Why oil forecasting is so difficult

Every year, a group of oil experts, analysts and aficionados on Twitter have a contest to predict the oil price. This year, I was fortunate enough to emerge as the winner. But I’m not here to crow about my foresight – perhaps a little – but to consider why oil price forecasting is so difficult, and about to get harder.

The friendly competition, managed by Leslie Hayward of US consultancy Rapidan, is assessed on the Brent crude closing price on December 31. With forecasts ranging from $28.75 to $78.37 per barrel, my guesstimate of $65.99 was a single cent below the actual figure. I had anticipated fairly effective US sanctions on Iran, likely geopolitical upsets, and continuing collapse in Venezuela, though I was more pessimistic on the world economy than transpired.

Oil markets last year saw numerous surprises. Perhaps the most surprising of all, though, was how stable prices were. Prices going into 2019 had touched a depth of $52.20 on December 28, 2018, and reached a year high of $72.21 on 17th May. But they were between $60 and $70 for nearly the entire year.

That was a year, which in April alone, marked the end of US sanctions waivers on Iran; a campaign by General Haftar to capture western Libya; and an attempted coup in Venezuela.

Later on, there were tanker and pipeline attacks in the Gulf; the September strikes against Saudi Aramco’s Abqaiq that led to the largest ever unplanned oil supply disruption. Then there were massive protests in Iraq, Iran and Algeria; deeper Opec cuts in December; rollercoaster US-China trade war ups and down and the preparation for much cleaner shipping fuel standards. In the end, perhaps all these factors cancelled out.

Oil prices are notoriously hard to predict. Because of the importance of oil for crucial tasks and the lack of ready substitutes, small imbalances lead to rapid escalations in price.

Demand does not vary too much from month to month, except when it falls sharply in a recession. But supply responds only sluggishly to price signals. It takes time to develop new fields and drill new wells, but they produce at fairly low costs once operational. So, a sustained strong world economy can outpace the industry’s ability to raise output.

Over the past decade, one factor has completely upended these traditional market dynamics. Shale wells in North America can be drilled and put on production quickly in huge numbers. They are quite expensive, so activity slows down sharply when prices fall. And their production naturally declines fast, falling up to 85 per cent in the first year, as compared to traditional wells that might drop 10 or 20 per cent.

This makes shale far more flexible than conventional fields. Because of this, the market was able to absorb major losses of output from Libya, Iran and Venezuela from 2011 onwards, and Opec’s ‘price war’ strategy of boosting production in 2015-16.

Opec has traditionally sought to stabilise prices, cutting output when they fall too low, though bedevilled by quota ‘cheating’. It has often failed to act quickly enough when prices spike and tried to maintain unrealistically high prices. Yet from December 2016 onwards, under its new pact with Russia and other non-Opec countries, compliance to agreed limits has been exceptionally high, and coverage of world production has been wider. This has helped ameliorate downward swings.

But two major factors are going to change the paradigm again in the coming decade.

Firstly, we cannot rely on shale to stabilise prices. Any major and extended disruption would still outstrip the US’s ability to keep up. Opec+ discipline may not be maintained indefinitely: Iraq chafes under production limits, Russian companies want to exit, Iran might negotiate a way out of sanctions, or Venezuela revive under a new government. And shale is beginning to show signs of strain, as investors tire of disappointing returns, and the best areas are drilled up.

Secondly, oil will for the first time begin to face a serious competitor in transport: electric vehicles. A small percentage of the market currently, battery cars are gaining range and falling in price. Increasing numbers of cities and countries are also mooting bans on petrol and diesel vehicles. Some credible forecasts see an overall drop in world oil demand as early as the mid to late 2020s.

In total, declining demand would be expected to reduce prices. But, with conventional field output falling 5-6 per cent per annum without new investment, while demand might be dropping 1 per cent per year or less, new production will still be required. Hostility to fossil fuels may lead to cycles of under-investment and price spikes, which would themselves accelerate the shift to battery vehicles. This is an intensification of Opec’s perennial big problem: it wants to support prices but driving them too high undermines its market.

Bearing all this in mind, contestants in this year’s oil price contest have proposed an even wider spread than in 2019, from $28.75 to $90, though most calls cluster in a range from the high $50s to low $70s.

My own estimate was made before the American killing of their Iranian nemesis, Qassem Sulaimani, but already expected a fair amount of geopolitical harum-scarum. A reasonably strong world economy, probably juiced by US government action ahead of November’s presidential election, will support prices. US shale output may slow down significantly, but other non-Opec output from countries such as Norway, Brazil and Guyana will compensate.

Past performance is no prediction of the future, but my figure is $70.14. No prize but glory awaits the winner.

Robin M. Mills is CEO of Qamar Energy, and author of The Myth of the Oil Crisis