Restatements by the Energy Information Administration (EIA) for its 914 data on U.S. natural gas production may not be as painful as some assume, Barclays Capital analysts said last week.

Beginning with its April report, the EIA plans to introduce a new way to calculate U.S. gas output (see NGI, April 5). The agency said the changes are based on a comprehensive review of its 2009 914 system. Some analysts recently have warned that if EIA restates some of its data on production, an underaccounting could leave the United States in a precarious situation once the economy recovers (see NGI, April 12).

However, Barclays analysts Jim Crandell, Biliana Pehlivanova and Michael Zenker said the restatements would only correct past errors and perhaps not cause harm to the markets as some assume.

“The change in 914 data will most likely diminish the balancing item or, in other words, simply reduce the reporting error,” wrote the trio in a report to clients. “Because the balancing item has run so large in 2009, our forward balances already reflect it.”

The “exact magnitude” of the revisions to EIA’s production data “is impossible to derive, even if one fully comprehends the nuances of the changes to the methodology. However, the direction is fairly clear: both the balancing item and EIA’s statements indicate that form 914 is most likely overstating production. Most importantly, in our view, the revision is unlikely to have a meaningful effect on balances — neither historically, nor for our forward projections.”

EIA’s monthly gas statistics, which serve as a basis for Barclays’ supply/demand models, “gather supply, demand and storage data from various surveys, each of which follows a different methodology, queries different market participants, and produces results independent from the other surveys. Thus, the 914 data restatement will not impact demand, nor storage levels, historically.”

Crandell and his colleagues said they didn’t predict the accounting errors, but the Barclays team “carried forward the average nonseasonal error from the past 12 reports (58 Bcf/ month, currently),” they noted. “Attributing the error to any particular category would be arbitrary. In our balances, we first forecast all individual supply and demand components, then subtract demand from supply to get to an implied monthly change in storage levels. Currently, we have adjusted the implied monthly injection/withdrawal to reflect the overreporting of supply or underreporting of demand by subtracting the 2009 average balancing item.”

If the revision by EIA lowers output and brings the balancing item “closer to zero,” the Barclays team would adjust its forward-looking balances, which “most likely” would leave demand unchanged and result in small changes, “if any,” to the end-of-season inventory estimate.

“This is not to say that a downward revision to production would be inconsequential,” said the trio. “The one area that the revision should impact, and this could potentially carry a more meaningful implication, is the relationship between rig counts and production. Many analysts, including ourselves, expected production to fall into sequential declines late last year as the rig count fell to multiyear lows. But EIA reported no such production decline.”

The market, they said, was left to adjust perceptions on the relationship between gas output and drilling, and specifically what appeared to be a quick acceleration in rig productivity gains.

Because EIA’s 2009 production figures likely were overstating supply, “the implied increase in rig productivity in 2009 versus 2008 would appear too high,” said the analysts. “A downward revision to 2009 production will imply that fewer Btus of gas were produced per running rig, and would thus impact calculations of the relationship of drilling and production.”

Assuming as an example that 2009 production data is revised lower by 1 Bcf/d for every month since March 2009, said the trio, the implied rig productivity drops by about 6% (relative to current assumptions) forward into 2010. Maintaining the Barclays current rig count trajectory assumptions in 2010 would result in production growing by 2.1 Bcf/d year/year (y/y), compared with a current estimate for y/y growth of 2.7 Bcf/d.

“Unfortunately, with the next Natural Gas Monthly, the EIA is only planning to revise January 2010 production numbers, while a revision to 2009 data is expected later in the fall of 2010,” noted the analysts. “Thus, we will be left with the knowledge that the data are most likely wrong and overstate production, but with no clarity on how to accurately address that.”

©Copyright 2010Intelligence Press Inc. All rights reserved. The preceding news reportmay not be republished or redistributed, in whole or in part, in anyform, without prior written consent of Intelligence Press, Inc.