The Trading and Price Discovery for Crude Oils

Abstract

What is the price of oil? Most oil in the world is never traded. Value of exchanged oil is set by key ‘benchmarks’ grades or baskets of crude oil that are commonly traded, both as physical and ‘paper’ barrels, in volumes many times greater than the world production. It is the derivative trades in these benchmarks such as Brent, WTI, Dubai and Oman that set the global price of oil and they are the most interesting feature of the oil market. Markets in these benchmarks are the stage on which two distinct categories of players are differentiated: the price makers and the price takers. How and why these benchmarks came to exist, how they work, interact and change over time is the subject of this chapter.

1 Introduction: Global Oil Markets

The oil market is by far the biggest commodity market in the world. Several billion dollars of physical oil is bought and sold every day. However, this is only tip of the iceberg. Not only are some cargoes often traded more than once, but a whole array of forward, futures, swaps, options and other derivative products have developed around physical exchanges. On August 16, 2019, in a single day, the two main oil exchanges traded, in two crude oil futures contracts alone, an equivalent of well over two months of global oil production.¹ This was an exceptional day, following an attack on Saudi oil facilities. However, this did not stop traders in the US, Europe or anywhere else in the world from buying US light sweet crude oil or North Sea Brent futures contracts, thousands of miles away from the areas affected by the conflict. Oil markets are linked by these commonly traded types of oil (or ‘benchmarks’), creating one global pool. This is oil markets’ most interesting feature: oil derivatives trade in far greater volumes than physical oil, and a majority of the physical barrels are not traded at all: they are supplied on long-term contracts at the price set by one of the global benchmarks. Thanks to the derivatives associated with these benchmarks, oil can seamlessly flow around the globe with relatively little risk. As we shall show, the world oil markets are all about benchmarks and their working.

It helps that oil is relatively easily transported and stored, so that factors influencing prices in one market very quickly spread and resonate in all other oil markets. However, global oil markets, prices and their interactions have taken decades to develop. The first section of this chapter will give a historical perspective and understanding of how and why oil markets developed, why they are so volatile and why they have their current form.

There are hundreds of different types of crude oil which can substantially differ in quality.² Over time, their prices evolved as differentials to most commonly traded benchmarks. Benchmarks are well accepted, commonly used, types of crude oil that have many buyers and sellers with prices which are very liquid and transparent.³ The main global benchmarks are West Texas Intermediate (WTI) in the US, Brent in Europe and Dubai and Oman in Asia. Their prices reflect their quality and fundamental factors prevailing in the regions in which they trade. We shall discuss the main regional benchmarks, how they work and all the main derivative markets associated with them. Benchmarks have their own dynamics and they change as markets change. They evolve over time and new, competing benchmarks emerge. All of this will be discussed in the second half of this chapter.

Global oil markets are linked by arbitrage. If gasoline rich oil from Vietnam becomes scarce and dear, similar quality of oil from Libya, Norway or the US will soon find its way to Asia. This is possible to do in a relatively riskless manner, thanks to well developed and mature ‘paper’ markets around the benchmarks in each region, with layers of associated derivative products designed to mitigate specific risks. For example, moving oil from Norway to Asia does not only involve buying and selling physical oil and shipping it. The whole process takes at least a couple of months and price risk over this period is enormous. Managing the risk between the purchase and sale is likely to involve ‘locking in’ the Brent—Dubai arbitrage⁴ (through a ‘swap spread’ or exchange for swaps—EFS⁵), hedging with forwards or futures and ‘contracts for differences’ (CFDs),⁶ exchange for physicals (EFPs),⁷ ‘rolling’ some contracts (through spreads) and so on. Interactions between the benchmarks, derivative instruments associated with them, arbitrage and global oil flows will be discussed in the last section.

Over time, market participants have evolved into two primary groups: ‘Price takers’ accept market prices as given and use them to manage risk. Usually, they are producing, or ‘upstream’ companies, refiners and other commercial entities involved in physical oil, and their primary focus is to reduce the price risk. ‘Price makers’ are companies actively involved in the price-making process by trading benchmarks and in the process, taking some price risk. In most cases, they are ‘buyers of risk’ and comprise of trading companies, banks, investment funds and the like. Of course, the picture is not clear-cut, and many major oil companies involve production, transport, refining as well as trading. The role of ‘risk takers’ or speculators in shaping the oil market and oil prices is discussed. Key market players will be discussed in detail.

Oil has become and asset class in its own right. High oil prices tend to be negatively correlated with a number of assets such as bonds and equities.⁸ This makes oil derivatives potentially attractive assets in any portfolio. ‘Financializaton’ of oil has introduced new actors into the oil markets and changed dynamics of trading and oil prices. ‘Financialization’ of oil markets, new methods of trading such as algorithmic trading (often referred to as ‘black box’ trading) and artificial intelligence (AI) and their impact will be discussed in the final section. The concluding remarks will summarise the main points of the chapter and hint at possible future development of the oil markets (Fig. 18.1).

Fig. 18.1

WTI Open Interest (OI) on CME Exchange (Number of contracts, each 1000 barrels) indicating a relentless growth in oil trading over time. Data from CME. (Source: Author’s elaboration)

2 Oil Industry and Markets: A Brief Historical Introduction

History of oil markets is a history of natural monopoly, booms and busts.

Natural monopoly can emerge in perfectly competitive markets. Economists define it by presence of both economies of scale and ‘sub-additivity’. The latter simply means that it is more efficient (cheaper) to have one provider than two or more. Oil projects tend to be big, capital intensive with long gestation periods, with assets specific that can be used for very specific purposes and usually, for very long-time periods. Capital spending tends to be front-loaded, with returns on investments enjoyed many years later. After the capital has been sunk, the operating costs are relatively small, making it harder to reduce or stop the use of assets. Drilling rigs, pipelines and refineries are only a few good examples. This makes energy projects sensitive to prices, interest rates and politics. They are risky.

Large energy investments usually come in discrete, indivisible ‘chunks’ to achieve economies of scale. Refineries, ships, pipelines and other assets are designed and built to an optimal size that is difficult to adjust and change quickly. Economies of scale make one larger project cheaper than several small ones for the same purpose. Natural monopolies are generally resolved either by breaking the integrated structures (caused by new technologies, entrants, supplies etc.) or by government regulation.

Breaking monopolies has generally led to emergence of competitive markets, lower prices and technological improvements. Competition can lead to extreme volatility and even waste while monopoly can give energy markets predictability and stability. Of course, on the downside, monopoly normally leads to higher prices and barriers to entry and change. For this reason, examining the market structure and the way it changes over time can give us an insight into development of markets and prices.

2.1 Beginnings of the Oil Industry in the US

In the early days of the industry in Pennsylvania in 1860s and 1870s, discoveries led to a rush for drilling and intense competition to produce, refine and transport the commodity. But the ‘boom’ soon led to overproduction, waste and the eventual collapse in prices. Then, as now, producers made attempts to reduce output,⁹ but cheating and free-riding was common, with disastrous results. Between 1860 and 1862, oil prices fell from $20 to $0.20 per barrel (McNally 2019).¹⁰ These booms and bust cycles went on well into the end of the century. It was the railroads, another ‘natural monopoly’, that helped Rockefeller tame and ultimately control the oil industry (Tarbell 1904).¹¹ This kind of control of the industry from drilling to marketing of products (and everything in between) is called ‘vertical integration’. It protected monopoly and prices by creating ‘barriers to entry’.¹² But during the Rockefeller monopoly, price volatility fell by about a half (McNally 2019).¹³ Although his Standard oil company was broken up in 1911 by the Roosevelt administration, its legacy will live on in companies that inherited it: ExxonMobil, Chevron and others and will dominate the oil markets in the 1950s and in some ways, ever since. Government intervention in the oil industry not only continued but also intensified from the British government’s purchase of 51% of British Petroleum (Anglo-Persian Oil company) in 1913, ‘Texas Railroad Commission’ in the 1930s, nationalisations and the birth of OPEC in 1960 to the lifting of the US oil exports ban in the 2015. Strategic importance of oil in the transportation sector facilitated the World War I and was probably one of the ultimate goals (or ‘The Prize’)¹⁴ (Yergin 1991) in the Second one.

2.2 The ‘Majors’

In the 1950s and 1960s, international oil markets were largely controlled by large oil companies or oligopoly of oil ‘Majors’ (Yergin 1991).¹⁵ Aside from being vertically integrated, the industry was also integrated ‘horizontally’; oil was carefully supplied from various geographic areas to ensure that supply and demand were balanced at lowest possible cost. Such integration enabled the supply of oil to be fine-tuned to the prevailing demand for end products, thus ensuring price stability.

Horizontal integration was done through joint ownership of several operating companies throughout the Middle East (ARAMCO, Anglo-Persian—later to become British Petroleum, Kuwait Oil Company etc.). These joint ventures continued historical and old colonial links involving the British, American, French and other governments. Companies worked with and closely followed the strategic interest of their governments. They operated through mutual agreements, preventing ‘harmful’ competition.¹⁶ The share of each Major oil company in any of the markets was to remain ‘As-Is’¹⁷ (in line with the prevailing market shares in 1928). Also, the world oil markets were to be supplied without ‘disruptive price competition’ (Yergin 1991).¹⁸

Mechanisms for controlling these large petroleum reserves were ingenious, generally based on various ‘rules’ such as Average Program Quantity (APQ) in Iran, the ‘Five Sevenths’ rule in Iraq and the ‘Dividend Squeeze’ in Saudi Arabia (Sampson 1988).¹⁹ Monopoly prices were charged using ‘Gulf Plus System’; all delivered oil prices included freight cost from the US Gulf regardless of its origin.²⁰ By 1950, the ‘Seven Sisters’²¹ owned 70% of the world refining capacity outside the Communist block and North America, almost 100% of the pipeline networks and over 60% of the world’s privately owned tanker fleet. They priced crude oil using ‘posted prices’ to maximise they own revenues within their vertically integrated systems. The producing countries were receiving royalties on percentage basis and posted prices were kept low to minimise this cost. They had done so through collusion while it was possible²² and through joint ventures which provided them with information to control the market and avoid competition. As Fig. 18.2 below illustrates, during this ‘golden period’ of their reign, oil prices were low and very stable.

Fig. 18.2

Major oil companies control of the market and prices. Author, prices, BP Statistical Review. (Source: Author’s elaboration)

2.3 The ‘Independents’

Towards the end of the 1950s, cracks were appearing in this structure. Existing oil producers such as Venezuela and Iran were pushing for higher production and revenues. Large profits of the oil ‘Majors’ were attracting ‘newcomers’ in terms of smaller ‘independent’²³ oil companies. These newcomers, such as J.P. Getty paid producing countries more for concessions and offered higher royalties (Sampson 1988).²⁴ Despite paying a lot more, Getty still made a fortune. It did not go unnoticed among the producers. In 1956, a French State oil company Elf²⁵ discovered oil in Algeria. In Libya, the 1955 Petroleum Law offered many smaller concessions and stricter terms for exploration than the existing producers did (Yergin 1991).²⁶ In 1956, BP and Shell found oil in Nigeria. More than half of the Libyan production ended up in the hands of companies which had no integrated systems in Europe (such as Conoco, Marathon and Amerada Hess) and hence no outlets for the oil. In 1959, ENI started importing cheap Russian oil from the Urals region into Italy, undercutting prices set by the Majors. By the end of 1950s, the USSR became the second largest producer in the world, after the US. It produced a volume of oil that could compete with the Middle East (Yergin 1991).²⁷ At the same time, the US had an import quota system, designed to protect the domestic oil producers (Sampson 1988).²⁸ This legislation left the independents ‘stranded’ with the oil which had to find markets elsewhere, in the world markets and put further pressure on prices. These newcomers, not unlike the shale producers now, were keen to get the oil out of the ground quickly and secure a return to their investment. Despite growing demand in this period (Yergin 1991),²⁹ the excess supply was becoming obvious. Oil had to be offered at a discount to the posted prices. This was making the royalties paid to the producing countries effectively higher than the ‘usual’ 50% prevalent at the time. The Majors were losing not only money and market share, but also the ability to balance the supply and demand and thus prices.

The integrated structure of the industry was crumbling. In 1946, nine oil companies operated in the Middle East. By 1970, this number reached 81 (Yergin 1991)³⁰! By 1966, very little crude was traded at posted prices. In the 1960–1965 period, Majors’ share of the European refining capacity fell from 67% to 54%. This intensified competition in the products markets reducing the refinery margin and putting further pressure on the oil prices. The growing competition³¹ for oil concessions between the Majors and the ‘independent oil companies’, coupled with the falling oil prices (Yergin 1991)³² put increasing pressure on the relationships between the oil producing countries and the Majors.

2.4 The Oil Cartel

In 1950s, a new breed of populist leaders emerged in some of the producing countries Nasser in Egypt, Muammar al-Qaddafi in Libya, Boumediene in Algiers, Abdullah Tariki in Saudi Arabia and Perez Alfonso in Venezuela. Apart from being better educated, they sheared strong anti-colonial feelings. The Algerian President argued for the producing countries to lead the ‘Third World’ towards a more equitable and just global world order (Sampson 1988).³³ Better terms offered to producers by the ‘newcomers’ made it very clear that the ‘old’ terms agreed with the Majors were a bad deal. On 9th of August 1969, Exxon, the largest Major and a price leader, announced a posted price cut of up to 14 cents per barrel without warning or consulting the governments of the producing countries. Other Majors in the M. East followed suit. The producing countries were outraged and swiftly arranged a meeting on September 10th in Baghdad.³⁴ Four days later, OPEC was born. Its major objective was to defend the price of oil (Bhattacharyya 2011).³⁵ OPEC was coloured by political ideas of the time. Nationalisation echoed from the OPEC Caracas meeting in December 1970 and Qaddafi implemented it after the Beirut meeting of the cartel in 1971.³⁶

The early 70s ended an era of the control of the crude oil industry by the Major oil companies. The control of production decisions shifted from these companies to the national governments of oil producing countries, usually through their national monopolies. The idea was that the national oil companies (NOCs) would give them power to decide output and hence influence the oil prices. Since NOCs had few assets such as ships, refineries and distribution networks, the distinction between buyers and sellers of oil in the international markets became obvious. As the market control by the Majors broke up, oil price (as illustrated in Fig. 18.2 below) became highly volatile.³⁷

3 ‘Spot’ Market and Prices

Following the collapse of the horizontal integration in the early 70s, the vertical integration of the industry ended with the fall of Shah of Iran in 1979. After the revolution, the Majors were forced to cancel the third-party oil deliveries from the country, which drove buyers of oil into the spot market (Treat 1990).³⁸ With no vertically integrated market tightly controlled by the companies, discrepancies between nomination dates, quantities, types and location of crude oil purchased by refiners became an issue.³⁹ To remedy these problems, long-term contract holders had to swap and trade different types of oil among themselves. The result was a massive growth of the volume of spot trades from some 3–5% in January 1979 to about 15–20% (Treat 1990)⁴⁰ by March of the same year.

Throughout its history, OPEC has had a fair share of infighting in their ranks, mainly over pricing policy and the fundamental long-term strategy. This was particularly obvious in the 1980 conference in Algiers where a lack of agreement led to a ‘free for all’ policy. By adhering to the system of ‘Official Prices’ which most of OPEC was abandoning (due to competition and oversupply, they were too high), Saudi Arabia was forced to reduce the volumes and take on the role of a ‘swing producer’.⁴¹ But the rigid official prices were falling out of line with the ‘real’ spot market. The House of Saud rejected a continuous decline in the volume of their exports (exports fell from about 10 mbd to just 3 mbd between 1980 and 1986!). They opted to recover their share of the world market by selling their oil at ‘netback’ prices.⁴² A year later, the oil prices fell to $8/ barrel (World Bank 1995).⁴³ With OPEC unable to control the supply, the industry resembled the early days of intense competition, booms and busts. With growing volatility,⁴⁴ the price risk had to be managed. This was especially the case as the term supplies from OPEC countries were generally of a long-haul nature.⁴⁵ In a volatile market, the price could significantly change between the time of purchase and delivery, exposing both seller and buyer to a large financial risk. The risk could be ‘hedged’ by selling relatively liquid forward Brent crude at the time of purchase and buying it back around the time of delivery. To make this risk management easier, the International Petroleum Exchange was launched in 1980, soon followed by an oil contract underpinned by the largest single grade in the North Sea, the Brent Blend.⁴⁶

Given that neither the official OPEC prices nor the ‘netback’ prices were acceptable any longer, a system of ‘spot’ related formulae prices was gradually adopted.⁴⁷ By 1987, over 60% of the oil prices were tied to the spot market prices.⁴⁸ This marked the birth of the modern oil market. This principle of setting prices for individual grades of oil against a published benchmark⁴⁹ has not changed to this day.

4 Oil Price Benchmarks

In all energy markets, government policy is critical, and oil is no exception. In response to the Arab oil embargo of 1973, US government imposed price controls through the Emergency Petroleum and Allocation Act (EPAA). In a well-supplied market of 1981, the US government lifted these controls, opening the markets to competition, trading and transparency. This ‘liberalisation’ of the market was instrumental for the success of the new, ‘paper’ contract for the domestic light sweet crude, West Texas Intermediate (WTI) with a delivery point in Cushing, Oklahoma. The contract was listed by the New York Mercantile Exchange (NYMEX) in March 1983, alongside then existing heating oil and gasoline contracts. It was a physical oil delivery contract, designed to mimic well established, physical trading around the Cushing hub.

OPEC pricing arrangements were also challenged by large discoveries outside the cartel members, particularly in the North Sea. The Ekofisk oilfield was discovered in 1969 in Norway by Amoco, the Forties field in 1970 by BP and the Brent field in 1971 by Shell (McGrandle 1975).⁵⁰ By 1980, the North Sea production was 2 million barrels per day (mbd), making the region an important supplier of non-OPEC crude oil. Underpinned by English law, standardised contracts,⁵¹ no destination restrictions and tax advantages in ‘spinning’ or ‘churning’ the cargoes, North Sea Brent market developed as the prime, transparent and liquid spot market (Mabro and Horsnell 1993; Fattouh 2010).⁵² Price reporting agencies (PRAs) Argus and Platts added to transparency of the market.⁵³ The benchmarks were set by participants in the spot and paper markets (‘price makers’) while the producers became ‘price takers’.⁵⁴ Oil price became the price of one of these grades. All the risk management involved derivative markets which grew alongside Brent, WTI and Dubai. These benchmarks became the backbone of oil trading.

4.1 West Texas Intermediate (WTI)

Throughout the modern history, the US has been the world single largest regional market.⁵⁵ Cushing, Oklahoma is at the crossroads of the US pipelines linking production in Oklahoma and West Texas (Midland), and the refining centres of Midwest, Midcontinent and US Gulf (see Fig. 18.3 below) in the country. Together with a massive storage capacity in tens of millions of barrels⁵⁶ and a large number of participants, the hub ensures unprecedented liquidity in trading light sweet crude oil,⁵⁷ surpassed at times only by Intercontinental Exchange (ICE)⁵⁸ Brent oil contract.

Fig. 18.3

US Petroleum Administration for Defence Districts or PADDs; The figure presents the key PADD 2 & PADD3 districts for WTI Benchmark formation. (Source: Author’s elaboration)

Interaction between the oil gathering centres, pipelines, refining and import/ export facilities is the key to understanding development and dynamics of the WTI benchmark. As domestic production and refining changed, so did the infrastructure linked to Cushing. The pipeline links to the US Gulf (USG) are essential in keeping the benchmark linked to the rest of the world. When domestic production fell in the mid-1980s, it was the imports of foreign sweet barrels that set the price of WTI.

Even though the US imposed a ban on oil exports between 1977 and 2015, the sheer size of the US market provided the trading liquidity, making WTI one of the two most important oil benchmarks. The key events leading to the birth of the benchmark happened in 1980s: Lifting of the price controls in 1981; Setting up of the WTI futures contract on the New York Mercantile Exchange (NYMEX) in March 1983 and the oil price collapse in 1986. Following ‘decontrol’ of prices in 1981, spot trade quickly grew and the price reporting agency (PRA) Platt’s started surveying and publishing prices for WTI as well as the sour grades, Louisiana Light Sweet (LLS) and West Texas Sour (WTS) (Purvin & Gertz 2010).⁵⁹ The NYMEX WTI contract had a good and well established underpinning in the physical trades (Purvin & Gertz 2010).⁶⁰

Being land-locked, oil balances in Cushing are subject to changes in flows and infrastructure in and around the hub. For this reason, the oil price crash in 1986 led to a fall in domestic production by over 1.5 mbd⁶¹ and change in flows around the hub (Purvin & Gertz 2010).⁶² Starved of the local crude oil, inland refineries had to import oil from the Gulf (USG) using reversed pipeline flows.⁶³ Given the ‘price war’ in the international markets at the time, foreign imports became competitive and soon the WTI introduced an ‘alternative delivery procedure’ which allowed for the foreign sweet crudes to be delivered into the contract.⁶⁴ This increased the ‘depth’ of the market; there was more oil to be delivered into the contract and more new players to do so. Open interest took off in 1986 and grew steadily from about 100,000 contracts then, to about half a million contracts in 1990s (see Fig. 18.1—each contract is one thousand barrels).

The following decade and a half saw relative price stability around $20 per barrel. Iraq war in 1991 was followed by a large release of oil from the US Strategic Petroleum Reserve (SPR) and eventually the price stabilised. Another market event in this period was the 1997–1999 Asian Financial Crisis when the prices fell from the peak of about $25 to $10, but eventually recovered and picked up in 2000. However, it was geopolitical events—the ‘9/11’ terrorist attack and the subsequent invasion of Iraq in March 2003—that rattled the market and increased demand for ‘paper barrels’ for mitigating this volatility. By the end of 2004, WTI had crossed $50 mark. The market was entering a new period of dizzyingly high economic growth in Asia and particularly China. Thus, began a period of ‘financialisaton’ of the oil market with price action being dominated by new players such as funds and other financial institution. This is an important development in the history of the oil markets and will be discussed separately, later in the chapter.

Like most benchmarks, WTI benchmark has faced some difficult times. Perhaps the most serious one was in the 2005–2015 period. It started somewhat earlier, in the mid-1990s with increasing Canadian oil sands production,⁶⁵ which found its main outlet in the US. These cheap barrels had a natural outlet in the Midwest and Midcontinent refineries which invested in upgrading their facilities to take advantage of them.⁶⁶ The greatest changes in the WTI benchmark came from explosion in the US shale production and the eventual lifting of the oil export ban in 2015. It can be seen in the Fig. 18.4 below that oil production picked up substantially after 2011, choking up international imports and resulting in domestic oversupply of oil. As there was a ban on US oil exports⁶⁷ at the time, it created an excess supply of oil and resulted in the price of WTI falling relative to other benchmarks. Figure 18.5 clearly shows this decoupling of WTI not only relative to Dated Brent but also relative to LLS. Oversupplied and banned from exports, WTI started trading at deep discounts to Brent as well as LLS. It decoupled from the international markets, and USG refineries as well as exporters of sweet crude from Europe and West Africa relied primarily on LLS as an indicator of market fundamentals in the USG refining hub. Unhappy with the state of affairs, Saudi Aramco switched from Platts WTI benchmark to The Argus Sour Crude Index (“ASCI”)⁶⁸ in 2009, followed by Kuwait and Iraq later.

Fig. 18.4

US Crude Oil Production, EIA Data. (Source: Author’s elaboration)

Fig. 18.5

Dated Brent vs. WTI and LLS (Argus spot assessments data). (Source: Author’s elaboration)

The isolation of WTI eased off with a reversal (yet again) of the Seaway pipeline from Cushing to USG in 2012.⁶⁹ However, as can be seen in Fig. 18.1, the open interest of WTI contract did not fall as much as one would expect. This proves one very important point regarding benchmarks in general: liquidity of the contract and the confidence that traders get from it is often far more important than the basis risk⁷⁰ involved. Traders value the ability to enter and exit contracts without fear of being ‘squeezed’ long or short. A contract that provides this assurance will be more successful than an illiquid contract which provides less basis risk (Williams 1986).⁷¹ This ‘rule’ will be confirmed in the case of Dubai, Oman, Murban and many other aspiring oil benchmarks.

Eventually, the US oil export ban was lifted in December 2015, relieving the pressure at Cushing, and WTI connected again with the international markets, increasing and eventually achieving record volumes of volume of open interest.⁷²

4.2 Dated Brent

Arguably, ‘Dated Brent’ is the world’s most important oil benchmark. It dominates as a pricing reference for the Atlantic basin (North Sea, Mediterranean and Africa) and for most ‘sweet’ (low sulphur) crude in Asia (Australia, Malaysia, Vietnam and others).⁷³ It is generally accepted that Dubai, the main benchmark in Asia, or more generally, ‘East of Suez’, is essentially (as discussed in the next section), a spread to Brent.

The Brent field was discovered in 1971 north of British Shetland Islands in the North Sea and the first ‘Bravo’ platform started production in 1976. In the 1980s, it was producing between 400,000 and 500,000 barrels or roughly just short of one cargo per day (one cargo = 600,000 barrels). It loads at Sullom Voe terminal in the Shetlands. Brent ‘paper’ has evolved from being a ‘forward’⁷⁴ market in physical cargoes in the 1980s, to become the most complex oil market in the world (Mabro and Horsnell 1993; Fattouh 2010).⁷⁵ ‘Brent’ is a brand name of a benchmark that has reinvented itself many times since 1980s. Due to falling production, other sweet North Sea grades were gradually introduced into the Brent delivery mechanism forming what we now call a Brent or ‘BFOET basket’ comprising and named after Brent, Forties, Oseberg, Ekofisk and Troll⁷⁶ crude oils. Physical volumes of oil in the ‘Brent basket’ have increased over time by widening the ‘window’ of cargo loadings which qualify for the price assessment of Dated Brent. From the beginning, in 1987, to 2002, this ‘window’ was up to 15 days ahead of the date of assessment (often referred to as 15-day Brent); in 2002, the window was expanded to 10–21 days ahead; in 2012 it was expanded to 10–25 days ahead; finally, in 2015, it was extended yet again to a 10 days–one calendar month forward ‘window’. Each of these changes added to the volume of oil trade included in the assessment. As we can see in Fig. 18.6 below, this has stabilised the volume of reported deals in the ‘Brent basket’.

Fig. 18.6

Number of physical BFOET or ‘Brent Basket’ deals (cargoes, 600 kbd each, including all the BFOET grades; Argus data). (Source: Author’s elaboration)

While Brent trading started as ‘forward’ physical or ‘cash’ market where oil was ‘churned’ for tax optimisation reasons (Mabro and Horsnell 1993),⁷⁷ Dated Brent is simply Brent with loading dates,⁷⁸ often referred to as ‘wet’ (as opposed to ‘paper’ Brent with no loading dates, in the forward markets). Dated Brent being a price of actual physical oil is generally used as a benchmark for physical trades of other types of crude oil. To understand what the actual Dated Brent benchmark price is, it is necessary to understand how its price is assessed. In a nutshell, it is based on four pillars:

• Physical assessment of the value of the BFOET (‘Brent’)⁷⁹ grades.

• A forward curve based on the Dated swaps market.

• The fixed price of the forward or futures ‘Brent’ contract.

• Quality differentials or premiums (QP) of crudes other than Brent or Forties.⁸⁰

Assessment of the Dated Brent curve allows for a wide range of crudes to be priced off this single benchmark. PRAs such as Platts and Argus report a wide variety of price differentials (for instance, Urals, West African, Mediterranean, North Sea and even many Asia-Pacific crudes) vis-à-vis the Forward Dated Brent (Argus refers to this as Anticipated North Sea Dated). This allows the refineries to compare the relative value of the different crudes and assess which crudes are being valued competitively relative to a single benchmark.

The most peculiar feature of the physical BFOET market is that it is generally traded as a differential to Dated Brent. Therefore, the PRAs are challenged to assess the Dated Brent price based on physical trades which are themselves differentials to Dated Brent! Fortunately, the expected assessments for Dated Brent are traded in a liquid derivatives market as weekly swaps,⁸¹ called Contracts for Difference or CFDs.⁸² Historically, CFDs developed from a need to convert an outright Forward Brent price into a Dated Brent price plus a differential (and vice versa). In the 1980s, most of the Forward Brent contracts were traded on an outright price basis. As alternative grades of oil used Dated Brent pricing around the Bill of Lading as a reference price (normally a five-day period taken two days before loading date and two days after the loading date), there was a need to value them using a common denominator. Given that CFDs trade at least six to eight weeks forward, they are used to construct the Brent Dated forward curve. The example below will help illustrate this.

CFD Brent swaps are differentials between Dated and forward Brent values. For example, 1–5 April CFD swaps may trade at June forward Brent, minus 50 cents per barrel (−$0.50). The following week (8–12 April), they may trade at −$0.30, and so on. PRAs need to establish these values (forward curve) as they are the key to resolving the circular problem where physical or Dated Brent normally trades as a differential to Dated Brent! Let’s take an example of a cargo of Forties crude (one of the grades in the ‘Brent basket), loading 2–4 April, traded at Dated+$0.50/bbl and another cargo of the same grade loading 9–11 April traded at Dated+$0.30/bbl. Given the above CFD values, they have both effectively traded at the same absolute price, equal to June forward Brent⁸³ (−$0.50 + $0.50 = 0 and −$0.30 + $0.30 = 0). The actual value for June forward Brent is established at the end of a ‘window’ at 16.30 London time (S&P Global Platts 2019; Argus Media 2019)⁸⁴ and the above differentials are added or subtracted from it.

The higher quality grades in the BFOET basket such as Ekofisk, Oseberg, and Troll have a quality premium (QP) applied⁸⁵ to ‘normalize’ the differentials for the assessment process. Brent’s most representative grade is usually Forties—due to its relatively, but not exclusively, high sulphur levels—and it commonly establishes the value of Dated Brent. The quality of Forties crude may sometimes vary depending on the contribution of the Buzzard field,⁸⁶ and a sulphur de-escalator is applied later, to compensate buyers when the level of sulphur is above 0.6%.⁸⁷ This whole process happens in the London ‘window’ between 16.00 and 16.30 BST, with most trades being done during the last minute of the ‘window’. This is a somewhat simplified rendition of the process.⁸⁸

What is clear is that derivatives markets, namely CFD swaps play a key role in establishing the value of Dated Brent. Hence, sometimes the criticism that ‘the (derivatives) tail is wagging the (physical crude oil value) dog’ is heard.⁸⁹

Brent has been trading as a futures contract since 1983 and it is listed on both ICE and CME exchanges. It is normally financially settled on the last day of trading based on an index⁹⁰ calculated based on the physical trades on the last day of the contract. Traders with a futures position in Brent can easily turn it into a physical delivery contract through an ‘EFP’ (Exchange for Physicals) trade, usually through a broker, in one simple transaction. This establishes a pretty seamless link between futures and physical oil, making Brent one of the most robust benchmarks in the world and it serves as a basis for another regional benchmark used ‘East of Suez’, Dubai. What is more, the most common grade setting the Brent benchmark, Forties, is popular in the Far East. With falling energy demand in Europe and growing Asian thirst for the commodity, after about 2010, the North Sea oil has depended on Asian demand to balance the market. This can be clearly seen on Fig. 18.7 below, as more and more Forties ended up being refined East.

Fig. 18.7

Destination for the North Sea Forties crude over time. Argus data. (Source: Author’s elaboration)

4.3 Dubai: Brent’s Asian Cousin

Dubai crude oil has been the main Asian benchmark since the mid-1980s (Fattouh 2012).⁹¹ It is responsible for the pricing of almost 30 million barrels per day (million b/d) of crude oil currently exported to Asia. Since its introduction, Dubai production has diminished substantially. Dubai does not release figures for its crude oil production but, from the loading data and sales, it can be deduced that production has fallen below 70,000 b/d in 2019 from a peak of about 400,000 b/d in 1991.⁹² However, just like Brent, the benchmark has evolved into a ‘brand name’, allowing for the delivery of Oman, Upper Zakum, Al Shahen and Murban grades of oil in into the ‘Dubai basket’ during the so called ‘Platts Dubai window’, between 16.00 and 16.30 Singapore time.⁹³ Dubai partials trade (on a fixed price basis, in dollars per barrel) only during this half hour window. For the remainder of the trading day, all Dubai trades are still differentials to Brent (EFS) or spreads to other Dubai swap months (swap spreads). A large derivatives market has grown around the Dubai ‘brand name’, feeding back into the price discovery of the benchmark itself (Fattouh 2012).⁹⁴

As discussed earlier, the 1990s and early 2000s have witnessed two main themes in the world oil markets. The first is a shift in demand from the developed to the developing world, particularly towards Asia and the Middle East (ME). The second is a large increase in light oil and gas production in the Americas (Imsirovic 2014).⁹⁵ The consequences for the crude oil and petroleum product flows to Asia, as well as main price benchmarks, have been profound.

United States East Coast (USEC) and Canadian refiners, traditional buyers of high gasoline yield crude oil from the North Sea (NS), West Africa (WAF) and North Africa (NA), have essentially stopped importing, given the availability of locally produced, light sweet shale crude oil. Sweet barrels from the Atlantic basin, which mainly trade on spot basis and have no destination restrictions (unlike most OPEC crude oil), have become ‘swing’ barrels for the Asian refiners looking for cheaper feedstock (Fattouh and Sen 2014).⁹⁶

Given weak European demand and poor margins, European refineries have been closing during this period.⁹⁷ Most of the new and more sophisticated refinery capacity was being built in the ME and Asia at the expense of Europe. This was exacerbated by the Russian tax incentives to increase product exports at the expense of the traditional crude oil, in order to maximise revenue.⁹⁸ Despite the slowing Chinese economy, the main demand drivers in the oil markets continued to be China and the ME.⁹⁹ Therefore, the oil flows shifted towards Asia from almost all the producing areas. Since the oil prices are set by the ‘marginal’ barrels and the region is a main buyer of these barrels, Asia was the main driver of global oil prices. Hence, delivered price of oil in Asia was the key signal for the world oil markets. At the same time, ME producers became more dependent on Asian buyers. This resulted in increased market power of the Asian consumers and the demise of the so called ‘Asian premium’ (Doshi and Imsirovic 2013).¹⁰⁰

The growing importance of Asia as a destination for oil from all over the world has profoundly impacted the Dubai market. ‘Arbitrage’ barrels that normally trade against Brent and WTI benchmarks are generally being evaluated by end users (refiners) and sold on Dubai-related prices.¹⁰¹ This means somewhere in the supply chain, prices may need to be converted from other benchmarks to Dubai prices. The process of arbitrage involves buying the benchmark other than Dubai (Brent, WTI)¹⁰² and selling Dubai swaps, in order to ‘lock in’ the differential that makes oil competitive in Asia. As Brent is the dominant international benchmark, Brent futures versus the Dubai swaps differential is also the dominant trading link between the two benchmarks.¹⁰³ It is known as EFS (exchange for swaps since Brent futures are ‘exchanged’ for Dubai swaps). For example, importing Brent-related barrels loading in the month of December to Asia and converting its price into a Dubai-related one, would normally involve a purchase of December EFS (buying December Brent futures and selling December Dubai swaps can be done as one trade by buying the December EFS). Then, as the cargo has ‘priced in’, usually during the loading period, December Brent futures are sold rateably. Since the cargo is placed with an end user at a Dubai-related price, the swap would be simply left to ‘price out’.

To understand what the ‘Dubai’ benchmark price actually is, it is worth briefly revisiting the process of assessing of the Dubai benchmark. Firstly, Asian refiners normally buy oil over a calendar month of loading.¹⁰⁴ Secondly, most physical crude trades as a differential to Platts Dubai assessment during a calendar month of loading, the value of which equates to the Dubai¹⁰⁵ swap for that month. Refinery Linear Programming models use Dubai swap values (normally based on an estimated forward curve) as a common denominator for comparing different grades of crude oil.¹⁰⁶

Unlike Brent and WTI, Dubai has no liquid functioning futures markets.¹⁰⁷ However, the over-the-counter (OTC)¹⁰⁸ markets for EFS, Brent-Dubai¹⁰⁹ swap spread and Dubai spreads frequently trade. Therefore, the way to establish a value of a Dubai swap is to use Brent and apply the prevailing EFS value to it. This is best illustrated by an example. Physical Dubai cargoes traditionally trade as a differential to Dubai assessments (equal to swap value during the month of loading). For example, physical Dubai loading in the month of October will trade as a premium or a discount to October Dubai, and would normally trade about two months earlier, during August. Also, in August, the most liquid EFS market will be October (October Brent futures and Dubai swaps). By applying October EFS to October Brent futures, a trader can obtain the October Dubai swap value. During August, when October Dubai normally trades, its value is equal to the calculated October swap, plus some differential (positive or negative), depending on the fundamentals of the market (Horsnell 1997).¹¹⁰

Like Dated Brent, The Platts window is dominated by a small elite of self-selected ‘price makers’ (Fattouh 2012),¹¹¹ with Shell, Glencore and Vitol accounting for about a half of all cash Market On Close (MOC) activity in 2019 (Fig. 18.8). Half a dozen players account for almost all the deals. Few, if any, of the NOCs are involved.¹¹² Of course, significant participation of one or more large producers could also produce a biased benchmark.

Fig. 18.8

Cash BFOE MOC Participation. Source: Platts. (Source: Author’s elaboration)

As already mentioned, other grades of oil are deliverable into the Dubai contract. As these two grades have higher net worth for most refiners, it is presumed they will be delivered at Dubai prices only when the Dubai price is above its ‘true market value’, providing the liquidity in the pricing ‘window’ and putting a ‘cap’ on the Dubai price, avoiding a possible ‘squeeze’.¹¹³

4.4 Oman: Dubai’s Neglected Sibling

A number of OPEC producers base their pricing formula on both Dubai and Oman.¹¹⁴ Oman is lighter with lower sulphur content, making it a higher value grade of oil and it normally trades at a premium to Dubai. It is also well accepted by most Asian refiners that it has no destination restrictions and it is frequently traded. With at least 50 physical cargoes produced every month and loaded outside the Strait of Hormuz, it has many characteristics of a good benchmark. It is a part of the ‘Dubai basket’ as it is deliverable into the Dubai contract.

Figure 18.9 below shows that Oman and Dubai prices, or a spread, can diverge by a dollar per barrel, and often more. While this difference can be significant, historically Asian refiners have not done much to hedge it. The evidence for this is a relatively low volume of Oman trading on the Dubai Merchantile Exchange (DME) outside the ‘pricing window’ and pure physical delivery. The DME Oman futures contracts settle daily, based on a weighted average of trades between 16.25 and 16.30 Singapore time (Exactly the same time as the ‘Singapore Dubai window’). In line with the usual timing of Asian oil purchases, this contract trades two months before the actual month of loading. So, during November 2019, the front month contract is January 2020. The Oman official selling price (OSP) is set using the monthly average of the DME Oman daily settlements. Physical Oman is generally traded as a differential to this OSP.¹¹⁵

Fig. 18.9

Oman vs. Dubai differential. Platts data for Dubai and DME Oman $/bbl. (Source: Author’s elaboration)

Even though Oman is used as a pricing basis by some of the most important producers in the world and is widely traded, it has remained a minor benchmark overshadowed by Dubai. This is partly to do with a difference in philosophy between some Middle East (ME) producers and Platts. For Platts, there is only one ME benchmark with Dubai ‘brand’ that currently encompasses the five grades of ME oil¹¹⁶ including Oman. But given the Saudis and the other ME producers’ preference to the ‘Oman/Dubai’ formula, Platts had no choice but to continue publishing the ‘Oman’ assessment as well.¹¹⁷ Indeed, the Oman assessment is not even included in the ‘Key Benchmarks’ section of their flagship publication, Crude Oil Marketwire.

In October 2018 Saudi Aramco switched to DME Oman daily settlement price (instead of Platts Oman assessment), in its pricing formula for Asian customers. However, the decision made little, if any, impact on the volumes of Oman traded. This seems to confirm the ‘rule’ about benchmarks in general (as discussed in the section about WTI) that traders prefer liquidity over basis risk. In other words, they are prepared to take some risk (pay some risk premium) and use an imperfect, but liquid instrument rather than eliminate this risk by trading an illiquid contract (which carries its own risks and possibly costs).

So far, we can conclude that the whole global oil market revolves around three (some would argue two: WTI and Brent, as Dubai is generally traded as a spread to Brent) main benchmarks. There is also a plethora of ‘quality’ or ‘regional’ benchmarks. Examples of these are Russian Urals, Kazakhstani CPC Blend, Nigerian Qua Iboe, UAE Murban, Indonesian Minas and many others. What they all have in common is that, while their price provides good signal regarding demand and supply of that particular quality and the market in which they trade, they all trade as a differential to Brent, WTI or Dubai. Strictly speaking, they are not benchmarks. True benchmarks trade at fixed price, in Dollars per barrel and thus set the ‘absolute price level’ for all other oil to trade against.

It should be clear from the discussion so far that the world oil market is dynamic and changing all the time. As demand and supply in different regions change, so do price benchmarks, sending signals to the market and thus facilitating global oil flows. Over time, these flows change and so do benchmarks. When the supply of oil in Cushing dried up following the 1986 price crash, ‘an alternative delivery procedure’ was introduced into the WTI contract incorporating many imported grades of oil. With the Asian boom of the early 2000s and ‘centre of gravity’ of the oil market moving East, Dubai benchmark needed major adjustments incorporating new types of crude such as Oman, Upper Zakum and later Al Shaheen and Murban. The same was true for Brent, especially as one of its main ‘basket’ crude, Forties started to move East, making it vulnerable to ‘squeezes’.¹¹⁸ Following the shale boom, the US exports swamped the European market and often exceeded the whole North Sea production. If the Brent benchmark did not change, it could become irrelevant (Imsirovic 2019).¹¹⁹ For the same reasons, the price of WTI in Cushing is less important for international buyers of the US crude. They are more concerned about the price of WTI in the USG area, where they load the oil. For this reason, all major PRAs and exchanges have launched some form of new WTI contract based there (Imsirovic 2019).¹²⁰ Finally, the Middle Eastern producers have long wanted more control over their export prices. While Oman has had some traction, there is a fair chance that Murban could emerge as a new benchmark in this region. This will require a lot of changes in the way this grade of oil is sold and traded, but rewards may well be worth it (Mehdi et al. 2019).¹²¹

5 Global Oil Markets: Benchmarks in Action

While benchmarks are the backbone of the global oil prices, not all market participants are involved in setting those prices. As briefly mentioned, the oil industry can be divided between those actively involved in setting up the benchmark prices and trading them and those who simply use those set prices for pricing of physical crude and hedging (Luciani 2015).¹²² It is important because prices of benchmarks are eventually set through participation in the market place, wether through exchanges, PRA ‘windows’ or any other way. In an ideal world, benchmarks should be set by all the market players participating in the price setting process; but this is not the case. This is a problem as it is sometimes suggested that the benchmark price is ‘too high’ or ‘too low’; but if the ‘too high’ prices are not sold into, and ‘too low’ prices are not bought by the market, there is no mechanism to ‘correct’ the benchmark price (Imsirovic 2013).¹²³ What is more, markets are not ‘perfect’ in the economic-theory sense: Like most other markets, oil is dominated by a handful of very large players and setting benchmark prices is often left to large traders or trading arms of large oil companies.¹²⁴ As a result, there is no perfect oil price benchmark.

However imperfect, benchmarks and their associated derivatives—futures, forwards, swaps, spreads, options and other instruments have created a market ecosystem facilitating smooth global movement of crude oil with relatively little price risk. Regional benchmarks such as Dubai in Asia clearly indicate the value of oil in that part of the world, set by local fundamentals of demand and supply. It is the function of the market to use this price signal and allocate the cheapest way to satisfy demand in Asia. It can be done with oil from the North Sea, West Africa, US shale regions or elsewhere, depending on the price of oil in those regions and the cost of moving oil to Asia. This is set by the spreads between benchmarks such as WTI/ Brent and Brent/Dubai. So, a tight Asian market will result in strong Dubai prices (and narrow WTI/ Dubai and Brent/ Dubai spreads). Narrow arbitrage spreads are akin to open doors. Once it is decided to move oil from one region to the other, spreads will be purchased to hedge risk. This will widen the spreads and ‘close the door’ for further arbitrage. Hence, these benchmark spreads are the key indicators of the relative oil market strengths around the globe. This requires the ‘paper’ oil to be traded many times the volume of the actual physical barrels being moved.¹²⁵

Of course, trading does not only involve the movement of oil through space. It also involves movement of oil in time and it this reflected in time-spreads. This is normally referred to as the ‘time structure’ of prices. Demand for oil is derived from the demand for products and it is highly seasonal. People drive more in the summer and heat houses more in the winter. Weak spot prices relative to delivery of oil in future is referred to as ‘contango’, and it is a signal to the market that cheap prompt oil can be purchased and stored profitably, as long as the ‘contango’ is greater than the overall cost of storage. Then it can be delivered later, when there is more demand for it. On the other hand, strong prompt price relative to future delivery, referred to in industry speak as ‘backwardation’, is a signal to the market that oil is needed and that it should come out of storage. In reality, trading involves movement of oil both in space and time. Contango in WTI and Brent are ideal for oil shipments over long distances such as Asia as the time spread is paying for at least a part of the shipping cost.

While oil benchmarks and spreads between them give a good signal where oil is most needed and offer instruments for moving it with little risk, there is no guarantee that the demand and supply of these ‘paper’ instruments are well matched (Imsirovic 2014).¹²⁶ This is because not every buyer and seller of oil mitigates risk through hedging. If they all had same internal measures and risk profiles, ‘paper’ markets would be ideally matched in terms of supply and demand. Unfortunately, this is not the case. So, a new class of market participants is necessary to take the excess risk: speculators. When returns on the commodity are high, speculators (investors) are attracted to the market and thus they provide the additional liquidity needed for the smooth functioning of the market.

6 Speculation and Financialization: Price Makers and Price Takers

It is generally accepted in virtually all the markets that speculators (Fattouh et al. 2012; Medlock 2013; Vansteenkiste 2011)¹²⁷ add liquidity, at a price. However, they may also add volatility (Einloth 2009),¹²⁸ especially in the global oil market with relatively low elasticity of demand and supply and high geopolitical risk. This was a particularly hot topic during the 2000s commodity boom when oil prices well exceeded $100 per barrel and ideas of ‘peak oil’ came back to vogue.¹²⁹

The commodity boom coincided with the growth in other financial markets. Large funds were shifting significant amounts of money into oil ‘paper’ markets looking for hedge against inflation as well as higher yields. The share of ‘non-commercial’¹³⁰ participants who are generally seen as ‘speculators’ increased from about 20% in 2001 to about 50% in 2006 and kept on growing until the financial crisis of 2008 (Medlock 2013).¹³¹ Some producers and OPEC in particular, were often deflecting the blame of high oil prices on ‘speculation’. However, the causality of the events is hard to establish: Did the ‘speculators’ and ‘financialization’ of the oil market cause the high oil prices or were they simply attracted by bullish market fundamentals and high returns? A large body of literature has been dedicated to this problem and their results widely wary. However, the general consensus seems to be that the prices were primarily driven by market fundamentals while speculation might have had influence in certain periods.¹³²

However, terms such as ‘financialization’, commoditization’ and innovation continue to dominate oil markets. Artificial Intelligence (AI), algorithms, ‘data mining’, ‘black box’ trading and so on are just some terms applied to various computerised trading strategies that increasingly dominate oil markets. It is impossible to read the future, but just like any other human activity, oil trading will be more and more dominated by the information technology.

7 Concluding Remarks

Oil markets, like most energy markets, are shaped by ‘natural monopolies’. These monopolies bring stability but also lack of transparency and lead to higher prices. They are usually broken when there is excess supply or by government intervention. Modern oil market has emerged following the collapse of the integrated structures of the oil ‘Majors’ in 1960 and the failure of OPEC to effectively control it.¹³³ The result has been more competition, spot trading, transparency, lower prices, but also higher volatility. The way markets deal with risk and volatility is by developing ‘paper’ or derivatives instruments which enable risk mitigation far into the future. In this respect, oil market is perhaps the most developed and sophisticated commodity market in the world. To facilitate trading in hundreds of different grades of oil, three dominant benchmarks have emerged—WTI, Brent and Dubai—representing the three major trading regions: US, Europe and ME/Asia. When we talk about the price of oil, we normally talk about the price of one of these three, especially Brent and WTI. All other types of crude oil are traded, one way or the other, using one of these three price markers. In order to manage price risk, market participants have developed a plethora of derivatives contracts to such an extent that futures, swaps, options and other contracts often far exceed the total global oil production many times over. With maturity, they have added to the complexity of the market. As we have shown, derivatives are often essential for the establishment of the physical benchmark prices themselves. Benchmarks have substantially changed over time, following changes in the underlying fundamentals of supply and demand. They are constantly being challenged by new potential benchmarks such as WTI in the USG, LLS, ASCI index, Oman, Murban and others. Adaptation and Schumpeterian ‘creative destruction’ are the way oil markets work, and this process will enable it to continue to function smoothly for as long as there is oil trading.