Thursday, July 18, 2019

EIA Explains The US Weekly Crude Oil "Adjustment" -- July 18, 2019

Updates

July 19, 2019: see first comment. The reader points out something incredibly important and something the EIA either purposely or inadvertently "forgot" to mention or point out.

In the EIA discussion, the EIA states: 
In recent weeks, the crude oil adjustment has been growing in absolute value, as high as 881,000 barrels for the week ending May 24. 
In my book, 881,000 bbls can be rounded to one million bbls. 

According to the reader the EIA omitted two incredibly important words, "per day." I've noted the change as it should have been written in bold red in the original post.

If that's accurate, it means we are talking about as much as 6.167 million bbls "adjustment" per week -- hardly trivial.

Original Post 

We've talked about this a couple of times. Sometime in the very recent past, the last month (?), the EIA has apparently changed its methodology for calculating US crude oil inventories.

And here it is, how timely, released yesterday by the EIA, released July 17, 2019, This Week In Petroleum, the crude oil "adjustment" accounts for differences in supply and disposition.

But it looks like this does not explain any change in methodology; it simply explains why "supply and disposition" seldom balance.

The EIA must have been getting a lot of phone calls from Wall Street analysts losing their shirts.

Link here.

The explanation begins:
The U.S. Energy Information Administration’s (EIA) Weekly Petroleum Status Report (WPSR) provides weekly estimates of U.S. crude oil supply, including a measure of how well the supply of crude oil and the disposition of crude oil balance with each other.
This measure—referred to as the adjustment—is a derived term equal to the difference between supply and disposition. If the reported supply and disposition of crude oil balanced perfectly each week, the adjustment would equal zero. For several reasons, however, this is rarely the case.

Weekly U.S. crude oil supply and disposition data are based on a combination of EIA survey data, U.S. Customs and Border Protection data, and modeled estimates. All statistical samples using survey data have small but unavoidable imprecisions, and model estimated data’s precision can vary. This imprecision in estimating each element of the crude oil balance can result in some over- and under-estimation in both supply and disposition.

In recent weeks, the crude oil adjustment has been growing in absolute value, as high as 881,000 barrels [sic; see comment below] for the week ending May 24.
However, this is still relatively small, when compared with the entire U.S. crude oil balance, less than 2.5% on a rolling four-week average basis (the sum of production, imports, exports, and refinery inputs) (Figure 1 at the linked article).
Although an increased adjustment is, in some part, the result of the inherent challenge of estimating perfectly each reporting period, increasing volumes of U.S. crude oil production and exports and other factors may also play a role. EIA will continue to evaluate crude oil data to identify possible sources of the higher crude oil adjustment.

8 comments:

  1. This is common knowledge. EIA has had this info for a long time in their FAQs and methodology summaries, publicly available by clicking on the links in their pages. They even did a webinar on it about a year and a half ago. The problem is bloggers and SA commenters and TV spokesmodel fluffheads who don't want to RTFM because it feels too much like work.

    This explanation still won't change people who are intellectually lazy. Or who are new to things but like to comment on Twitter and blogs with conspiracy theories about evil EIA.

    The smart Wall Street money (really good analysts, not TV news talking heads or SA article writers) already understands this well.

    ReplyDelete
    Replies
    1. "Fluffheads." LOL. I think EIA's explanation that a 2% "error" on production 4 million bopd is a lot different than a 2% error on production of 12 million bopd.

      Delete
    2. Whether EIA is right or wrong is different from these people not having even read the explanations yet. It's like arguing about some part of an airplane and not bothering to look at the tech manuals on the desk.

      Delete
    3. I do have some readers that watch this very, very closely and do come up with some inconsistencies in the various government reports. But I know you are correct: vast number of"analysts" have not read the "manual."

      Having said all that, reading the comments over at "peak oil" sites suggest that there are many folks who live in a different universe than I do. LOL.

      Delete
    4. They are incorrigible. They've had years--many over a decade--of amateur reading on oil production and still don't know basic things. Like what's the difference between lease condensate and plant condensate. And then add onto it that they don't even feel the gap. That they think they know everything. But really they're just kooks.

      Delete
    5. Wow, you are so correct on that, about not knowing the basics.

      I probably know only 1% of all that goes on in the oil sector, but it was a huge aha! when I sorted out the differences between heavy oil and light oil, and the implications of those differences. I don't think a lot of "analysts" (until perhaps recently) were even aware of that.

      But reading the comments at some of those other sites does "make my day" when looking for a bit of humor.

      Delete
  2. there was a minor omission in that explanation, Bruce; "per day"...it should read "as high as 881,000 barrels per day for the week ending May 24...that means we're talking about as much as 6.167 barrels in a week...moreover, often times the sign on the adjustment switches from positive to negative, to give us a week over week shift of over a million barrels per day...a million barrels per day is a hell of a big number if your inventory figure turns out to be off by that much...imagine the difference in the price of oil on the day of a report that is 7 million barrels smaller or greater than what they reported otherwise..

    ReplyDelete
    Replies
    1. In the comment above, the word "million" should follow "6.167 bbls." The writer is suggesting that the adjustment is upwards of 6 million bbls/week.

      I've brought this up to the post itself. This gets complicated; I don't understand all of it; I don't follow it closely. If there are errors in what I'm reporting, it is not intentional. I do not my best not to make errors of "material" value/importance.

      Delete