Kate Mackenzie What’s in an oil storage data flaw?

Dow Jones says it has uncovered some serious shortcomings in the widely-watched weekly oil storage data published by the US Energy Information Agency.

It says the methodology is too dated for today’s complex futures markets. For example, the data is entered manually and difficult to check, and it is not adequately secured, according to the report. Dow Jones is citing a report commissioned by consultants SAIC, and internal EIA emails obtained under the Freedom of Information Act. How important is the flaw, and how are markets likely to react?

The story gives a specific example of errors creeping into the EIA data, and quite possibly affecting crude oil futures:

On Sept. 16, the EIA released data showing almost four million barrels of oil had vanished from the Cushing storage hub in Oklahoma during a single week. The market paid particular attention because Cushing is the nation’s most important commercial storage facility. Its oil is used to fill orders from buyers on the New York Mercantile Exchange. Oil futures jumped 2.2% after the report

But out of the sizable drop at Cushing, 1.7 million barrels represented a correction made after the EIA discovered a previous error in one company’s reporting, according to the emails. James Beck, who heads the team that conducts the weekly survey, confirmed the correction in an interview.

The EIA weekly storage data is certainly closely watched by market participants, and the WSJ report underlines this with a quote from Bank of America Merrill Lynch oil analyst Francisco Blanch who says questions about the data would “limit the validity of some of our own analysis, particularly as it relates to the US market”.

This is a dramatic story, no doubt. But three points should be made that temper it, somewhat, or at least explain why EIA oil and gas chief Stephen Harvey didn’t exactly sound mortified when Dow Jones asked him about it:

“Should you be concerned? Yes. Is it as good as we’d like it to be? No. Is it better or worse than some other countries where we’d like to know this information? It’s probably a whole lot better,” Mr. Harvey said.

First, as Harvey suggests, the US statistics, despite their flaws, may still hold up relatively well on the world stage. The IEA publishes OECD stocks, but only monthly data. Efforts are under way in the EU to improve reporting methodology. As for China, it’s anyone’s guess. And don’t even start on the raging debate over the opacity of oil data from Saudi Arabia, the world’s biggest oil producer.

Second, although the US remains, for now, the biggest consumer of crude oil and products, it’s not exactly calling the shots in world oil markets these days. OECD oil demand is widely thought to have peaked, if not quite yet, then imminently. The real drivers of changing oil demand are Asian countries, particularly China.

And in fact, the EIA data is not seen as all-important, or as infallible. JBC Energy’s analysts fumed in October over the market’s misreading of the data.  The American Petroleum Institute also publishes weekly US crude inventory data and the two figures rarely, if ever, match. Furthermore, if the EIA data was considered rock-solid, presumably there would be no interest in the infra-red and other spy style technologies employed by the likes of Genscape, nor any market for the estimates it publishes based on its findings. And the usefulness of storage levels at the widely-referenced delivery point in Cushing, Oklahoma, was the subject of fierce debate last year.

All that said, a new uncertainty over the US storage statistics could add another layer of intrigue to the already opaque world of oil data.

Related links:

China, setting the world’s oil prices (FT Energy Source)
Crude interpretations of stocks data (FT Energy Source)