energy Studies, November 2012. ———. ‘OPEC: What Difference Has It Made?’. MEP 3. Oxford: Oxford Institute for En- ergy Studies, January 2013. http://www.oxfordenergy.org/wpcms/wp- content/uploads/2013/01/MEP-3.pdf. Fattouh, Bassam, and Pasquale Scaramozzino. ‘Uncertainty, Expectations, and Fundamen- tals: Whatever Happened to Long-Term Oil Prices?’. Oxford Energy Comment. Ox- ford: Oxford Institute for Energy Studies, March 2011. Fattouh, Bassam, and Amrita Sen. ‘New Swings for West African Crudes’. Oxford Energy Comment. Oxford: Oxford Institute for Energy Studies, August 2014.

-46- Rethinking the Asian Marker Debate

Fix, Christopher. Setting a Benchmark. Interview by Akshay Bhatnagar. Oman Economic Review, April 2014. http://www.oeronline.com/php/2014/Apirl/trade_finance.php. Gorst, Isabel. ‘Russia – ESPO: Asia’s Gain, Europe’s Pain’. Financial Times , 19 February 2014. blogs.ft.com/beyond-brics/2014/02/19/russia-espo-asias-gain-eu. Gronholt-Pedersen, Jacob. ‘REFILE-FACTBOX-Russia-China Crude Oil Deals’. Reuters . 19 February 2014. http://uk.reuters.com/assets/print?aid=UKL3N0LI3K720140219. Henderson, James. ‘Russia’s ESPO Crude: A New Benchmark for Asia?’. Oxford Energy Forum , no. 94 (November 2013): 24–27. ———. ‘The Strategic Implications of Russia’s Eastern Oil Resources’. WPM 41. Oxford: Oxford Institute for Energy Studies, January 2011. Horsnell, Paul, and Robert Mabro. Oil Markets and Prices: The Brent Market and the For- mation of World Oil Prices . Oxford: Oxford University Press, 1993. IHS. ‘Dubai Mercantile Exchange Begins Oil Futures Trading Today as New Benchmark Contract Is Sought’. Same-Day Analysis , 31 May 2007. https://www.ihs.com/country- industry-forecasting.html?ID=106598030. ———. ‘Transneft Formally Launches First Leg of ESPO Pipeline’. Same-Day Analysis , 29 December 2009. https://www.ihs.com/country-industry- forecasting.html?ID=106594735. Imsirovic, Adi. ‘Asian Oil Markets in Transition’. Journal of Energy Security , 24 April 2014. http://ensec.org/index.php?option=com_content&view=article&id=520:asian-oil- markets-in-transition&catid=143:issue-content&Itemid=435. ———. ‘Oil Markets in Transition and the Dubai Crude Oil Benchmark’. Oxford Energy Comment. Oxford: Oxford Institute for Energy Studies, 2014. Imsirovic, Adi, and Tilak Doshi. ‘Exploring the Asian Premium in Crude Oil Markets’. Jour- nal of Energy Security , 21 November 2012. http://ensec.org/index.php?option=com_content&view=article&id=393:exploring-theasian-premium-in-crude-oil-markets&catid=130:issue-content&Itemid=405. Intercontinental Exchange. ‘ICE Brent Crude Oil: Frequently-Asked Questions’. ICE, Janu- ary 2013. https://www.theice.com/publicdocs/futures/ICE_Brent_FAQ.pdf. International Energy Agency. ‘Oil Medium-Term Market Report 2014: Market Analysis and Forecasts to 2019’. Paris: OECD/IEA, 2014. ———. ‘World Energy Outlook 2014’. Paris: OECD/IEA, 2014. Kono ńczuk, Wojciech. ‘The East Siberia/Pacific Ocean (ESPO) Oil Pipeline: A Strategic Project – an Organisational Failure?’. Centre for Eastern Studies Commentary , no. 12 (22 October 2008). Mabro, Robert. ‘On Oil Price Concepts’. WPM 3. Oxford: Oxford Institute for Energy Stud- ies, 1984. ———. ‘The International Oil Price Regime: Origins, Rationale and Assessment’. The Jour- nal of Energy Literature XI, no. 1 (2005): 3–20. Montepeque, Jorge, and Peter Stewart. ‘Sour Crude Pricing: A Pressing Global Issue’. Middle East Economic Survey 48, no. 14 (4 April 2005). Pirrong, Craig. ‘The Economics of Commodity Trading Firms’. Trafigura White Paper, 2014. http://www.trafigura.com/media/1364/economics-commodity-trading-firms.pdf. Platts. ‘Dated Brent: The Pricing Benchmark for Asia-Pacific Sweet Crude Oil’. Market Is- sues: Oil, 2011. http://www.platts.com/IM.Platts.Content/InsightAnalysis/IndustrySolutionPapers/olm i003_datedbrent.pdf. ———. ‘Dubai Mercantile Exchange 2014 Traded Volume Surged 33% from 2013’. Latest News Headlines , 12 January 2015. http://www.platts.com/latest- news/oil/singapore/dubai-mercantile-exchange-2014-traded-volume-26980073.

-47- Rethinking the Asian Marker Debate

———. ‘Interview: CME Sees Push into Europe, Asia Energy Markets as Gaining Momen- tum’. Platts Commodity News , 26 March 2014. ———. ‘Japan’s Tocom, Dubai’s DME Ink Deal to Beef up Cooperation on Energy Con- tracts’. Platts Commodity News , 16 March 2014. ———. ‘Malaysian Tapis Oil Field EOR Project Set to Start up End 2013: ExxonMobil’. Latest News Headlines , 10 July 2013. http://www.platts.com/latest- news/oil/singapore/malaysian-tapis-oil-field-eor-project-set-to-27172899. ———. ‘Review of the Partial Dubai and Oman Price Discovery Mechanism’. Discussion Paper. Market Issues, November 2004. http://www.platts.com/elqNow/elqNotAuthRedir.htm?ref=/IM.Platts.Content/Insight Analysis/MarketIssues/discussion_partials.pdf. ———. ‘Russian Crude Oil Exports to the Pacific Basin – an ESPO Update’. Special Report, February 2011. http://www.platts.com/im.platts.content/insightanalysis/industrysolutionpapers/espo0 211.pdf. ———. ‘Shanghai Exchange Eyes Bunker Contract before Oil’. Platts Oilgram News , n.d., 22.11.2013 edition. ———. ‘The Dubai Crude Oil Price’. Dubai Crude Price Assessment Explained . Accessed 5 February 2015. http://www.platts.com/price-assessments/oil/dubai-crude. ———. ‘The Structure of Global Oil Markets’. Backgrounder, 2010. http://www.platts.com/IM.Platts.Content/InsightAnalysis/IndustrySolutionPapers/oil markets.pdf. ‘Price Reporting Probe Puts Oman/Dubai Assessments Under Scrutiny’. Middle East Eco- nomic Survey 56, no. 21 (24 May 2013): 4. Radyuhin, Vladimir. ‘Changing the Rules of the Energy Game’. The Hindu , 12 January 2010. http://www.thehindu.com/todays-paper/tp-opinion/changing-the-rules-of-the-energy- game/article682254.ece. Reed, Andrew. ‘Russia’s ESPO Blend Crude Has a Bright Future Ahead’. FSU Oil & Gas Monitor , no. 699 (12 September 2012): 2–3. http://www.esai.com/petroleum/12/pdfs/FSUOGMexcerpt.pdf. Saidi, Nasser, Fabio Scacciavillani, and Fahad Ali. ‘The Dubai Mercantile Exchange: Trad- ing, Prices and Market Efficiency’. Economic Note. Dubai: Dubai International Fi- nancial Centre, May 2011. ‘Saudi Arabia Increases Focus With Japan Oil Storage Deal’. Middle East Economic Survey 57, no. 08 (21 February 2014): 14. Shanghai Futures Information Technology Co., Ltd. ‘Company’s Briefly Information’. SFIT . Accessed 11 February 2015. http://www.sfit.com.cn/egs1.htm. Shiryaevskaya, Anna. ‘Rosneft Completes $55 Billion TNK-BP Russian Oil Acquisition’. Bloomberg , 21 March 2013. http://www.bloomberg.com/news/2013-03-21/rosneft- completes-55-billion-acquisition-of-oil-producer-tnk-bp.html. U.S. Energy Information Administration. ‘Exports’. Petroleum & Other Liquids , 27 February 2015. http://www.eia.gov/dnav/pet/pet_move_exp_dc_NUS-Z00_mbblpd_m.htm. ———. ‘U.S. Field Production of Crude Oil’. Petroleum & Other Liquids , 28 May 2015. http://www.eia.gov/dnav/pet/hist/LeafHandler.ashx?n=PET&s=MCRFPUS2&f=M. Wong, Fayen, and Chen Aizhu. ‘UPDATE 2-China Private Energy Firm Wins Rare Crude Oil Import Permit’. Reuters , 28 August 2014. http://uk.reuters.com/article/2014/08/28/china-crude-guanghui-stone- idUKL3N0QY0FY20140828.

-48-