EIA crude inventories preview: Oil up after API report smashes estimates

Commodities

Crude oil and Brent oil are moving higher for a second session today after another massive draw in US oil stockpiles.

WTI has gained $0.70 (+1.8%) and Brent oil is $0.73 (+1.8%) higher. Crude oil has now almost erased the losses incurred since prices tumbled on September 8th.

Crude, Brent up on falling oil stockpiles

The latest oil report from the American Petroleum Institute revealed that US oil inventories fell by 9.517 million barrels in the week ending September 11th. Analysts had expected a draw of 1.271 million barrels.

While crude stockpiles fell last week, gasoline inventories rose. The Labor Day holiday marks the end of the US summer driving season and falling gasoline demand is expected heading into the winter months.

This may already be priced into the market, however, with crude and Brent having languished near their lowest levels since June at the start of the week. Both the Organisation of the Petroleum Exporting Countries and the International Energy Administration revealed more bearish forecasts for the recovery in global oil demand this week.

Hurricane Sally to provide short-term boost for crude oil?

Oil is being leant short-term support thanks to the approach of Hurricane Sally, which is expected to cut between 3 million and 6 million barrels of energy production along the Gulf Coast.

However, the weather system has also shuttered some refineries, meaning that oil demand has also taken a hit.

Will US EIA oil inventories data contradict API numbers again?

While the API numbers are huge it’s worth remembering that latest week’s report was later contradicted by the official Energy Information Administration figures. While the API report showed a draw of nearly 3 million barrels for the week ending September 4th, the EIA data revealed a 2 million barrel build.

It seems unlikely that the EIA numbers would diverge so heavily from the API figures, but it is worth remembering that there are discrepancies between the two data sets.