Constant revisions to the weekly natural gas storage surveys contribute to undue market volatility, according to commenters on the Energy Information Administration’s (EIA) proposed survey changes. The comments ranged from warnings that changes to the storage revisions policy could lead to market manipulation, to recommendations that companies frequently reporting revisions be penalized.

EIA’s Roy Kass said proposed changes to the weekly storage survey probably won’t be made anytime soon because of the variety and complexity of the comments and the nature of the inquiry. “There were some comments in there that, frankly, I hadn’t thought of,” said Kass. “It’s going to take us a while to digest them. This has never happened before. It is uncharted territory certainly for the natural gas division, and as far as I can tell, for the EIA. We want to get it over with fast, but we don’t want to rush to something that is going to end up being undone.”

Kass said EIA is exploring many options but is focusing on suggestions about changes in its estimation procedure, particularly using three months of complete historical data rather than one month in coming up with the weekly estimation ratio. “We have a statistician working on and exploring what the implications of that would be,” he said (see NGI, Aug. 19, July 15). “The theory is if you average two or three months together, then if you have a significant change in any one month that is going to be blunted.”

Changes in EIA’s estimation ratio due to a shift to a new month of complete historical data has been blamed for repeated revisions to EIA’s weekly data. EIA requested comments on its storage methodology in a July 11 Federal Register notice. Currently all revisions over 7 Bcf are announced at the same time the new weekly storage figures are announced. Under consideration are whether revisions under 7 Bcf should be revealed and whether there should be more immediacy in announcing large revisions of over 35 Bcf.

A majority of comments from futures traders, which appeared on the EIA web site (https://tonto.eia.doe.gov/oog/info/ngs/ngscomments/comments.html ), advised against making “unscheduled major revisions,” arguing that the current policy of announcing revisions at the same time as weekly survey data enabled the market to focus on a specific time to react to changes. The New York Mercantile Exchange, however, supported EIA’s suggested announcement of a major revision at 10:30 a.m. on the business day following one on which there was an announcement that a revision would be made.

Others, however, argued against accommodating speculation on the futures market, suggesting there would be less volatility if the weekly futures reports were issued after the market closed. Out of cycle announcements of large revisions should be made as close as possible to the notification that there will be a revision to allow less time for market volatility, according to the Natural Gas Supply Association (NGSA). But Duke Energy Trading and Marketing thinks there is no benefit to EIA’s proposal for any “unscheduled major revisions.” The “timing uncertainty and surprise factor” of the proposal has the “potential for significant market disruption, without any offsetting benefits,” Duke said.

NGSA, fearing that storage operators would game the market, also said EIA should take steps to prevent companies from manipulating data by expanding the number of companies reporting to the survey to the universe of storage operators, alerting companies their data reporting is being monitored and will be reviewed by the Commodity Futures Trading Commission (CFTC) and the Securities and Exchange Commission (SEC), and by requiring companies to state a reason for revisions and publish that reason.

NGSA also said information about changes should be dispersed uniformly through notice on EIA’s web site, simultaneous with an email notice. While 7 Bcf is an appropriate number for denoting a scheduled change, there are times when that would be a very large number relative to the total being reported in a given week. “Historically, the first week of the injection period (first week in April) and the first week of the withdrawal period (first week in November) have very low injection/withdrawal numbers.” EIA should hold off releasing adjustments during those transition weeks to help minimize volatility.

One commenter, who identified himself as a veteran of the industry who is not currently working for any company and speaking as a consumer, pointed out that “the primary purpose of the storage survey should be as a planning tool for government to obtain timely information to develop a response plan for regional and national emergencies involving natural gas and related fuels.” Further, John Cochener said, “any information which artificially triggers gas price volatility is not in the best interest of gas consumers and plays into the hands of those attempting to capitalize on natural gas markets for short-term financial gain.”

He also pointed out that the “survey” is just that; it is not an accounting, and by definition includes a margin of error. Cochener warned against using the survey numbers in other formulas, such as attempting to derive implied production. “The national gas balance contains many variable moving parts, including demand, and gas and LNG imports.” He said he believed EIA’s weekly storage report “presents a fair and impartial view of gas storage, and EIA does not display bias or an agenda in the reporting.”

Others also pointed out the dangers of market manipulators who might report a false number, take a futures position and then report a revision. There should be a penalty for inaccurate reporting, they said.

David Finch, formerly in the gas storage department of a large company, said it was his experience that companies collected data on storage every day with only “rare” inaccuracies. Consequently any adjustments reported to the American Gas Association (AGA) were very minor. He questioned EIA’s procedures and processing of the numbers.

Suggestions were made about EIA’s procedure for extrapolating its sample survey to the full market, which calls for a periodic change in the base month and results in regularly programmed revisions (see Daily GPI, Aug. 13 ). When a revision is due to EIA’s changing of the reference month, that fact should be announced in advance or at least footnoted, so market-watchers will be aware there was no actual change in the amount of gas in storage, one commenter said. Another suggested using a three-month rolling average ratio to estimate the “full” from the sample. Using a three month average, dropping one month and adding another when new data is available, would minimize the changeover.

An AEP Energy commenter pointed out that the survey sample is disproportionately weighted towards large storage facilities and there is not adequate representation of the increasing numbers of high deliverability salt storage fields. Salt storage represented less than 4% of total reported working gas in storage in November, 2001, at the same time it actually was 15% of the deliverability of all storage, Keith Barnett said.

He also objected to EIA’s collection of data in whole numbers, pointing out that could distort the survey by up to 14 Bcf in a single week. “This degree of error may be minor in a large winter week with withdrawals approaching 200 Bcf, but could be damaging to the market and consumers if it occurred in the summer. He recommends reporting results to one decimal point. Barnett also suggested EIA should advise in advance when a change in reference month is coming up.

The Coalition for Energy Market Integrity and Transparency said that instead of collecting weekly results from 45 storage operators, EIA should require filings by all of the approximately 110 storage operators. The coalition, which has a significant industry membership that includes the American Public Gas Association, American Public Power Association, Halliburton, Louisiana Independent Oil and Gas Association, Noble Drilling, Schlumberger, and the Texas Independent Producers and Royalty Owners Association, also bemoans the lack of any other supply and demand data, and suggests that collecting accurate daily scheduled volumes on pipelines would provide a good picture of the market. This information could show the volume and flow direction of natural gas in the United States, which could be used to show demand and supply in different regions. Much of this flow data appears on pipeline bulletin boards at this time, but the information posted varies from company to company.

The Coalition also suggests that gas delivered into a consuming area could be broken down into gas for consumption, injection into storage or gas in transit to another region. Imports also could be monitored.

The Coalition also believes price reporting publications and electronic trading systems can be manipulated. Price reporting agencies should collect and retain detailed records on transactions which would be subject to audit by EIA or the Federal Energy Regulatory Commission, the Coalition said. EIA itself should collect and publish price information.

Responding to commenters, Kass said EIA simply lacks the resources for monitoring wholesale spot gas prices and pipeline daily throughput. He noted that this inquiry is strictly on the revision policy for the storage report and it is unlikely that EIA will be making any changes to anything other than the storage report as a result.

Pete Frost of Conoco Inc. recommended a three-step process to establish and publish and accurate storage survey: determine the root causes of incongruent or inaccurate data; eliminate the reporting inaccuracies; and determine how and when revisions should be reported. Conoco suggests EIA should conduct an open conference or rulemaking to determine the causes of the current revisions and potential solutions. “Meaningful and consistent penalties” should be part of the establishment of an accurate survey, he said.

©Copyright 2002 Intelligence Press Inc. Allrights reserved. The preceding news report may not be republishedor redistributed, in whole or in part, in any form, without priorwritten consent of Intelligence Press, Inc.