Quants puzzle over how to handle negative oil prices

Firms are choosing to cut ‘outlier’ prices from data or to rely more on fundamental inputs


When the oil price went negative last month, quants were confronted with something their models had never before seen. Some firms have treated the episode as an outlier that can be removed from the data. Others are falling back on fundamental analysis for a steer on whether the same could happen again.

Treating the event as a one-off avoids the risk of updating strategies in ways that curb their effectiveness in normal conditions. However, dismissing the oil plunge as a freak episode risks overlooking the possibility of a repeat in future, which some see as possible.

“We have to be on the lookout for market conditions that were not considered carefully when we built our models,” says Roberto Croce, head of risk parity and liquid alts and portfolio manager for the risk parity and managed futures strategies at Mellon Investments.

West Texas Intermediate crude futures for May delivery traded on the CME settled at -$37.63 a barrel on April 20 due to storage issues at Cushing, Oklahoma, the delivery point for the WTI contract.

The price fall left some investors with losses perhaps as high as a billion dollars.

When events like this happen quants have to “step up” and take discretionary decisions, Croce says, even though Mellon is a purely systematic firm.

“If you were holding the June contract in crude oil and the May contract went to -$50 because storage for the physical commodity is all full, then you are sitting in a contract that [could be] the next in line to get squeezed,” says Croce.

“In this case, your model has no way of seeing that the prior month’s contract became much riskier approaching expiry, nor of observing the amount of available storage; you might use discretion to decide to roll out to December, to take the position off entirely if it’s not material to the strategy, or to trim it based on higher risk that you’re seeing. Those are all reasonable responses.”

Mellon itself was not invested in the May contract, he says.

Quants that chose to remove negative prices from the data have assumed the impact of the episode on specific strategies would be small.

Michael Heldmann, who heads multi-factor equity investing for North America at Allianz Global Investors in San Francisco, says the negative oil price had limited repercussions in equities – at least in the short term.

“We have removed those outliers from the data on a discretionary basis for fitting [models]. I think that’s absolutely warranted,” Heldmann says.

The firm’s alternative risk premia funds include in their models the susceptibility of stocks to oil moves.

“We have never been shy of modifying the way we see the world and adjusting our models to what’s going on in the market, especially removing really big outliers,” Heldmann says. “For the oil price, I cannot imagine a repeat of this situation in the future. It was a very specific situation with a contract expiring and the lack of storage facilities at that point in time when the oil market was already in huge distress.”

May demand looks a little bit better with the economy opening up. But no one really knows when it’s going to bounce back
Brent Belote, Cayler Capital

Elsewhere, Christian Kjær, head of liquid markets at Danish pension scheme ATP, says the organisation is exposed to oil from its beta exposure as well as its cross asset trend and carry strategies, but since the models are “not built” to make predictions, the group is getting out of positions entirely.

“When you look at these projections of the storage capacity, it could be an issue again. The models are not built to be able to predict [such capacity concerns], so we just feel safer leaving a relatively small part of the universe out [of the portfolio] for this short period of time,” he says.

Kjær is also concerned about whether oil can be relied on as a diversifier going forward.

Some firms are relying on both technical and fundamental data to alert them to changes in market behaviour that may have emerged since last month’s fall.

Acadian Asset Management did not hold the May contract, but negative prices have affected pricing on longer-dated oil futures also, says Michael Ponikiewicz, vice-president and portfolio manager.

Investors are now looking to roll contracts earlier lest storage constraints recur when the June contract expires, he says. To form a view on oil going forward, Acadian is using real-time pricing from futures curves along with data on supply and demand.

“We do have more fundamental supply and demand type of information and inventory information in the in the petroleum models,” Ponikiewicz says. “If that confirms what the futures curve is pricing in, that is a pretty robust signal.”

Acadian is also of the view that a negative oil price is unlikely to happen again, but notes the supply/demand imbalance.

“I think it’s less likely this time, largely because the smaller speculators that were there are now aware that if you’re holding that close to delivery, logically you would not want to take that risk again. But that said the supply and demand balance is still the worst ever.”

Acadian is also looking at government data commonly used by fundamental oil analysts to measure supply and demand. “If that’s the best source, then we’ll use that,” he says.

Relying more heavily on fundamental signals has its challenges, though, not least because of uncertainty around the Covid-19 epidemic.

“April demand has been absolutely horrendous,” says Brent Belote, chief executive and portfolio manager of Cayler Capital, a systematic fund that trades commodities. “May demand looks a little bit better with the economy opening up. But no one really knows when it’s going to bounce back. We’ve lost 70% of global jet fuel demand, and even if airlines open up: how many people are going to come back and fly? Usually demand is fairly static and you have an idea of what it’s going to look like.”

Cayler exclusively uses fundamental data – such as field-level production, import and exports from the gulf, pipeline flows, refinery economics and refinery turnarounds – to assess demand.

  • LinkedIn  
  • Save this article
  • Print this page  

Only users who have a paid subscription or are part of a corporate subscription are able to print or copy content.

To access these options, along with all other subscription benefits, please contact [email protected] or view our subscription options here: http://subscriptions.risk.net/subscribe

You are currently unable to copy this content. Please contact [email protected] to find out more.

You need to sign in to use this feature. If you don’t have a Risk.net account, please register for a trial.

Sign in
You are currently on corporate access.

To use this feature you will need an individual account. If you have one already please sign in.

Sign in.

Alternatively you can request an individual account here: