Andreoli on Oil and Fuel: This time the trend is your friend

Over the last 20 years, the volume of oil consumed per unit of real GDP has contracted on a very consistent basis; consequently, the global economy could grow at more than 2% per year without oil demand rising. Meanwhile shale oil producers—the world’s swing producers—are increasing well productivity at double-digit rates.

A common theme among the Oil & Fuel columns that I’ve written over the last five years is that, ultimately, oil prices are driven by the fundamentals of supply and demand.

Back in April, I explained that much of the price volatility that we had seen and continue to witness has been driven by heightened sensitivity to imperfect information. More specifically, oil traders don’t know how big the current surplus is, nor do they know exactly how much oil is being produced and consumed on even a quarterly basis.

That said, it’s widely believed that the current market oversupply is a temporary phenomenon that will eventually work itself out sooner than later. The problem is that this has been the prevalent view for more than a year and a half—since oil prices tanked at the end of 2014. Increasingly, it looks like “later” is the more likely option, and that “later” in fact means much later.

The fundamentals

Let’s take a fresh look at the fundamentals. At the global level, approximately 95 million barrels of oil are consumed per day, and there’s a very strong correlation between real gross domestic product (GDP) growth and oil demand. Each percentage point of growth in global GDP has been associated with just under a half a percentage point of growth in oil demand. This, of course, means that any growth in real GDP will generate incremental oil demand. 

But there is another way to think through this problem. What if we look instead at oil efficiency—the volume of oil consumed per unit of real GDP—and analyze how this metric has developed over time? Consider this: If the world is becoming more oil efficient, then real GDP should be able to grow even if oil consumption is held constant.

Back in 1995, the world consumed slightly more than 1,830 barrels of oil for every billion U.S. dollars (USD) of real GDP. Ten years later, the quantity of oil consumed per billion USD of real GDP had declined to just over 1,580, which accords with a total decline of 13.6%.

The economist in me wants to argue that the push for oil efficiency over this period was clearly in line with—and largely due to—the market fundamentals. After all, the price for a barrel of oil increased by more than 220%, rising from just over $17 to $54 per barrel over this period.

Then over the next 10 years—from 2005 to 2015—global consumption fell from just over 1,580 to just under 1,350 barrels per billion USD of real GDP, which pencils out to a decline of 14.8%. Over this period, prices actually fell in real terms from just over $54 to just over $52 per barrel, but only after climbing to well over $100 per barrel. So, clearly, the steady decline was not greatly influenced by oil prices—which moved up and down.

Careful inspection reveals that the decline rate in oil consumption per unit of real GDP has been remarkably consistent at the global level. On average, just over 25 barrels per billion USD of real GDP have been cut from consumption annually since 1995.

Surprisingly, if you estimate the number of barrels of oil consumed per unit of real GDP by simply subtracting the average decline on an annual basis, you will end up with estimates of annual global oil consumption that are extremely accurate. More than half of the estimates would be within a half a percentage point of the observed level, three-quarters would be within one percentage point of the observed level, and more than 90% would be within 2 percentage points of the observed level.

To explain in comparative terms, estimating observed oil consumption by using the historical trend in oil consumption per unit of real GDP generates predictions that are more accurate than predictions based on a linear regression model with a coefficient of determination (i.e. R-squared) of 0.986. This means that in a regression model 98.6% of the observed variation in oil consumption can be explained or predicted based on the variation in real GDP. Yet, the estimates based on oil efficiency gains is even more accurate.

Determining market price

It’s fascinating to me how consistently linear the trend in oil efficiency is. Despite demand increasing by 35%, and average annual oil prices ranging from a low of just $12 per barrel to a high of $112 per barrel, a linear trend line of the 21 years of data is within 2% of the observed level on 19 occasions.

If we simply assume that this downward trend in the volume of oil consumed per unit of real GDP continues, we can calculate that real GDP could grow at a compound rate of 2.15% per year when oil consumption is held perfectly flat. That is a significant amount of economic expansion under a circumstance in which oil consumption does not grow.

Even if the world economy grows at double this pace (i.e. 4.3% per year), oil demand would only be expected to grow by just 2.1%, which should be easily manageable given the significant volume of proven reserves of shale oil that technological innovations have allowed the industry to tap in an economical manner over the last five years.

If we ignore the fact that the global surplus has grown to more than 1% of global consumption, meaning that a year of 3.15% real GDP growth could be accommodated without lifting production, we might assume that any incremental oil demand beyond that which can be accommodated by gains in efficiency would be satisfied by the output from the world’s swing producers exploring and producing oil from shale formations across Texas and North Dakota.

So, naturally, the marginal cost of these producers should be the point that determines the market price.

The trend is your friend

But here we see another consumer-friendly and promising trend. The volume of oil produced per shale well has been increasing at fantastic rates over the last half-decade. In Texas’ Permian shale play, which is the most prolific shale formation (currently producing around 2 million barrels per day), new well production per rig increased sixfold from 100 in January 2012 to nearly 600 barrels per day by October 2016.

Similarly, new well production per rig in the Eagle Ford (currently producing around 900 thousand barrels per day) has increased nearly fivefold from slightly less than 250 barrels per day in January 2012 to nearly 1,200 barrels per day in October 2016. Finally, new well production per rig in the slightly less prolific Bakken Formation (currently producing around 900 thousand barrels per day) has increased fourfold from around 250 barrels per day to around 950 barrels per day over the same period.

These multifold increases in well productivity reflect new technologies that have allowed well laterals to increase in length, and this has, of course, come at a cost. That said, productivity has increased many times more quickly than production costs. This is important because, ultimately, the costs are spread across all the barrels of oil that are produced, so by increasing productivity at a much more rapid rate than production costs, the ultimate cost per barrel has been declining rapidly.

There can be wide variability across wells due to geology, but recognizing this, average production costs may have fallen by more than half since 2012, falling from somewhere around $100 per barrel to somewhere around $50 per barrel.

Take-home point

In a way, one could argue that the world has two swing producers, and neither of them are members of OPEC.

The first swing producer is not even a producer at all. Rather, efficiency gains that permit the global economy to grow by over 2.1% per year without increasing demand for oil are a more powerful swing producer than OPEC.

The second is comprised of the group of shale oil producers in the U.S., who, it seems, are increasingly able to turn a modest profit when oil prices fall in the $50 to $60 range—but looking forward, we should expect the breakeven price to continue to fall.

Increasingly it looks like a price rebound won’t happen unless the world returns to exceptional growth rates in excess of 4.5% per year, which has occurred only rarely over the last two decades. There is an old saying: “The trend is your friend until the trend ends.” In this case, I think the trend will be our friend for quite some time, and oil prices should stay within the $50 to $60 range.