Andreoli on Oil and Fuel: Disagreements on the fundamentals of oil supply and demand

Google is kind enough to aggregate all the oil news stories of the day and send them to me in an e-mail. And as part of my morning routine, I read all the headlines and click through to dig more deeply into those that promise to bring new insight.

I’ve been struck by the regularity with which a headline confidently proclaiming that oil prices are going to plummet will be immediately followed by a headline proclaiming the opposite—and with equal confidence. While opposing views are normal, the schizophrenic nature of my newsfeed has been particularly pronounced as of late.

How prices evolve over the coming months has everything to do with how large the current global oversupply is and the degree to which oil futures traders believe this glut is going to expand or contract. In order to educate their views, oil traders evaluate myriad data to formulate a view on the extent to which oil markets will tighten or loosen in the coming months.

The problem is that this data is untimely, incomplete, opaque and often erroneous.

Just how much is there?
This may come as a surprise to those who are less familiar with the market, but nobody really knows how much oil is produced, how much is consumed and how much goes into storage. Consequently, nobody really knows how large the current glut is and whether it’s growing or contracting. Consider the following production and consumption statistics from the three most prominent sources of oil data.

The U.S. Energy Information Agency (EIA) estimates that oil production averaged 96.4 million barrels per day over the third quarter of 2015 while demand averaged 95.7 million barrels per day. Hence, the EIA data suggest that, over the quarter, the glut grew by a bit more than 700,000 barrels per day.

By contrast, the International Energy Agency (IEA) estimates that during the same period, global oil production averaged 97 million barrels per day while demand averaged 95.4 million barrels per day. Using these estimates, the glut would have expanded by 1.6 million barrels per day over the quarter.

During the same quarter, the Organization of Petroleum Exporting Countries (OPEC) estimates that production averaged 95.5 million barrels per day while consumption averaged 93.8 million barrels per day, thus the glut would have expanded by 1.7 million barrels per day on average.
Interestingly, OPEC reports two series of production data for its own members. The totals stated above reflect the data in the table titled, “OPEC crude oil production based on secondary sources.” This table contains estimates for each of OPEC’s 13 member states. The volumes reported from secondary sources vary from the production volumes reported in the table titled “OPEC crude oil production based on direct communication,” and this table does not include volumes for each of the member states and neither Indonesia nor Libya are reported.

In itself, this is very interesting in light of the number of articles that reflect a belief that OPEC is a well-functioning body that is making calculated decisions. In reality, OPEC is completely dysfunctional, and the low prices are wreaking havoc on these country’s economies. They have every reason to mislead each other, and production from these countries accounts for nearly half of global oil production.

Where is it going?
Beyond OPEC member countries deliberately attempting to mislead one another and the larger market, estimating oil production is challenging even in the most advanced regions. One needs to look no further than Oklahoma to see this.

Back in February, the EIA changed the methodology for estimating the state’s oil production, and this resulted in an upward revision of nearly 100,000 barrels per day. On a base of 300,000 barrels per day, this change in methodology reveals that the previous estimates were off by approximately 25%. With production estimates this far off in the United States, imagine how far off they might be in other places where gathering such statistics is neither well funded nor appreciated.

To come at this from a different angle, one might ask where all this excess oil production is going. At 700,000 barrels per day (EIA), global stocks would need to have increased by nearly 64 million barrels over the third quarter of 2015 alone.

And if the surplus production was closer to OPEC’s estimate of 1.67 million barrels per day, global stocks would need to increase by 155 million barrels. Over the course of the year, we are looking at stocks growing by more than 500 million barrels. This is a lot of oil, so surely we should have some insight into where it is going.

It would be nice to square the production/consumption differential against changes in stocks, but stocks data is even less complete and less reliable than production and consumption data.

According to the IEA, and again looking at third quarter 2015 numbers, the net change in the Organization for Economic Cooperation and Development (OECD) member country stocks amounted to 750,000 barrels per day while the amount of oil in floating storage/transit increased by 150,000 barrels per day. Thus, this leaves approximately 725,000 barrels per day unaccounted for.

It has been reported that China has been filling up strategic reserves while the price is suppressed (not a bad idea), but the country is not forthcoming regarding the details. In a rare public statement, China announced that it more than doubled the size of its strategic crude oil reserves to 190 million barrels between November 2014 and mid-2015—which is equivalent to roughly one month of the country’s net oil imports.

Given that the country’s stated goal is to lift strategic reserves to 550 million barrels by 2020, it’s fair to assume that China maintained a similar purchasing pace through the second half of the year. Thus we can assume—perhaps correctly, or perhaps incorrectly—that China increased its strategic reserves by around 200 million barrels, which pencils out to around 550,000 barrels per day, or roughly two tanker deliveries per week.
Under these assumptions, there are still 175,000 barrels per day that are unaccounted for. In other words, squaring the production/consumption differential against changes in stocks is fraught with uncertainty. Consequently, we can’t really know for certain how large the current glut is.
Of course knowing the size of the glut is only one part of the equation. Estimating how fast it will grow or decline is the other half of the equation.

How many wells?
Some analysts believe that the low prices will bring the North American shale oil boom crashing to the floor. Indeed, the number of drilling rigs in operation in the United States has fallen from around 1,600 in the third quarter of 2014 to just 400 today.

This is especially important because the decline rate of a typical shale oil well is extremely steep. By the end of the first year of production, the typical shale oil well is producing at a rate that is 70% lower than the initial flow rate. By contrast, conventional oil fields can produce oil at the initial rate for years, if not decades.

These steep decline rates mean that new wells must be completed on a continuous basis to keep production from declining. The question that oil traders must answer is how many wells need to be brought online in order to counter the declines. Answering this question is like throwing a dart at a moving dartboard.

During the last five years, drillers have become much more productive. Whereas the average shale well in the Eagle Ford shale play produced an initial flow of just over 100 barrels per day back in 2010, the average initial flow of a well completed in 2015 is closer to 430 barrels. So, clearly one well completed today will more than accommodate the decline in productivity of a well completed one year ago.

The problem is that the number of wells being completed has fallen below the breakeven threshold. Consequently, the EIA estimates that production from shale fields peaked in March of last year and has declined by 9.4%, or 515,000 barrels per day since then. It may in fact be significantly higher or lower.

Whether or not the decline in U.S. oil production means that global production is also down depends largely on the rates at which Russia and OPEC are producing oil, and none of these countries is forthcoming with reliable data.

On the other side of the equation, the demand side, analysts must make estimates of oil consumption for a number of economies whose growth is consistently being downgraded (see: China). On top of that, they must account for increases in efficiency. On the whole, consumption among the countries that comprise the OECD peaked in 2005 at 50 million barrels per day, and has since declined to approximately 45 million barrels per day.

Minimize risk
This leaves the oil traders—those that are ultimately responsible for global prices—swimming in a world of uncertainty. It’s no wonder that my current news feed is so full of divergent views.

When you consider the level of uncertainty in each of the main variables that will ultimately determine how the current glut evolves, it is not difficult to imagine a set of quite reasonable assumptions that would result in the global glut growing while imagining another set of equally reasonable assumptions that would cause the glut to dissipate.

In the end, those of us who are involved in logistics should always seek to be as efficient as possible in all that we do—and this includes energy and fuel efficiency. The markets will do what the markets do, and we can’t change that. What we can do is minimize exposure to risks associated with rising oil/fuel prices.