The Mechanics of a Crude Oil Price Recovery
Prognostications of crude oil price are a staple of financial journalism- everyone has an opinion, from bankers to governments. While many unknowable forces like changing geopolitical tides or potential E&P bankruptcies may affect crude oil price, these factors must always manifest themselves as market signals that inspire physical players to make moves that create changes in the physical balance of the commodity. For example, a bankruptcy filing by a medium-sized oil producer may mean a pause in production and shut-in of unprofitable wells, which in turn leads to refiners who had been purchasing those barrels to look for new supply, competing with each other and perhaps causing a traders to sell from storage, causing a drawdown in inventory. Or to follow another chain, lifting of sanctions on Iran causes an increase in production and competition for US market share and perhaps an overall increase in imports. In turn the physical market balances by reducing marginal production or increasing injections into storage. We can even follow the market signals which will drive these changes as they happen.
Though there is further complexity inherent in a physical market (such as fungibility of differing grades and the market-sensitivity of storage costs and limitations on storage capacity), it’s not impossible to follow the fundamentals to begin to know where exactly in the cycle of commodity price we may be now.
Traditionally analysts have done this by following a path that looks something like:
- Drop in prices
- Drop in rig count
- Drop in production
- Price recovery
But the confounding feature of the oil and gas industry in recent years is the linear relationship above has fallen apart- that is, a drop in price did not cause a drop in supply. Below is a chart of crude oil production and Cushing prices since 2013:
The persistence in production levels has a lot to do with increased productivity in the oil patch, meaning that advanced completion technologies have allowed onshore producers to do more with less, exponentially increasing productivity per rig. This means that the traditional method of using rig count as a leading indicator for production decline, market rebalancing, and price recovery has fallen apart.
Since this relationship has broken down and we can no longer rely on rig count and production volume as a shorthand indicator, we’re left to look more broadly at the steps of a physical market going from oversupplied to balanced:
- Drop in production capacity
- Flattening of contango on the futures curve
- Drop in production
- Drawdown of inventories
- Price recovery
Let’s look at these in detail and see where we are.
Drop in production capacity. Here is where rig count is still important, but we must combine with rig productivity. As I’ve already mentioned, E&Ps have exponentially increased rig efficiency. This is due to “high grading” the most productive acreage and shortening the days needed to drill each well, therefore each rig (and therefore each drilling dollar) has been able to drill more and better wells and generate more and more production.
If we look back and combine the nation’s overall production per drilling rig, we can combine the efficiency data with the precipitous drop in rig count to come up with an “effective rig count” as a proxy for production capacity. When these productivity gains begin to level out, we will be able to apply a constant efficiency factor to adjust rig count to predict production capacity. At this time though, it looks like rig efficiencies, at least in the case of crude oil drilling, are still increasing. This means that ever-steeper drops in rig count are required to drop production.
What we can say is that production capacity has just recently begun to decline, even though the number of rigs drilling for crude oil has been falling since the end of 2014. What we will be watching closely for now is a leveling off of rig efficiency, combined with a continued drop in rig count. Only then will we see a real decrease in capacity, and a subsequent significant drop in production.
Flattening of contango on the futures curve. Typically in any commodity, oversupply results in a steep contango in the futures curve, meaning that near term prices are discounted to future months. What this means in the physical market is that traders are incented to store barrels rather than selling in the current market. As long as premiums for crude in later months exceed storage costs, total inventories will increase. In February of this year, the difference between the March 2016 contract and March 2017 was at times over $11. Keep in mind this was in a $38 per barrel oil market, representing nearly a 30% premium on one year’s time. The situation drew the attention of financial media, with Bloomberg quoting a BP executive as saying that “every tank and swimming pool” would be filled with the commodity in the second half of the year.
At the time of this writing, the difference between the prompt (June 2016) contract and one year out (June 2017) has dropped to $3.19. Which is a relief, even though I don’t own a swimming pool.
How does this compare with normal market conditions, not reflecting massive oversupply? As in any commodity, the difference between one month and the next should reflect storage costs as well as any decay in value due to inflation. The current 7% premium on a year’s time in the crude market seems to be approaching normal, suggesting by this measure that the peak of the oversupply situation may have passed.
Drop in Production. As shown in the first chart with EIA production data, overall US crude oil production is just beginning to fall. However, in order to achieve a drop in production volumes substantial enough to begin to balance the market, we need to see a leveling off in rig efficiency accompanied by a continued fall in rig count.
Drawdown of inventories. A “super-contango” market like that described above obviously leads to increasing inventories in a big way, and we can see this reflected in the data:
Though the contango has abated, traders are still injecting crude oil into storage, resulting in an unprecedented level of commercial crude stocks. To give some perspective, the previous record high inventory level before the shale boom was in 1990, at just over 387 million barrels. The 540.6 million barrels of inventory reported last week is not only above all previous record levels, but also exceeds previously available storage capacity, filling a majority of capacity additions in recent years.
So why is storage inventory continuing to increase even though the contango has flattened?
It appears that the crude being stored is not of the exact quality needed by today’s refiners. There is now an oversupply of the types of light crude produced in the Eagle Ford and Bakken that do not work well for refiners to produce profitable “middle distillates” such as fuel oil, jet fuel, and diesel. For a thorough explanation as to why there is now limited opportunity to blend this light crude at Cushing to meet WTI spec, check out RBN’s detailed blog post on the subject.
This represents a precarious situation for a physical market balance. While the range of grades of crude produced in the US and Canada aren’t all completely fungible, space in storage tanks is. This means that an actual drawdown of stored WTI crude may be masked by continuing injections of lighter grades that have no current physical buyer. In that scenario (which may be starting to play out right now), buyers of WTI crude will continue to bid prices higher, even though the market looks oversupplied because of extremely high inventory.
Price recovery. At this point in time, the prospects for different types of crude oil begin to diverge. It does appear that WTI may already be moving toward equilibrium based on the above factors. Contango has flattened, reducing the incentive to stockpile. Total production has leveled off and begun to decline, which should accelerate once drilling efficiency gains are maxed out. But the overall numbers mask an underlying disparity.
Because of the limitations on blending and lower potential refining margins, lighter crude oil from shale tends to be discounted from the benchmark WTI. For example, Flint Hills Resources (a Koch Industries-owned refiner) is publishing a current bid for Eagle Ford light crude of $34.75 per barrel and Bakken light sweet at $35.75. This is in comparison to the $42.75 that the refiner is willing to pay for WTI.
Even while WTI continues its recovery, discounts for barrels from shale could grow until either exports or producer shut-ins accelerate to bring physical supply of lighter grades back to or below refinery needs.
Investors in E&Ps producing crude oil from shale may be disappointed to find that an increase in WTI prices does not result in proportionate revenue growth.