Something is rotten in the state of Denmark.
U.S. crude oil futures prices made 15-month highs today following the release of the Energy Information Administration's (EIA) weekly petroleum statistics. The EIA reported a 5.2 million barrel (mmb) decline, extending the crude stock draw to 26.5 mmb from the week ending August 26th. The market may have rallied based on bad data.
The EIA has made two major changes in its weekly statistics, which may have corrupted them to a greater or lesser extent. I have found one error and have strong suspicions there are errors in its crude oil imports figures, which have mysteriously dropped from unnamed countries. In addition, net crude imports from Saudi Arabia account for 84% of the drop in imports from the U.S. top 10 importing sources - but that may be by design as a means for supporting oil prices, as explained further below.
The bottom line is that the statistics raise more questions than they answer, and so, it is far too premature to conclude a major global shift in the supply-demand balance has occurred.
EIA Change in Exports Estimation
Beginning with the week ending August 26th, the EIA revised its methodology for estimating crude oil and petroleum product exports. I had previously discussed this change. Though the EIA admitted that its prior export estimation process caused its demand figures to be overestimated, I now suspect that the new method may be causing its crude import estimates to be underestimated (as discussed further below).
EIA Change to Stocks' Accounting Basis
As I explained last week, the EIA changed its reporting basis in the week ending October 7th to exclude about 31 million barrels of crude oil lease stocks in the U.S. total commercial crude oil inventory data series. This does not mean stocks are lower, have been "estimated away," or have disappeared. The EIA is just not including them anymore.
Historical data has also been adjusted so that comparisons to past data will have the same accounting basis. This also affects the EIA's Short-Term Energy Outlook (STEO) historical data and projections of future U.S. and OECD commercial stocks.
I suspect that making changes to the organization's reporting system has resulted in errors that are corrupting the data. I can point to one such error and have reason for suspecting others (as described further below).
On line 13 of the EIA's "U.S. Petroleum Balance Sheet," (Table 1), there is an item named "Adjustment." (It had been previously named "Unaccounted for," and is a balancing item to make the supply and demand numbers equal to the crude stock change for the week.) The adjustment for this week is -312 and the number for the 4 weeks ending 10/14/16 is -1,041. This 4-week number implies that it cannot explain a stock draw of 29 million barrels (1,041,000 b/d times 28 days). This number is wrong. The adjustment reported for the past 4 weeks (240, -130, 415, and -312) averages 53. The graph below shows the 4-week trends going back in time and highlights how far this number is out of line.
By the way, the -312,000 b/d adjustment means that the EIA's supply/demand numbers cannot explain 2.2 million barrels of the 5.2 mmb crude stock draw reported for the week.
Imports Numbers Corrupted?
Based on the supply and demand reported, it is obvious that the 26.5 million barrel stock draw since the week ending August 26th is due to a sharp decrease in net crude imports.
Specifically, according to the EIA's weekly numbers, the 4-week trend in net crude imports dropped by 745,000 b/d from the week ending August 26th, 7.857 million barrels per day (mmbd) to 7.112 mmbd for the week ending October 14th.
The EIA reports the top 10 sources of net crude imports on a weekly and monthly basis. Its weekly data show that imports from the top 10 sources dropped by 291,000 b/d over the same period mentioned above. On a net basis, crude imports from Saudi Arabia dropped by 243,000 b/d, 84% of the net drop.
The Saudi joint venture Motiva operates U.S. refineries with more than 1 million barrels per day of capacity. While Motiva may have decreased imports for operational reasons, it also seems possible that Saudi Arabia reduced exports to the U.S. after it changed its strategy, preferring higher prices, to reduce the widely reported EIA crude stocks.
A larger, more uncertain question is why imports from other sources than the top 10 countries from which the U.S. imports crude fell by 454,000 b/d. According to the EIA's monthly numbers for the first half of 2016, these other sources accounted for only 342,000 b/d.
Analyzing the EIA's weekly data provides no further explanation. It raises the question of how and why net crude oil imports, and whether there are errors contributing to the EIA's mis-estimation of crude stock changes.
The EIA has made two significant changes to its methodologies and databases since the week ending August 26th. I have pointed out one glaring error that resulted in its reporting.
It has reported an unusual drop in net crude imports, a number which is impacted by the change in its crude exports methodology.
The large drop in crude imports justifies a large drop in crude stocks. But if the crude import number had been higher, the EIA would have been justified in reporting higher crude stocks.
I view the crude import and stock data with a high degree of skepticism. I think we will know the status of imports and stocks when the EIA completes its monthly surveys in about two months. For now, I would not place much confidence in these numbers.
Disclosure: I/we have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours.
I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.