LONDON (Reuters) - Price volatility is the lifeblood of commodity futures exchanges. But the amount of volatility has to be just right.
Too little and there is no incentive to trade. Too much and trading becomes dangerous, prompting traders to give future contracts a wide berth.
The extreme volatility that occurred in U.S. light sweet crude futures for May delivery on Monday, the day before the contract expired, is likely to become a textbook example of too much volatility.
Oil traders have already started to avoid the June futures contract for fear of similar wild price moves next month.
The volume of futures contracts still open for delivery in June has fallen by the equivalent of 217 million barrels, more than a third, since Monday.
Unusually, the front-month June futures contract now has less open interest (364 million barrels) than the second-month July contract (385 million barrels).
Open interest is normally concentrated in the front-month contract until a week or so before expiry, by which time most positions have been rolled forward into the second month.
In this case, traders have started to abandon the front-month for the second-month with more than three weeks left. In fact, the number of open contracts is falling across the next four months as a whole; decreasing open interest for June is not being fully offset by increased interest in the July, August and September contracts.
Total open interest in the four contracts from June to September has fallen by 170 million barrels, or 13%, since Monday and some brokers are restricting clients from initiating new positions in the June and July contracts because the risk is too great to manage.
Most criticism about what occurred in light sweet crude oil futures, commonly known as West Texas Intermediate (WTI), has focused on the plunge into negative territory for the first time in history. But that is the wrong focus.
Negative prices are actually fairly common for commodities that cannot be stored (such as electricity) or are expensive to store because they require specialist facilities (natural gas).
So it should be no surprise in that oil turned negative, given that consumption has collapsed and increasing volumes of crude need to be stored in specialist tank farms.
Monday’s settlement price of almost -$38 a barrel for the May contract reflected lack of available tank space at Cushing in Oklahoma and the resulting high cost of storage.
The Chicago Mercantile Exchange (CME) has vigorously defended the performance of its futures contract and said that the negative price reflected storage conditions.
“CME Group markets worked as designed,” the exchange said in a statement emailed to Reuters on Thursday.
“Our futures prices reflect fundamentals in the physical crude oil market, driven by the unprecedented global impact of the coronavirus, including decreased demand for crude, global oversupply and high levels of U.S. storage utilisation.”
The exchange said that it had given advance notice to its regulator and the marketplace that it would accommodate negative futures prices so that clients could manage their risk amid dramatic price moves while also ensuring the convergence of futures and cash prices.
The real concern about the contract’s price move was not that prices went negative but that they dropped by so much, so quickly. Between settlement on Friday and settlement on Monday the price plunged by almost $56 a barrel, representing a decline of more than 300%.
In percentage terms, the one-day price move was equivalent to 67 standard deviations for all daily price moves since 1990.
By contrast, the front-month Brent futures contract declined by only 9% between Friday and Monday, equivalent to four standard deviations.
Brent’s daily move was large but not abnormal. WTI’s move was so extraordinary it would not have been foreseen by any realistic risk-management system.
No trader can protect themselves against such moves, other than by not running a position at all.
The decision of some brokers to restrict some clients from opening new positions in the June and July contracts reflects that extreme level of risk.
The price crash occurred on the penultimate day of trading in the May contract, on a relatively small volume of transactions, when most positions had already been rolled into June.
Traders had relatively few positions left in the soon-to-expire May contract so the number of market participants directly affected by the price plunge was fairly small.
Prices recovered most of their losses the following day before the contract expired, but the heightened concern remains. No other monthly contract for either WTI or Brent showed a similar decline, with all the volatility concentrated in May WTI futures.
In the continuous price history of front-month WTI and Brent futures, the settlement of the front-month WTI contract on Monday is a clear aberration.
Prices for May WTI experienced by far their largest one-day decline in history despite the absence of any obvious new information about production, consumption or inventories.
Traders have expressed concerns about tank farms becoming full, especially around the contract’s delivery location at Cushing, but those worries had been around for several weeks.
If the sell-off was sparked by concern over storage running out, why did prices for the May contract rebound the next day, and why have prices for the June contract been rising since Wednesday?
The nosedive in prices shares many characteristics with a flash crash rather than a move driven by fundamentals – and it occurred on thin volumes shortly before expiry, when contracts are vulnerable to dislocation.
In defence of the futures contract, prices on Monday may have reflected trader assessments of the real supply, demand and storage fundamentals for oil at a specific time (May) and location (Cushing).
But most traders do not intend to make or take delivery of physical crude at Cushing. They use the contract as a benchmark for crude prices more generally in the United States.
The contract’s usefulness depends on most participants employing it as a reference for purely financial transactions or physical transactions at multiple locations.
If the contract reflects only conditions at Cushing, and Cushing becomes delinked from conditions in the rest of the country, the contract ceases to be a useful benchmark.
CME has staunchly defended the performance of its WTI contract. But users may have several questions:
- Would CME be comfortable if the expiry of the June contract next month was marked by the same volatility?
- How will the contract’s physical delivery mechanism work if tank farms at Cushing become full in the next few weeks?
- Can WTI futures function as an effective benchmark for hedging and trading crude if Cushing prices dislocate from the rest of the country?
- How should most market users protect themselves from extreme volatility in near-dated WTI contracts, other than by not using them at all?
- What special measures, if any, will CME take to ensure contracts trade in an orderly manner through to expiry?
John Kemp is a Reuters market analyst. The views expressed are his own.
Editing by David Goodman