Paris, France - July 22, 2020
Written by The Commodities Team at Murex
On a baseline, human level, the reality of coronavirus has the Murex team dealing with a “new normal.” This means a (sigh) heavy stream of virtual meetings across multiple platforms, digital (sigh) drinks and dinners. Those of us who hate quiz nights have found this moment ... trying.
In commodities trading, meanwhile, an unthinkable was spurred by coronavirus: Oil futures prices settled in the negative territory. The MX.3 Commodity Solution managed the switch to a valuation model built to anticipate negative prices, a potentially important adaptation in a very uncertain period.
Back to basics...
In the real physical world, commodity prices can become negative. This phenomenon is largely seen in the spot market, triggered by an acute disequilibrium between supply and demand. The pandemic-impacted global economy has caused crude oil demand to plunge into an abyss. In April, supply was still abundant. Oil storage capacity was literally packed in every imaginable monetizable crude storage facility.
Pipeline giant Energy Transfer had requested permission to idle two pipelines in Texas and convert them to storage for around 2 million barrels of oil, according to Forbes.
With all these cocktail ingredients in the shaker, no commodities expertise is needed to predict that the crude oil price is under extreme downward pressure with high volatility.
Intense fluctuations and a very rocky ride
The crude futures market has been used by the physical market players for pricing. And the WTI contract is widely used for this purpose. Drilling into Chicago Mercantile Exchange (CME) viscous data, on the penultimate day of WTI May-20 futures (April 20), at 2:08 p.m. New York time, the contract was trading “freely” (~0 $/bbl), before sliding into negative territory within 20 minutes and settling at a historic –37.63 $/bbl 2 minutes later. This first-time event came to most as a complete shock. On April 20, this contract’s final trading day, it settled at ~10 $/bbl. It was a very gnarly ride for crude traders, indeed.
The fakes and the facts
Did the physical market really react so promptly to swing from zero to available crude storage spaces within such a short period of time to absorb undesired oil? There was an acute mismatch of supply and demand, but only on the paper oil market. United States Oil ETC (USO), a fund allowing investors to be exposed to the oil market without really going into the physical market, rolls their WTI front-month positions every month. According to Bloomberg, the week before the event, USO owned 25 percent of the outstanding volume of May-20 Futures. A market share of 25 percent is daunting. Positions needed to be rolled over. In plain English, sell May-20 and buy Jun-20 Futures (and/or Jul-20, etc.). It’s unknown when did they sell (stretched over how many days). But big selling was definitely in the pipeline.
A deep liquidity gap
CME data showed that May-20 open interest was fairly large on April 20. Apart from being the penultimate trading day for May-20 Futures, April 20 was the last trading day for WTI Calendar Spread Option (CSO) for May-Jun where, according to Risk.net, the open interest in put options was large, as well. These put options were deep in-the-money that the generated short May-20 Futures simply offset the traders’ outstanding long futures (bashful traders rarely trade naked options). Many players did not have to trade much.
The liquidity gap became obvious when desperate sellers pushed until the trading limit mechanism kicked in at around 2 p.m. New York time. (This trading mechanism is factored in by MX.3, by the way.) Those who did not close their positions had to close them (approximately 10,000 contracts) in the open market before closing.
The protagonists unwittingly became mere spectators experiencing the unfolding events until the clock struck 2:30 p.m. A black swan—negative oil—was born.
One of the direct consequences of this event is on option pricing and valuation. Traditionally, the Black-Scholes (BS) model has enjoyed market dominance. The main assumption of this model is that underlying prices of stocks or commodities follow a lognormal distribution.
In layman’s terms, a commodity price can never go negative in the BS model. Whoops.
Two approaches from MX.3 clients
The simple, closed-formula Bachelier model allows for the possibility that prices can go negative. The Chicago Mercantile Exchange and the InterContinental Exchange have decided to switch to this model for valuation under certain conditions to calculate the margin requirements for the listed commodity options. One could reasonably wonder the hassle of switching valuation model in a big organization. The MX.3 Commodity Solution handles the switch from BS to Bachelier for pricing and valuation.
Our clients can be broadly classified into two categories: those who demand adaptability whenever Bachelier valuation is required, and those who believe oil going negative was an aberrant thing—a freak occurrence. OPEC+ decided to cut supply, after all, and crude prices have been since trending in the positive territory. So can commodity prices go negative? It could have been worse, an optimist might say. A pessimist would say things could always be worse. No one knows. It is probably best to be prepared.