Time Bomb In Oil Markets: Goldman Sachs Issues Warning
While energy traders remain focused on weekly changes in crude supply and demand, manifesting in shifts in inventory of which yesterday’s API data and today’s EIA data was a breathtaking example, a much more troubling data point was revealed by the Energy Information Administration last week when it reported implied gasoline demand.
To be sure, surging gasoline supply and inventories are hardly surprising or new: they remain a byproduct of the unprecedented global crude inventories leftover from two years of failed OPEC policy which resulted in a historic glut. Last January, overall crude runs were up 500,000 bpd as refiners shifted away from diesel and other products to gasoline to chase more attractive margins amid a mild winter and sluggish diesel demand. The move led to an overbuild of gasoline stocks that lingered into the summer, punishing margins when they should have been at their strongest. This January, crude runs are at historic levels, up by roughly 300,000 bpd over last year.
So yes, both gasoline stocks and supply remain at extremely high levels, but what set off alarm bells is not supply, but demand: the EIA last week reported that the 4-week average of gasoline supplied – or implied gasoline demand – in the United States was 8.2 million barrels per day, the lowest since February 2012. And, as Reuters adds, U.S. refiners are now facing the prospects of weakening gasoline demand for the first time in five years.
Unlike excess supply, which may have numerous factors, when it comes to a plunge in end product demand there can only be one implication: the U.S. consumer is very ill, especially when considering that gasoline use has grown every year since 2012, despite fears that demand has topped out amid the growth of fuel efficient cars, urbanization and a graying population.
Upon learning the data, the industry’s immediate concern was about refiners and what it means for already sagging margins: U.S. gasoline demand is closely watched by traders as it accounts for roughly 10 percent of global consumption. U.S. refiners amassed large inventories that punished margins last year, but record gasoline demand and robust exports helped provided a firewall against further slippage. Now the industry faces the prospects of higher crude prices following global production cuts and fresh federal data that suggests their gasoline demand safety net may be eroding.
“It’s tough to base conclusions solely on the weekly data, which can be off significantly,” said Mark Broadbent, a refinery analyst with Wood Mackenzie. “If the demand is low as the data shows, then it’s going to be real problem for refiners.”
But it could be a far bigger problem in what it means for the broader economy.
Enter Goldman, which cuts right to the point: “A 6 percent fall in U.S. demand would require a U.S. recession”
As Goldman analyst Damien Courvalin notes, “implied demand data points to U.S. gasoline demand in January declining 460 kb/d or 5.2 percent year-on-year. In the absence of a base effect, such a decline has only occurred in four periods since 1960 during which time PCE contracted.”
Goldman then adds that “to achieve the 5.9 percent decline suggested by the weekly data, our model requires PCE to contract 6 percent, in other words, a recession.”
So is the gasoline demand data accurate, and is a recession quietly gripping over the U.S., even as most other indicators are calmly flashing green?
Here Goldman refuses to believe the official data, instead reverting to its own model, which “adjusts” the data, to goal-seek the decline to appear more manageable.
Given that the December PCE printed 2.8 percent growth, in line with its performance throughout 2016, we find such a sudden collapse unlikely… our revised model for gasoline demand, which regresses year-on-year change in demand on analogous growth in PCE, pump prices, efficiency, number of public holidays and base effect, points to a 30 kb/d or 0.3 percent decline. Alternatively, given our economists’ forecast for PCE to grow 2.6 percent in 1Q17, such a decline would require a year-on-year efficiency gain of almost 20 percent vs. the maximum historical gain of 8 percent. Finally, the potential reduction in demand on account of the Presidential Inauguration on 20 January is offset by one less weekend day vs. the same period in January 2016.
Goldman then calculates what it believes is the accurate collapse in implied gasoline demand, instead of the 460k b/d reported by the EIA:
Our analysis identifies weekly yield and exports as systematically deviating from their final values and such biases suggest that demand could be revised higher by 190 kb/d…..