Q: You’re dialed in right now on the issue of compression of capital expenditures—or capex compression—in the oil industry. Can you give us a quick definition of what that is?
Kopits: Capex compression is a term we use to describe the reduction of upstream spending by the oil companies when their exploration and production costs are rising faster than their oil revenues. That’s what’s happening today. Hess is divesting oil producing properties to increase profits; BP has shelved the deepwater Mad Dog Phase 2 project in the Gulf of Mexico. This is occurring because oil prices haven’t been increasing, and costs have. So oil companies are looking at their portfolio of projects and deciding to postpone or cancel some of them. Were the oil supply rising quickly and oil prices falling, this sort of capital restraint would be normal—the usual boom-bust cycle of the industry. But oil is still in short supply, and very few of the large oil companies have been able to hold oil production over the last few years—even as they were investing massively in oil exploration and production. Now, they are actually reducing investment in upstream projects, even in the face of historically high oil prices and falling production. That’s capex compression.
Q: And here I thought investments in exploration and development were still on their way up. What’s changed?
Kopits: In aggregate, upstream spend is still rising, but at a decreasing pace.
If we look at the issue more broadly though, there are some things happening in the oil business that are beginning to validate views that we, and analysts like Chris SkrebowKopits:ski, have held regarding economic peak oil.
Peak oil does not occur when we run out of oil. Peak oil occurs when the marginal consumer is no longer willing to pay the cost of extracting and processing the marginal barrel of oil. And we can actually calculate what the related numbers are.
Q: How do we do that?
Kopits: To begin with, we refer to the price a nation’s oil consumers are willing to pay as its “carrying capacity.” For the US, carrying capacity is about $95-100 Brent [per-barrel oil price in London]. If the oil price is above this level, oil consumption will decline—which is exactly what we see and what we predicted four years ago. But carrying capacity is not a static number. It changes over time, specifically, with three things: GDP growth, efficiency gains in the use of oil, and dollar inflation. So if GDP goes up, efficiency goes up and the CPI goes up, then the amount that consumers are willing to pay for oil will increase. For China, by the way, we estimate the carrying capacity at around $115-120 / barrel Brent. So oil consumption will increase in China at $115 Brent, but fall in the advanced economies—exactly the pattern we’ve seen in the last few years.
On the supply side, the global oil supply and related costs are determined primarily by two factors: geology and technology. Geology is driving costs by forcing us to frontier areas like ultra deepwater and the Arctic. Technology, on the other hand, is allowing us to access new resources like shale gas and shale / tight oil. So, for any given oil price, depletion will always drive us to more difficult geologies and thus higher costs. Technology, on the other hand, can move us back to easier geologies and lower costs. Hydrofracking of shale oil and gas wells, for example, has done just that.
Also, if you are so inclined, you can add above-ground constraints—Saudi policy or Venezuelan policy or Alaskan tax and royalty rates, for example. But assuming these latter factors are relatively constant, geology and technology will determine supply for any given oil price.
So, to sum all this up: we hit peak production when the marginal consumer is no longer willing to buy the marginal barrel.
Q: I think I’ve read in your work elsewhere that you believe the consumer is already there.
Kopits: The marginal consumer banged into the price of the marginal barrel, on a static basis, somewhere in 2011 at about $110-115 Brent. And then, oil prices essentially stopped rising. Those of us who use supply-constrained forecasting weren’t surprised. It’s entirely consistent with the historical record. But I think many in the oil business still thought, somehow, that oil prices would continue to rise as they had done in the 2000s. After all, the oil supply is widely acknowledged as constrained, even by those who are not necessarily believers in peak oil. So why wouldn’t prices continue to rise if we’re supply short? Well, because there was a price at which the marginal global consumer would rather reduce oil consumption than pay more. And that price is around $110-115 Brent, and from here on in, we should expect that number to rise only with the purchasing power of the marginal consumer.
On the other hand, the cost of extraction development has continued to increase. Last year costs increased somewhere between 10% and 13%, depending on who you talk to. Exxon’s costs rose about 7% in excess of its increase in revenues, which were also falling. And Petrobras’ costs were rising 10% to 13% faster than its revenues. So what we can see is that in the contest between technology and geology, in recent times geology has been winning. Oil has become more expensive to extract.
Q: But when costs increase to a certain level, production should fall; yet we haven’t seen that.
Kopits: In fact, oil production is falling at most the of the oil majors. It was even down at 2% at Petrobras last year. But on a global scale, you’re right. Oil production hasn’t fallen—for three reasons. First, much of what passes for increased “oil” production is actually natural gas production. This includes natural gas liquids from “wet” natural gas wells; LNG [liquefied natural gas] from gas wells; and gas-to-liquids diesel made from natural gas. That’s about half of global oil supply growth in the last six years right there. Check out any investor presentation from the majors. LNG features prominently.
Second, we started throwing massive amounts of upstream spend into this business. Upstream expenditures essentially went from $250 billion around 2005 to about $650 billion this year. In essence, by really jacking up how much money we were putting into the system, we were able to increase production…a little bit. To that we can add some changes in above-ground constraints, primarily in Iraq, which is a very important part of supply growth.
Finally, we made some important technological advances with hydrofracking technology. US tight oil production and Canadian oil sands growth represent just about 100% of net oil supply growth in the last two years.
But leaving these aside, the system hit a wall in 2005—Ken Deffeyes was really spot on with his prediction—and the way we maintained and only slightly grew production after that was essentially by throwing money at it.
This was facilitated by dramatic oil prices jumps, from $25 in 2002 to $112 in 2012. But since 2011, depending on rapidly rising oil prices is no longer a viable strategy. The global economy has said, “this is how much we’ll pay and no more.” At the same time, geology just kept marching along right down the back half of Hubbert’s peak, and costs have continued to rise.
That’s where we are today: price resistance from the consumer and E and P costs that just continue rising. Despite the very high oil price environment, the upstream financial performance at most of the oil majors, including Exxon and Petrobras, has deteriorated. True, Petrobras’ performance is distorted by government interference, but Exxon is arguably the most disciplined investor in the world. But both of them face deteriorating upstream performance for oil.
Q: Given that emerging reality, how are these companies responding?
Kopits: Well, if you look at their capex plans then you see that Shell, BP, Total, Exxon and Hess are all cutting their upstream spend in their 2013-2017 plans going forward. Only Chevron is raising theirs, and only modestly. So in a world where we are struggling to increase global oil supply and the price itself remains high, the major oil companies are in fact beginning to carve back on their exploration and production investments. It’s capex compression.
Q: Why are they going that route?
Kopits: It’s because they’re not getting the bang for their buck. Their megaprojects—ultra deepwater and LNG—are often not able to hold the line on costs. The growing hit-list here includes Australia’s Browse, a $45 billion LNG project that was just cancelled. It includes the Arctic, specifically Alaska, where Shell is sitting out the coming season, in part because they ran their drilling rig aground. But Statoil has said they won’t proceed in Alaska until Shell has shown some progress. ConocoPhillips has just cancelled a jack-up rig order that was intended for the Alaskan market. Total pulled out of Canadian oil sands at a loss. Then we see just last week that BP pulled the plug on Mad Dog Phase 2, which would have been one of the major developments in the Gulf of Mexico—a $10 billion megaproject—and that cancellation was a surprise.
What we’re seeing is that the majors are looking at these high-cost projects, and they are beginning to take a more critical eye. This is very much in line with what our model says, which is that oil prices can’t rise much faster than GDP and inflation, plus or minus. And in fact geological costs, as you come down the back side of Hubbert’s peak, will increase and will do so at an accelerating rate. I think we are beginning to see that process now.
Even when we look at the “good-news” shale / tight oil, some investment is slowing. In the Bakken, for example, the rig count actually peaked in September of 2012, and the year-over-year production growth rate peaked at 90% three months earlier in June. Today the growth rate, while still impressive, is down to about 40%. If that trend continues, we could see single-digit growth in the Bakken much sooner than most think.
Q: So the shale oils won’t be the ever-growing cavalry that everyone expects them to be?
Kopits: If you take the plain vanilla interpretation of this, unless the shales start picking up rapidly from non-exploited plays—not the Permian and the Eagle Ford and the Bakken, but places like the Utica and Monterey, where results have been disappointing, or some other plays or even abroad—you are looking at a world in which the marginal consumer is beginning to reject the marginal barrel. And if you run this out for a period of time, you will peak out the oil supply. I think the peak occurs in a finite time frame—not 2030, not 2020. Maybe 2014 or 2016—I’m not exactly sure, but sometime pretty soon, unless shale oil really takes off in new plays.
Q: So the story line getting a ton of ink of late—peak oil is dead….it isn’t actually quite dead yet?
Kopits: No. But importantly, we’re going to peak out production not because we’re “running out of oil,” but because the marginal consumer is not willing to pay for the marginal barrel. We seem to be pretty much at that level today.
We need to understand these dynamics better. What are the combined effects of flat oil prices and rising production costs, that’s where I think the challenge is and where our professional work is focusing on the macro side…to better understand what these trends are, what they mean, and how companies in the industry should respond to it.
I’ll give you an example. Normally, if you look at an oil production system, it tends to be symmetrical around the peak. The rate at which you approach the peak is the rate at which you depart from the peak. We haven’t done that. What we’ve done is that we’ve approached the peak and we’ve leveled out production, the so-called “undulating plateau”. But we’ve maintained that plateau by turning to non-oil liquids, by dramatic increases in upstream spend, and also by technological innovation related to hydrofracking. All of these, as of today, look to be running their course. Even shale oil. Yes, it will grow for the next few years from the three majors plays in the US, but the peak of production growth is already behind us in the Bakken, for example. On current trends, Bakken production will be increasing by single digits within two years. Not a tragedy by any means, but not enough to move the global oil supply at that time, either.
Of course, we have one more arrow in the quiver after that: government take. Governments typically take 60-90% of revenues of oil production. There’s nothing wrong with that, as in most cases the oil belongs to the respective government. But if the cost of production is increasing, then the value of reserves is falling. Put another way, current levels of government take, whether production or profit sharing, royalties, lease payments or taxes of any sort, are likely unsustainable. Oil companies will need tax relief in one form or another. Far from being able to raise taxes on oil companies, the sober reality is that governments are going to have to get used to getting less. Expect this theme to come front and center in the next couple of years. If government take is reduced quickly, then oil production levels could be sustained for a few more years.
But what then? What’s the outlook for oil production globally? Will production at the high cost producers just ease off gently, or will global production rejoin the anticipated trend line from a 2005 peak sharply and quickly? Will the major oil companies invest just a bit less, or do they start culling their new project list aggressively and without material replacement?
I don’t know what the answer to that is. But that’s what we’re trying to find out. That’s the focus of our macro thinking today.
Steven, thanks for your time and your thoughts.
Steven Kopits has been Managing Director for the New York office of energy business advisors Douglas-Westwood since 2008. He is solely responsible for the views expressed.
