Prof. Jayanth R. Varma's Financial Markets Blog

About me       Latest Posts       Posts by Year       Posts by Categories

WTI crude futures in India

India’s commodity derivatives exchange, MCX trades crude oil contracts that mirror the WTI Futures contract traded at CME/NYMEX in the US. When the US contract settled at an unprecedented negative price this week (the seller had to pay the buyer to take their crude away), the Indian contract followed suit. Press reports state that brokers who had bought MCX crude futures suffered a loss of Rs 4.35 billion and have gone to court against the exchange’s decision to settle at a negative price.

I like to think that I have better things to do than take sides in this fight, but I also think that everybody involved in the Indian crude futures market has behaved recklessly. Since around mid-March, it has been clear that WTI crude in the US was experiencing extreme dislocation, and that highly perverse outcomes were likely, though nobody could have predicted the precise outcome. Prudent traders should have stopped trading the MCX crude oil futures in late March, and a prudent derivative exchange should have suspended trading in the contract in early April. The Securities and Exchange Board of India (SEBI) is currently overburdened with keeping the markets functional during Covid-19, but otherwise, they should have forced MCX to suspend the contract.

Unfortunately the Indian commodity derivatives ecosystem is borrowed lock, stock and barrel from the equity derivative ecosystem. Everybody thinks that a commodity is just another stock price ticker (to trade), another stock price chart (to do technical analysis) and another time series of prices (to compute VaR margins). People tend to forget that commodities are intensely physical, and, unlike stocks, do not come with limited liability. However much we may try to “financialize” and “virtualize” the commodity, its “physicality” never really goes away.

This will therefore be a long post discussing the gory details of crude oil (and WTI crude futures in particular) to explain why I think all parties involved in the MCX crude oil futures behaved recklessly this month.

Economic rationale for Indian crude oil futures

The biggest difference between a futures market and matka gambling is that the futures contract has an economic rationale: it helps economic agents to hedge risks that they are exposed to. Many entities in India are exposed to energy price risk and crude oil futures are useful to hedge that risk. Of course, the crude basket that India imports is closer to Brent crude than to WTI, but then MCX has an MOU with CME that gives them access to CME/NYMEX contracts, and that obviously determined their choice.

So long as WTI crude is highly correlated with Brent (and the Indian crude basket), the choice of underlying does not matter too much. In normal times, the correlation is high and the Indian crude oil futures serves a valuable hedging function. In abnormal times, this correlation can break down and then MCX/NYMEX WTI crude futures would cease to have an economic rationale, and its continued trading would become questionable.

Logistical constraints on WTI crude

Historically, it has been observed that there are occasions when logistical constraints create a big divergence between (a) Brent and the Indian crude basket and (b) WTI. The best known example of this was in 2011 when rising production of shale oil led to a glut of crude at Cushing, Oklahama (the delivery location for the WTI futures contract). At the peak of the dislocation in 2011, WTI crude fell to a discount of around $20 to Brent crude (in the pre-shale era, WTI traded at a premium reflecting its sweetness and lightness).

The critical difference between the two major global crude benchmarks is that Brent is a waterborne crude while WTI is a pipeline-delivered crude. Market forces will move waterborne crude from a region of excess supply to one of excess demand. There are relatively few constraints and frictions in this process of market equilibrium. Pipelines are much more rigid: they have limited capacity and fixed endpoints. A pipeline from point A to B is useless if stuff needs to be moved from A to C or from D to B. Even if the movement required is from A to B, the quantity might exceed the capacity of the pipeline and it would then of limited utility. The sea by contrast has practically unlimited transportation capacity and its directional preferences (winds and ocean currents) can be largely ignored in modern times.

The 2011 experience shows that when logistical constraints arise in landlocked WTI crude, its price diverges not just from Brent, but also from prices in the rest of the world, and indeed even from prices in the rest of the US. In the last decade, benchmarks like LLS and ASCI based on crude prices in the coastal US (Louisiana) have risen in importance. The other observation from 2011 was that refined petroleum products in the US tend to track Brent crude better than they track WTI crude when logistical constraints emerge on WTI. All of this means that WTI starts losing its economic rationale as a hedging instrument for Indian entities in such situations. The behaviour of Indian players in April 2020 needs to be evaluated against this background.

WTI Futures Contract

The WTI futures contract is physically settled: all positions outstanding at the expiry of the contract have to give/take delivery of WTI crude at Cushing, Oklahama which is a major pipeline hub of the US. The delivery procedure at CME/NYMEX reflects the rigidity of pipeline logistics:

Rolling out of WTI futures before expiry

What this means is that the only people who can afford to hold the May futures at expiry would be those who have lined up the pipelines and storage facilities at Cushing, Oklahama. Everybody else should trade out of the contract on or before the last trading session (either by closing the position or by rolling it into the next month contract). In fact, it is very risky to wait till the last trading session to exit the contract. If a buyer is trying to trade out of the contract, and other players know or suspect that the buyer does not have access to a pipeline/storage to take delivery of the crude, the sellers will try to take advantage of his predicament. The buyer’s only hope is to find a seller who does not have the crude to deliver and is equally eager to trade out of the contract. This means that prudent traders who want to square out should do so several days in advance when there is a lot of trading by hedgers and speculators who are not physical players in Cushing.

For example, USO (United States Oil), which is the largest crude oil ETF, typically starts rolling out of a contract two weeks before expiry and completely exits it one week before expiry. In fact, both the exchange and the regulator monitor large positions in the near month contract and encourage traders to reduce positions. They are much more relaxed about positions in the next month or more distant months.

The situation in March/April 2020

Early in March, when crude prices were falling due to Covid-19, Russia and Saudi Arabia held talks on cutting output to stabilize the price. When these talks failed, the Saudis (who are among the lowest cost producers in the world) responded by increasing output to crash prices and remind other oil producers of the perils of not cooperating with Saudi Arabia. In the meantime, demand collapsed as the Covid-19 situation worsened and there was a massive glut of oil worldwide. Prices fell far below what even the Saudis had anticipated, but all players hoped that the demand would bounce back as and when Covid-19 lockdowns were relaxed. The natural response was to store cheap oil so that they could be drawn down when prices rose in future.

By mid/late March, concerns were mounting that storage in Cushing was getting full. On March 19, 2020, Izabella Kaminska wrote in the Financial Times’ Alphaville blog that “in a scenario where there’s literally nowhere to put oil, it’s not inconceivable prices could go negative.” She also explained that some oilfields simply cannot be turned off during a short term glut: “temporary shutdowns pose the risk of them never being able to be revived at the same rates again.” So they would keep pumping crude even at negative prices. On March 27, 2020, Bloomberg reported that “In an obscure corner of the American physical oil market, crude prices have turned negative – producers are actually paying consumers to take away the black stuff.” The price was only 19 cents negative, and the grade of crude was a heavy oil that fetched only around $40 a barrel at the beginning of the year when WTI traded at around $60 (in April, this grade went much more negative). The importance of this Bloomberg report was that it confirmed that negative oil prices were not merely theoretical speculation.

Commodities do not come with limited liability

Many people find it counter intuitive that a commodity can have a negative price. However, the assumption of “free disposal” which is beloved of economics text book writers is not valid in the real world, and there are many examples of negative prices for commodities that are normally quite valuable. Before refrigeration became commercially available, it was quite common for fishermen to pay farmers to collect the day’s unsold fish catch and use it as manure. To understand the plausibility of the negative price, you need only imagine a bunch of fishermen trying to catch a night’s sleep with a boatload of rotting fish just outside their huts. Similarly, I have been told that restaurants often pay pig farmers to collect the waste food at the end of day and feed it to their pigs. On a more sombre note, the Bhopal gas tragedy was an unintended “free disposal” of a hazardous substance. The ultimate negative price of that “free disposal” bankrupted the company.

Please remember that crude oil is also a hazardous substance. Almost everywhere in the world, you would need an explosive licence to stock any significant quantity of it. Anybody who has seen images of oil spills at sea and the damage that it does to sea beaches knows that crude is an ugly and foul thing. Paying somebody to take this stuff from you is not at all unreasonable.

Physical versus Cash Settlement

Some people seem to think that the problems of WTI arise from the fact that it uses physical settlement while the Brent futures uses cash settlement. I have been arguing for more than 15 years that apart from transaction costs, there is no difference between cash settlement and physical settlement. The key difference between Brent and WTI is not in the futures market but in the spot market: one is waterborne and the other is pipeline-delivered. (The Brent physical market is in some ways even more out of reach of ordinary hedgers and speculators than WTI: the typical delivery is a ship load of 600,000 barrels. I described the Brent market in gory detail more than a decade ago and I am not masochistic enough to try and summarize that again.)

What about the theory that the Indian MCX futures is cash settled and therefore should not be subject to the travails of the NYMEX physically settled contract. This reminds me of the story about Medusa in Greek mythology: anybody who looked at Medusa would be turned into stone, but Perseus was able to slay her while looking at her reflection in the mirror. Myth is probably the most charitable word to describe the “Medusa” theory that the NYMEX contract was dangerous, but its reflection in the MCX mirror was safe.

What should Indians have done in early/mid April

First of all, anybody who was actually trying to hedge energy price should have run away from the May WTI futures contract as it was crystal clear that it would no longer provide any meaningful hedge of crude price risk or energy price risk in India.

Second, anybody who was in this contract purely as a speculator needed to understand that in early April, the WTI future was no longer a bet on crude; it was purely and simply a bet on storage space in Cushing. If storage remained available at Cushing, then the May future price could not fall too low. The floor for the price was the expected post Covid-lockdown price (proxied by the July/August futures prices) less the cost of storing crude for a few months. In mid April, this floor might have been estimated at around $15 a barrel. On the other hand, if storage got full, there was no floor on the futures price at all. The price could go negative, and if you did not exit the contract in time, the potential for a hugely negative price was clear as daylight. I would put it this way: when you went long WTI May futures in mid April, you were actually shorting Cushing storage space. Unless you had the discipline, attitude and nerves of a short seller, you should again have run away from this contract.

Third, any prudent broker should have stopped allowing their retail clients to trade this contract purely for risk management reasons. The only clients to whom this contract should have been made available were those with deep pockets and known integrity who could be counted on to pay up when things go wrong. Please remember that when prices can go from positive to negative, even 100% margins are inadequate as the loss exceeds the notional value of the position. This is another way of saying that the long crude position is actually a short storage position and there is no limit to the losses of a short position.

Fourth, if the exchange observed a sizeable open interest in this contract in early/mid April, it should have realized that market participants were ignoring one or more of the above three prudential principles. If market participants are reckless, they pose a risk to the exchange if they are unable to meet their obligations. Also, as mentioned above even 100% margins do not cover the worst case risk in this situation. Faced with this problem, I think the exchange should have done two things:

Posted at 9:10 pm IST on Thu, 23 Apr 2020         permanent link

Categories: arbitrage, commodities, derivatives, exchanges

Comments

Comments