Under Fire, EIA Takes Public Comment on Storage Revision Policy
Stung by criticism of a revision made in its weekly gas storage report two weeks ago, the Energy Information Administration (EIA) said last Wednesday that it will publish a notice in the Federal Register in January soliciting public comment on its storage data revision policy. The agency also said it is taking a number of other actions to assure the quality of weekly natural gas storage data and the appropriate timing of its release.
The decision to take comments follows a barrage of public criticism over an erroneous EIA storage number and the subsequent handling of a large revision. The revision was published on Dec. 2 to storage data for the week ending Nov. 19 that was necessitated by a resubmission of data by one or more EIA storage survey respondents.
The revision caused the working gas levels for Nov. 19 to change from 3,272 Bcf to 3,304 Bcf. The change was confined to the East region where working gas levels were revised to 1,915 Bcf from 1,947 Bcf. As a result of the revision, the implied net change to national working gas levels between the weeks ending Nov. 12 and Nov. 19 changed from -49 Bcf to -17 Bcf.
While a 32 Bcf revision was large on its face, the more significant factor was the impact the original incorrect -49 Bcf weekly change had on the futures market at the time of its release.
The storage report for the week ending Nov. 19 was released a day early on Nov. 24 because of the Thanksgiving holiday the next day. It also happened to be the last day of trading on the December natural gas futures contract.
The erroneous 49 Bcf withdrawal was much larger than the market had expected and sent December futures soaring $1.183 to an expiration day settlement of $7.976. The contract reached a daily high of $8.20, while the daily low was $6.66, for a trading range of $1.54.
At the time the 49 Bcf withdrawal was announced, traders were dumfounded because the total of heating degree days for the week was close to the total reported the previous week when only 6 Bcf was withdrawn.
But the market didn't find out until a week later that the number was false. EIA's revision policy, which was based on extensive public comment, required that, no matter when it received revised information, it wait a full week until the next scheduled release of the storage report to inform the market that the original numbers were incorrect.
News that the original withdrawal was erroneous outraged many consumers who set their baseload gas purchases based on the Nymex contract settlement for the month which was set just hours after the original storage report came out. Only one week later futures prices had fallen back to where they had been prior to the release of the erroneous storage report (see NGI, Dec. 6). For those indexed to the December contract, however, the die was cast.
The Nov. 24 run-up "significantly influenced" the price for December physical deliveries of approximately 73 Bcf/d of gas, which could cost gas consumers as much as $1 billion in higher gas prices, said Paul Cicio, executive director of the Industrial Energy Consumers Association.
Nevertheless, the revision EIA made was done in accordance with the existing storage revision policy. "EIA is obligated to adhere to its announced revision policy, as it did in this case, to assure that all market participants know when and where revisions might be forthcoming," EIA Administrator Guy Caruso said last Wednesday.
"However, recent events indicate a need for EIA to again seek public input regarding possible improvements in the current revision policy for the weekly gas storage survey," he added. "Once we have this input, we will be positioned to implement any changes to revision policies that are needed to best serve the public interest."
EIA also said it already is making adjustments in the way it assures the quality of the data it receives. The agency said it already has contacted respondents to re-emphasize the importance of accurate, timely data submissions.
"Accurate and timely data submissions from all respondents, 52 weeks a year, are needed to assure the quality of natural gas storage estimates," noted Caruso. Second, EIA is continuing ongoing efforts to improve its own quality processes and systems. Finally, EIA will consider, on a case-by-case basis, adjustments to the date and timing of storage data releases during holiday weeks.
"The number was way out of sync. They should have questioned it when they received it," Cicio said in an interview with NGI last week. "Anytime there is a major deviation there should be a phone call, not an e-mail, but a phone call and an answer. If they need to start earlier in the morning in terms of collecting that data so that by a certain time in the day they can release it and have accurate data, having talked to people, so be it. They really need to challenge anything [that looks remotely incorrect]. There should be a standard deviation [model].
"At this point they should know all this," he said. "They've been doing this report for some time now."