The Price of Oil: When Does “Complex” Become “Too Complicated”?

"Everything is complicated; if that were not so, life and poetry and everything else would be a bore." Wallace Stevens

It is often trotted out that the “right” price of any good or service is the highest price at which a buyer will buy and the lowest price at which a seller will sell. That’s all very well, but if only one party, or in some cases, neither party understands the price they are agreeing who is to say that the price is fair or a true reflection of market forces?

This blog will give a brief chronicle of how oil prices are agreed, not from the perspective of the supply/demand balance and the absolute level of oil prices, but by focusing on the oil price negotiation that takes place when oil cargoes change hands. Traders know all about this but oil industry professionals who are not directly involved in trading may wish to skim the ensuing explanation. It is not necessary for those readers whose day job is not trading to absorb the minutiae of the price formation process. However, an appreciation of the complexity of all the moving parts that interlock in pursuit of a fair market price for oil is advisable.

Safe to say, if you wanted to design the perfect oil pricing mechanism, crude oil or refined products, you would not start from here. But here is where we are!

Formula Pricing

To recap, the oil market abandoned fixed pricing, such at $X per barrel (“$X/bbl”) or $X/tonne (“$X/Mt)”, in contracts for the majority of physical cargoes starting in the 1980s. Instead, we went to formula pricing which reflected the published price of one of the fixed benchmark contracts, such as forward Brent or Dubai or the futures contract in WTI, averaged over an agreed period of time plus/minus a differential to reflect the value of any differences between the benchmark price and the oil in question. The market for refined products followed suit on formula pricing, developing benchmarks from Rotterdam, New York Harbour, the US Gulf Coast and Singapore, amongst others. This is old news!

The important point is that benchmarks have at least one of their contracts that are traded at a transparent fixed price, $X/bbl or $X/Mt, that can be used to solve the oil price formula in non-benchmark contracts.

Non-benchmark contracts contain a formula that allows the trader to agree a differential, $Z/bbl or $Z/Mt, to the benchmark number $X/bbl or $X/Mt, where Z itself may be a formula that expresses a value for all the differences between the characteristics of the benchmark grade of oil and the characteristics of the grade of the non-benchmark grade of oil question. The value of $X/bbl or $X/Mt in a non-benchmark price formula is usually the published value of X, averaged over a number of days publications. To this value of X a price differential, $Z/bbl or $Z/Mt, is added or subtracted.

Maintaining the Benchmarks

That was all fine, so long as there were sufficient observed fixed price deals transacted to establish the standard benchmark price, X. This benchmark price, averaged over the pre-agreed period, was then plugged into the price formulae in non-benchmark, which we will call “physical”, oil contracts for convenience, to deliver a fixed price result for physical oil. This is the number that appears on the invoice, i.e. X+Z.

Over time the benchmarks have altered out of all recognition. Production of some benchmark grades of oil has declined, trade routes and delivery points have transformed in response to legislation, sanctions, shifting consumer demand and improvements in infrastructure and, particularly for refined products, tightening environmental regulations. Typical efficient cargo sizes (and therefore freight costs) have also evolved. If the characteristics of a benchmark, X, change, the formula price differential for physical contracts, Z, also has to change.

Often those negotiating physical contracts do not appreciate the, sometimes subtle, changes that are happening to the benchmarks and therefore fail to recognise that the formula price differential that has applied to their physical oil in the past has now to be adjusted for future cargoes.

The logical response to changes in the level of production of benchmark grades and therefore the number of transactions informing the published benchmark prices, might have been to allow the benchmarks to evolve too, some disappearing and new ones growing. If that had happened physical oil contract price formulae would have contained new benchmarks with price averaging periods and price differentials for physical grades reflecting departures from the new benchmark characteristics.

But that is not what has happened, particularly for crude oil benchmarks. Instead, the market has gone to extraordinary lengths in some cases to preserve the semblance of old benchmarks, such as Brent, which is discussed in some detail below. This benchmark has included more and increasingly disparate grades of oil from different locations into the benchmark price discovery database to shore up the old benchmark.

If a party to a physical oil contract does not understand what a changed benchmark, X, now represents and how it is compiled, how can it hope to agree a fair price differential, Z, in a physical cargo oil price formula?

Deal Data Underlying Benchmark Quotations

Often it is not actual deals negotiated and finalised in a contract that form the benchmark database for each grade of oil. It is frequently just price indications to buy or sell that do not result in an actual deal. And those deals and indications are those submitted to a price reporting agency (“PRA”) during a specific 15-minute period on an electronic platform each day. Only those companies that have been pre-approved by the PRA are permitted to input price indications to the PRA’s electronic platform.

The PRAs are at pains to consider other transactions not input directly to the electronic platform in arriving at a final published number. It is my understanding that the PRA considers all the information at its disposal, not just electronic platform data, and normalises non-standard transactions onto the same basis as the characteristics of its benchmark when it has sufficient data to do so.

What could possibly go wrong?

Basket Cases

A significant level of complexity arises when a price benchmark reflects the price of, not one specific grade of oil, but a basket of similar grades of oil, any of which can be delivered in satisfaction of a contract.

This is a predominantly crude oil phenomenon, rather than for refined products. This is because refined products are typically defined by their quality specifications rather than by a “brand name” that defines their origin. For refined products, baskets usually only arise for delivered (CFR, CIF, DAP etc.) transactions where the freight component of the price may involve a basket of possible origins for product delivered to the delivery point specified in the benchmark.

In the case of crude oil, a distinction must be drawn between a blend and a basket.

Many grades of crude oil are actually a blend of oil arising from many different oil fields that are gathered in a common pipeline or that are commingled and use shared storage and export facilities. The quality of blends can vary depending on the relative production rates of the different fields contributing to the blend.

Baskets are something different. Baskets are employed when a range of different crudes from different origins can be delivered in satisfaction of a contractual commitment. Examples of baskets include:

The Brent Basket

In the case of the ubiquitous “Brent” benchmark, by now most actors in the oil industry have cottoned on to the fact that the price of Brent as published by the PRA, Platts, has very little to do with the Brent of old. It is now the lowest of the prices for a basket of grades including Brent, Forties, Oseberg, Ekofisk, Troll or, since June 2023, WTI Midland (“BFOETWTIM”).  Yes, West Texas Intermediate gathered at Midland Texas! We’ll come back to WTI later in this blog.

Quality

Forties, Oseberg, Ekofisk and Troll are of different quality to the Brent and Ninian System crude oils, loaded at Sullom Voe in the Shetlands. Brent and Ninian System crude oils reach landfall through two separate pipelines, but these were commingled in onshore tank to become Brent Blend from 1990 onwards. Some of the PRAs have chosen to call this Brent/Ninian Blend (“BNB”). To the cognoscenti, “Brent” continues to be the brand name of this benchmark basket, although Brent now contributes very little to the benchmark volume.

When compiling the daily benchmark price quotation, it is expected by Platts that Platts’ monthly sulphur price de-escalator is applied to the reported price of Forties by the two parties to the deal. This is done when the precise sulphur content of each specific cargo is determined at the loading port. Platts assumes a maximum of 0.6% Sulphur content (“S”) in its assessment of the price of Forties and the de-escalator is expected by Platts to be applied by the parties to the contract for every 0.1% above 0.6% S that is actually measured at the loadport.

The Platts benchmark standard assumes that Platts’ monthly quality premia (“QPs”) apply to the prices of Oseberg, Ekofisk and Troll sold FOB their loading ports. QP’s for Oseberg, Ekofisk and Troll are derived by Platts from, currently, 60% of the prevailing market differentials between each of these three grades and Dated Brent, assessed two months ahead, as expressed by the forward oil price curve, often referred to as the “forward Dated Brent strip”. QP’s are assessed for the current month of publication and the subsequent month.

These QP’s are then subtracted from the observable indication/trade levels for these grades during the reference period so that they can be included in the Dated Brent price assessment on a comparable quality basis as Brent Blend. If Oseberg, Ekofisk or Troll is delivered into a forward cash Brent trade, then the buyer compensates the seller by the QP amount because the buyer is receiving delivery of a higher-valued Oseberg, Ekofisk or Troll barrel.

These premia/de-escalators are determined empirically and announced monthly in advance by Platts. Which precise month’s sulphur de-escalator or QP is used in a contract depends on the deemed B/L date, not the actual B/L date. So which party does the deeming can make a measurable difference to the price for cargo deliveries around the turn of a month.  

The Limitations of the Gross Product Worth Approach to Quality

In the olden days when Brent was Brent and Brent alone, one of the tools used to establish the quality component of the price differential between the Brent benchmark and other grades of oil was to look at the value of the refined products that could be extracted from the physical oil and compare it with the refined products that could be extracted from Brent.

For most, if not all, grades of oil worldwide a refining assay is usually commissioned from a specialised laboratory. This explores under laboratory conditions how the grade in question will perform in a refinery. The assay spells out the quantity and quality of finished and semi-finished products that can be derived from the grade under laboratory conditions. This gives us the starting point for evaluating the Gross Product Worth (“GPW”) of the crude oil. The GPW can be used to clarify the likely value of the new grade relative to existing benchmark grades.

GPWs are based on simple refined products yields and qualities multiplied by the prices of those refined products that can be derived from the crude oil in question:

GPW = Ʃ ((Yield of Product 1 X Price of Product 1) + (Yield of Product 2 X Price of Product 2) + ((Yield of Product 3 X Price of Product 3) etc.)

The GPW of the physical oil is then compared with the GPW of the benchmark oil for which, by definition if it is a benchmark, a transparent market price exists. From this a market price for the physical oil may begin to be inferred.

This approach was never a panacea for several reasons, not least of which is that each refinery and each geographic region extracts different amounts of product from the same crude depending on the equipment of the refinery and consumer demand for end products in the region. Also, each refinery sells into different local markets with their own pricing and taxation architecture. Furthermore, there is more to price than quality. But a GPW comparison is a reasonable starting point to begin price differential negotiations.

However, in the case of the Brent benchmark on any given day the market price may be established by any one of the other grades in the basket, Forties, Oseberg, Ekofisk, Troll or WTI. So, although we know the market price of the benchmark Dated Brent physical benchmark, we do not know if we should be applying a GPW differential between Forties and the grade we are assessing, or Oseberg and the grade in question, or Ekofisk and the grade in question etc.

Usually, this difficulty is addressed by performing a GPW comparison between the physical oil being evaluated and a different, more established, physical oil that is not a basket and for which the market price differential to Brent is known. The market price differential to Dated Brent for the more established physical oil, it is hoped, will have been tested in the market over time.

If the market for that established grade is limited either in volume or in the number of companies trading it, the problem is not necessarily solved. Because inserting the market price differential between two physical grades of oil, one of which is established relative to Dated Brent, into the formula price for the other physical grade could merely be perpetuating an inaccuracy.

Delivery Location and Freight

Platts started assessing the price of the individual Brent basket grades delivered to Rotterdam in 2016. To boost the pool of transactions or indications that form the daily Dated Brent published price assessment of these grades, transactions and indication reported on a delivered CIF Rotterdam basis have been included in the database since 2019. These are adjusted by the PRA to net them back to an FOB North Sea equivalent by a freight adjustment factor (“FAF”). FAF is a 10-day rolling average of the UK to Rotterdam freight assessments, as compiled by the PRA.

Differential to Varying Benchmark Price Averaging Periods

A further adjustment to the CIF price is made to reflect the fact that it takes about 1-2 days to get from the North Sea loading ports to Rotterdam. So, an FOB price refers to a different price averaging period than a CIF price. Platts assumes that, in the case of North Sea grades, contract price formulae use the 5-day average of published prices around the bill of lading (“B/L”) date for physical oil. This is so-called “2-1-2” pricing. In the case of West African grades, the norm is to use the 5-day average of published prices after the B/L.  

If the deals or indications that are reported in the 15-minute assessment period on the electronic platform are on a different basis from the 5-day assumption made by the PRA, the dealer is expected to report this to the PRA so that a further adjustment can be made by the PRA to normalise non-standard deals onto the basis that the PRA recognises.  Only the PRA knows how rigorously the companies provide this information.

Not all actual deals done in the market are transacted on a 5-day average period, particularly in the case of refined products.  Some are done on a whole month average (“WMA”), with the month being determined either by the actual B/L date, a deemed B/L date or the date of the Notice of Readiness (“NOR”) to load or discharge the cargo. Others are done on a balance of month average (“BALMO”).

If the trader reporting to the PRA omits to mention the fact that its deal is not based on a 5-day average, this introduces a skew in the pricing. For example, if a deal for a late date-range cargo in a month is entered into the electronic price discovery platform by a trader without mentioning that the pricing is WMA, and the market is in backwardation, the reported price will be over-stated, and a deduction should be made to the differential that would apply if pricing was 2-1-2. If the market is in contango it will be under-stated, and an addition should be made to the differential that would apply if pricing was 2-1-2.

The Inclusion of WTI in the Brent Basket

In 2023, in response to a continuing decline in the volume of crude oil making up the Brent basket, WTI transactions and indications became part of the database that is used to compile the published price assessment for Brent. This took effect from cargoes in the June 2023 forward cash Brent contract.

With production and exports of US crude becoming a dominant component of the international supply/demand balance, it is only to be expected, and consistent with market forces, that WTI has become an increasingly important component of the absolute price of oil. Hence, while it is true that WTI has brought down the value of Brent because its price is typically lower than that of the other grades in the basket, this is a logical development. There is more WTI being exported, displacing other more local grades in refiners’ historic slates, so it is to be expected that the price of the more traditional grades would decline to compete.

Nevertheless, traders continue to cling slavishly to Brent as their benchmark of choice and PRAs have introduced even more complexity into the Brent price discovery process to maintain Brent’s benchmark status by annexing the WTI export volume into the Brent basket.

If producers, refiners and traders had stopped using Brent as the benchmark component of physical oil contracts, then WTI exported from the US Gulf coast is likely to have taken over naturally as the benchmark of choice. Instead, an appetite remained in the trading community for the dwindling Brent benchmark, which has had to be shored up by shoe-horning the burgeoning volume of WTI exports into the Brent suite of contracts.

What we call “Brent” now predominantly, but not exclusively, reflects WTI delivered to Rotterdam, or price adjusted delivered to other European ports, and netted back to the North Sea. To achieve this, some precise rules have been introduced by the PRAs in their price assessments to make WTI fit into this new Brent model.   

WTI, in particular the CME/NYMEX futures contract, has been a pricing benchmark in its own right since the 1980s. This very active futures contract has always been a pipeline contract deliverable inland in lots of 1,000 barrels (“bbls”) at Cushing, Oklahoma. It is an important benchmark for the US domestic market but, arguably, has little direct relevance to the international market. It continues to have an indirect relevance because some of the exported US domestic grades price as a differential to the WTI futures contract.

Since the repeal of the US congress’ ban on the export of domestic oil in 2015 and the increase in US production from about 13 million b/d in 2015 to about an estimated 20 million b/d in 2024, WTI has taken a different role in world trade and the price discovery process.

It is necessary to be precise about what is meant by WTI. As mentioned above, the futures contract uses WTI delivered to Cushing as the underlying commodity, but WTI is also gathered and traded at Magellan East Houston and at Midland Texas.  Physically, it can be exported from an increasing number of locations along the US Gulf coast including from Corpus Christi and from the Louisiana Offshore Oil Port (“LOOP”).

US export infrastructure was going to have to change to accommodate its growing export volume, but the inclusion of WTI into Brent price assessments has steered developments in favour of smaller cargo sizes.

Cargo Size

It would be most economically efficient to export WTI to Europe or further afield to the Far East in big tankers such as VLCCs, which can transport up to 2.2 million bbls. This would bring down the unit cost of freight: usually the larger the tanker, the lower the per barrel cost of freight.

However, since the inclusion of WTI in the Brent basket, the “Brent” contracts now specify a cargo size of 700,000 bbls. Brent started life trading in the 1980s usually in cargo lots of 500,000 bbls. This was in line with the Sullom Voe terminal lifting agreements between producers and the terminal and pipeline operators, then BP and Shell. This was increased to 600,000 bbls in 2016 when CIF Rotterdam cargoes began to be assessed and published. When WTI became a candidate for the Brent basket, the Brent cargo size was increased to 700,000 bbls. It would have been difficult to increase it further because the North Sea terminals would have found it hard to comply. The North Sea terminals accommodated the increase to 700,000 bbls without too much fuss.

Even with the increased North Sea cargo size, bringing WTI into the Brent basket means that WTI also has to comply with the new 700,000 bbl cargo size, which is sub-optimal for transatlantic voyages. To export a 700,000 bbl cargo, transport would have to be to be in, usually more expensive, Aframax tankers rather than the more logical VLCC choice.

Alternatively, a tanker charterer delivering WTI into the Brent basket might decide to use a VLCC anyway and incur deadfreight costs on a larger tanker by sailing partly laden with only 700,000 bbls onboard. This is unlikely for economic reasons. So, the charterer may have to organise two port discharges or other non-standard contractual arrangements with other shippers, which involve additional costs, to ensure that their cargoes can be delivered into the Brent complex. Platts recognises the price of cargoes that are the subject of such adjustments.

The US Plays Ball

New US export infrastructure has connived at the insertion of WTI into Brent and many, though not all, have sought Platts’ approval to include cargoes from their pipelines and export terminals into the Platts Brent price assessment process. For example, cargoes exported from LOOP are not currently recognised in the Platts Brent basket.

In order to gain such approval for inclusion in the Brent basket, cargoes of WTI Midland oil must meet quality specifications set down by Platts if they wish to be included in the Platts price assessment. These specifications are intended to be typical of US Permian basin crude and are quite different from the quality of the WTI that is delivered to Cushing Oklahoma under the CME/NYMEX WTI futures contract.

Once the WTI price is assessed by the PRAs delivered to Rotterdam, it is similarly netted back to an FOB North Sea basis in the same way as CIF cargoes of the other basket grades. A commensurate adjustment is made to the 2-1-2 price averaging period around a deemed B/l. This deemed B/L reflects a 1 to 2 days sailing time from the North Sea to the continent.

Rolling Schedules versus Monthly Schedules

Platts assumes that it takes 17 days to cross the Atlantic from the US Gulf coast to Rotterdam. That is the case for a speed of about 12.5 knots, but the time can be pared down to 15 days at 14 knots or dragged out to 21 days at the environmentally friendlier rate of 10 knots.

This introduces a further complication. The North Sea has historically scheduled its cargo liftings from the various shared terminals on a monthly cycle. WTI has hitherto operated on a rolling basis with shipments, or “liftings”, agreed ad hoc between producers and the various pipeline/terminal operators.

In the North Sea the producers nominate their preferred cargo lifting dates in month M (“M”) to the terminal operator usually by about 20-25th M-2. The allocation of cargoes to individual lifters by the terminal operator is determined before the last day of M-2. This process has become earlier and earlier over the years to fit with the market’s need to comply with the notice period that a seller must give a buyer before delivering a forward cash Brent contract.

It should be recalled that the Brent forward contract, sometimes referred to as “cash” Brent, allows a buyer and seller to trade a Brent cargo for delivery at any time during a specified future month up to several years ahead. The seller must tell the buyer which of the grades in the Brent basket - Brent, Forties, Oseberg, Ekofisk, Troll or WTI- it will deliver and on which 3-day delivery date range in the specified month it will deliver it, by 30 days before the first day of that particular delivery date range being supplied. So, to ensure that the first cargo in month M can be delivered into the forward contract for month M, the schedule of liftings must be known 30 days before 1st-3rd Month M.

This timetable does not sit easily with the rolling date ranges employed, and historically kept confidential, by US Gulf Coast terminals, particularly since cargoes exported in month M are about 17 days away from Europe. It is therefore very difficult to track how many cargoes of all grades are potentially deliverable into the Brent cash contract. This gives traders considerable flexibility to speed up or slow down cargoes coming from the US Gulf to allow them to be delivered into an earlier or later month cash Brent contract. It also makes the cash contract difficult to “squeeze”, as it was in the past, when certain companies bought up the whole supply of cargoes that qualified for delivery into the cash contract.

So What?

The complexities of the oil price formation process described above have emerged over time as a result of consultations between the PRAs and the trading community, including the trading divisions of major oil producers and refiners.

It should be stressed that oil pricing is complex because traders and trading departments have agreed to have it that way by default. The trading community did not have to agree to benchmark changes as they became more convoluted and more dependent on the ability of PRAs to normalise the price of the deals, or the buy/sell indications they see, onto a standardised basis. They do not have to use PRA or futures contract data in their price formulae at all. But they do. They bend over backwards to accommodate the PRA price assessment process as it changes over time.

Obviously, between two consenting companies trading any cargo, they can enter into any type of agreement or pricing mechanism that suits them both. But, in the case of Brent, if either or both companies wish a deal to qualify for delivery into the forward cash contract, or in the case of other benchmarks, be included in the PRA assessment database, the deal must conform to the reporting standards of the PRAs.

There are very few oil industry executives who are not directly involved in trading who understand the intricacies of the evolving benchmarks. This is true for all benchmarks, not just Brent. If the benchmark in a physical oil price formula is not understood, then the differential to the benchmark agreed in the formula price of physical non-benchmark cargoes may not be a true reflection of the market for that grade of oil. 

For the fleet of foot trader, complexity provides opportunities to make money, or to make themselves look good to their uninformed non-trading colleagues. For example, an apparently more favourable differential for the same cargo can be established by using a different benchmark, or a different price averaging period, or by moving the delivery date range back or forward, particularly between months, in a market showing backwardation or contango.

In the case of the big companies, their loyalty to existing benchmark labels is partly because there is a whole suite of price management tools – futures, forwards, swaps and options- that are based on these benchmark brand names, which they can use to manage their risk or trade for profit.

In the case of small producers and refiners without their own trading departments, it is very difficult to make a meaningful contribution to the consultation process when benchmarks or benchmark changes are being debated. Most small companies simply use the existing benchmarks because their peer group companies do so.

It is dangerous for these smaller companies to simply roll over annual or other long-term sale and purchase contracts on the same basis as they did last time without checking if the benchmark, let alone the market, has changed in the meantime.  If it has, the price differential to the benchmark in the formula price has to be reviewed.

Coping with Complexity

Consilience’s advice would be, when in doubt always run a tender. It is advisable, when possible, to give as wide a range of traders and oil companies as achievable the opportunity to compete to bid or offer the most favourable differential to a consistent benchmark formula. The call for tenders absolutely has to be consistent or else the differentials submitted in response are not comparable.

If a responder to a tender suggests that any form of optionality in the delivery dates or the price averaging period should be inserted into the price formula, then we enter new and even more complex territory. That will be the subject of a subsequent blog.

The only way to ensure that small producers or refiners, usually referred to as “price takers”, are getting the best price for their shareholders is for them to award the tender, always to a reliable company, that gives the most favourable differential to a price formula that is identical in every other respect.

This approach is probably more realistic than trying to understand and keep up to date with constantly changing benchmarks

“Nothing in life is to be feared, it is only to be understood. Now is the time to understand more, so that we may fear less”: Marie Curie
Yeah! And look what happened to her: Liz Bossley 

Gulf Intelligence – Daily Energy Markets Podcast

Liz Bossley and the other guest panelists, Maleecha Bengali and Jose Chalhoub, are discussing the direction of travel in the Asian Oil Markets. Host: Sean Evers.

To watch the interview CLICK HERE

Twitter

https://twitter.com/gulf_intel/status/1704020284410667343

Gulf Intelligence Half Time Talk

Liz Bossley, alongside Kurt Chapman of Zenith and Neil Fleming of Argus, is interviewed by Sean Evers of GI about the new Palgrave MacMillan Book “Brent Crude Oil Genesis and Development of the World’s Most Important Oil Benchmark” [Print ISBN 978-3-031-28231-7; (eBook) ISBN 978-3-031-28232-4. Each interviewee authored one chapter in this new textbook.

To watch the interview CLICK HERE

LinkedIn

https://bit.ly/3C1Xhyn

https://bit.ly/3C47muJ

https://bit.ly/43C6327

Twitter

https://twitter.com/gulf_intel/status/1663818321253236736

https://twitter.com/gulf_intel/status/1663794942559522816

https://twitter.com/gulf_intel/status/1663778306792079361

YouTube

https://www.youtube.com/watch?v=QNUvcIxDg3s

Developing Emissions Trading Schemes

Liz Bossley, with Cui Ying, Deputy Director General of the International Institute Of Green Finance and Adrian Rimmer, Director of Finance LSEG present their views to Michael Mainelli of  ZYen on the subject of emissions trading schemes worldwide.

To watch the interview CLICK HERE

For a copy of Liz’s slides entitled “Investing in the Real World” CLICK HERE