US Gasoline Demand has slumped so far in 2017 – that could start to matter a lot, as it drives global oil demand
Oil demand growth remains heavily dependent on transportation fuel growth. About 60% of the oil barrel is made up gasoline and diesel fuel, both of which have historically been foundations of demand.
One level down, US gasoline is increasingly the most important component of overall global transport fuel demand, not least as China’s energy use starts to decouple from GDP (gasoline growth in China in 2016 was only 3%, less than half of the US’s gasoline growth in absolute terms).
Although the US accounts for only 30% of the world’s passenger cars, it consumes 45% of global gasoline, some 9.2m b/d, of a 20 mb/d total.
This is due to two simple facts as the chart below shows: US car owners drive much further per year than anyone else –at 13,000 miles that’s 50% more in fact. And US cars are less fuel efficient on average than rest of world vehicles – about 35% less.
source – dollarsperbbl data, IEA, US federal transport
Over the past few years vehicle miles driven have increased by about 2% yoy , whilst fuel efficiency has remained flat as the US preference has passed to SUVs rather than light vehicles. This means that the increased mileage has more or less passed straight through to more barrels consumed – some 180,000 b/d on average per annum since 2015.
A robust economy and becalmed gasoline price as low as $1.90 gallon during 2016 supported this trend. Given global oil demand increased about 1.5 mb/d in 2016, then US gasoline demand increases alone accounted for about 12% of its growth.
US gasoline demand is therefore important – it accounts for the majority of marginal global gasoline consumption, and supports overall oil demand.
US Gasoline Demand Evaporates
But over the past two months, this gasoline engine of global oil demand has started to weaken considerably.
Current reports of demand show it down around 450,000 b/d from previous trends, with buying at levels not seen since 2002.
OPEC and US oil – stimulating supply, diminishing demand
Of course this slump in demand coincides suspiciously with the agreement in November with OPEC and non-OPEC states to restrain oil production in a bid to increase prices.
This has partly worked by keeping the oil price range-bound at $50-55/bbl, yet it is has also sent US and other national gasoline prices up from the $2/gallon range to near $2.50.
Conventional analysis – see here and chart below from the US EIA – suggests that fuel consumption is relatively price inelastic in the short-term – that is, most folk tend to keep on driving the same miles, and buying the same level of gasoline even when prices hike by nearly 20% as they have done.
These studies suggest that the response is about 5% elastic, – that is for every 20% price increase, demand will drop by 1%. If so, then even OPEC’s price hike should only hit sales by about 90,000 b/d, not the 400-500,000 b/d seen in January and February.
However, other recent work has suggested that consumers are much more price sensitive than previously thought, with maybe 25-30% reaction to price rises that then stick. Price rises above $2 and then $2.50/gallon may be quite symbolic and trigger consumer responses. This would suggest a demand decline of eventually 450-500,000 b/d.
Demand Reduction Models in Résponse to 20% Price Rise
source: dollarsperbbl model
OPEC’s actions might therefore have the perverse effect of stimulating US shale supply, whilst at the same time diminishing US gasoline oil demand – an ironic double-whammy for the cartel that would clearly act as a curb on oil price.
Glitch, Blip or Trend?
First – we ought to wait for another few weeks’ data to ensure that glitches or some very short-term phenomenon isn’t to blame.
If not, there are several possible types of causes ranging from short-term reactions, to more systemic movements.
– Buyer’s remorse – the surge in purchases of SUVs and the use of highly leveraged financing to do so will start to make the family car a significant part of household budgets when insurance, loans, running and fuelling costs are combined. The latest price move of gasoline would add about $265/pa to the average household budget, or about $10-15 each fill up, which would be noticeable.
– EVs and Sailing ships – the surge in new car purchases, however, do take a lot of older gas-guzzlers off the road, and perhaps quicker than forecast. Whilst SUVs are fuel-hungry, newer models are less so than their 10 year-old counterparts . So, whilst its unlikely that EVs have any direct impact on fuel consumption yet, their presence may be starting to cause many drivers, and manufacturers, to be a bit more fuel-conscious than before in a sailing ship effect on gasoline-powered cars.
– Behavioural Elasticity has increased – Maybe the increased elasticity arguments hold water. The chart below shows historical response gasoline price hikes in the US. Clearly a longer-term trend linked to demographics , urbanization and driving patterns might be at work behind the numbers too. Even geo-political awareness may contribute – every increase of 50 cents/gallon passes over another $60bn per year to producing states, mainly OPEC. The price increase might crytsallise several of these behavioural factors together.
Of course, to state it again, the data may just be incorrect – as Goldman Sachs believe.
The size of the change is acute, and does suggest that a shorter-term action may be the cause. However, there may be a combination of near-term price sensitivity intertwined with longer-range shifts in vehicle technology and driving patterns. The next few weeks will reveal more.
Whatever the case, any impact on US gasoline demand this year, perhaps causing trend growth of 100-200,000 b/d to transform into a reduction of say 250,000 b/d, would force the oil market into having to find that net 350-450,000 b/d demand from elsewhere.
And that seems unlikely given the trends in Europe and China toward more conservative fuel policies, and non-gasoline technologies.
With global oil demand growing weakly, a reversal of this type in a prime driver of consumption will provide more weight to the notion of a looming peak in demand.