Oil Data: A Real Problem

            One major problem not often discussed is the data we get regarding oil. We all await reports from OPEC, EIA, IEA, and others to help us determine what is happening with oil markets. These reports detail storage, production, and other key indicators that should help us determine when the market will return to balance and how long it will stay there. But the issues with the data are numerous.

The problem is that this data isn’t terribly accurate, especially figures on oil production that come out of countries that are known for their secrecy – like Saudi Arabia.

“There is a persistent lack of agreement among estimates made by Saudi Aramco, the International Energy Agency, the US EIA/DOE and BP’s annual Statistical Review of World Energy of the amount of Saudi Arabian crude produced each year from 1988 to 2004,” writes Matthew R. Simmons in Twilight In the Desert (2005). “The variances between these sources highlight the uncertainty about the volume of oil that Saudi Arabia produces.”

The problem on production data is so bad that OPEC actually looks to both reports from member countries on what they produced and reports from secondary sources on what each country produces because countries tend to overstate or understate their production. Just for this past month, for example, Iraq told OPEC it produced 4.775 million bpd in September, while the secondary sources put output at 4.455 million bpd. From Iraq's point of view, joining the OPEC supply cut deal from the higher figure would be more favorable. At times there will be incentives for OPEC members to over-report and at times there will be incentives for them to underreport,

And then there is the issue of storage. China, for example is reported to have stored substantially more oil then they have reported. Whether it is for strategic reasons, or simply an accounting gaffe, who knows?

But oil data is not just a problem because of secretive countries. Even the U.S. has problems reporting oil storage figures. And herein, perhaps, lies the biggest rub.

In an announcement at the end of August, the EIA announced changes to their weekly statistics. And many familiar with the market rejoiced. “The improvements to the EIA data will definitely offer a better read for those tracking the fundamentals of the energy markets compared to the less accurate estimates’ that were previously published,” said Tyler Richey, co-editor of The 7:00’s Report.

One significant change is that on a monthly basis, the EIA will now “re-benchmark that weekly estimate, if necessary, if we see that the series are diverging materially.” In addition, oil export data, not too important before the lifting of export restrictions but we can expect to become increasingly more important over time, will now be based on “near-real-time export data” provided by U.S. Customs and Border Protection. The EIA previously used weekly export estimates based on monthly official export data published by the U.S. Census Bureau about six weeks following the end of each reporting month.

But even the EIA has acknowledged that their data won’t be fool proof – just that we should have a better ability to observe overall trends in the data. Which to me is a de facto acknowledgment that problems will remain.

And this is the problem: there is a great deal of variance in what is reported and what makes sense. Art Berman of Oilprice.com lays out a fascinating article detailing all of the problems with the statistics – and I encourage you strongly to read the full article – and reaches the conclusion that the problems with the data may call into question reliability of the EIA and other storage reports (incidentally this discrepancy is something I have been trying to figure out since the downturn started). Click here to read the full article with all the nice graphs and charts.

           But if you do not want to read the full article, here is a summary. In effect, Mr. Berman argues that the EIA makes estimates, and then goes back to square the books. And they notate their report simply as an “Adjustment”. Makes sense. And logically, Stock Change = Domestic Production + Net Imports – Crude Oil Input to Refineries.

But that is not what happens. Instead the EIA employs the following logic: estimated stock levels in tank farms and underground storage are relatively dependable and that any imbalance must be from less reliable production, net import or refinery intake data. The problem is that the results of the adjustment sometimes overwhelms the original data. So much so that in September 2016, the adjustment is 60 percent of the stock change. An astronomical figure. The adjustment for unaccounted-for oil averaged about 15 percent from 2002 through 2010. In 2016, almost 80 percent of reported stocks are from unaccounted-for oil.

There is no obvious solution for the mystery of unaccounted-for oil in U.S. inventories. Possible explanations, however, include:

1. Crude field production is underestimated
2. Net crude oil imports are underestimated
3. Refinery inputs are over-reported
4. Crude oil stocks are over-reported

or any combination of those possibilities.

But we don’t know which one it is.

So even setting aside the more secretive nations who do not want to report their actual storage levels, even in the U.S. we have problems reporting storage levels.

And of course, we have API data to contend with as well, which often offers us a dramatically different view of oil storage than the EIA report. I will not begin to get into all of the problems associated with the API data.

However, there are several moves around the world to increase transparency. In late September, it was reported that The United Arab Emirates port of Fujairah, now the Middle East's largest commercial storage hub for refined oil products, will publish weekly inventory data to improve transparency and boost the credentials of the Emirates' only Indian Ocean port that has oil export facilities. The UAE has clear hopes that more transparency will increase business for port, its storage and refining capabilities. And these are steps in the right direction to help oil traders and those of us dependent on oil understand market conditions and anticipate those conditions more timely.

So, one question we asked last week that seems pertinent now: how much oil is in storage? Several commentators have stated that U.S. Inventories will have to drop by 150 million barrels before there will be a meaningful recovery. And if we do not know how many barrels are truly in storage, when that number will occur, even assuming a production freeze, becomes more questionable. And the bigger problem: OECD inventories report liquids, not crude oil. So again, we see a lack of transparency. All of this adds wrinkles to the question of when to reasonably anticipate recovery in the oil patch. But it is fascinating to me that at the heart of the matter, where numbers do truly matter, as sophisticated and complicated as our global economy has become we still can not seem to account for what one would think would be something fairly easy to account for: just how much oil is out there…

By: Ty Chapman

Five Star Metals, Inc.

Raising the Bar for Customer Service and Quality

Twitter: @FSM_TY

Follow me on Twitter to get the latest updates throughout the week!