Sunday, May 20, 2018

Crude Oil -- The Next Five Years -- SeekingAlpha -- May 20, 2018

Updates

May 30, 2018: WTI has plunged well below $72. Now down to $66. Discussion changed from "demand" to "supply." OPEC-Russia suggested "they" may add one million bopd production -- note: current global production is 100 million bopd -- as if one million additional bopd would make a difference.

May 22, 2018: Over at Bloomberg via Rigzone -- "forget $80-oil The big rally is in forward prices."
There are several reasons for the sudden surge in forward prices. Oil consumption is expanding much faster than anticipated, adding growth in two years that would normally take three. At the same, oil investment has dropped significantly over the past three years, particularly in projects that take longer to develop such as ultra-deep water offshore, raising doubts about future supply growth despite the gains in Texas, North Dakota and other U.S. shale regions.
Original Post

For the record, I really do not like these kinds of articles over at SeekingAlpha. I don't know why. I would "love" RBN Energy to do a similar story but something tells me there is a very, very good reason why RBN Energy does not (post such an article).

But a reader sent it to me. I read it quickly, and thought it interesting. I did have some snarky comments which I will post later.

This is a very, very long article. I'm impressed that the writer took that much time -- it must have taken a fair amount of time -- to write that article, and then post it for free (of course, he will be remunerated through "clicks").

Warning: beware of confirmation bias:
Confirmation bias, also called confirmatory bias or myside bias, is the tendency to search for, interpret, favor, and recall information in a way that confirms one's preexisting beliefs or hypotheses.
It is a type of cognitive bias and a systematic error of inductive reasoning.
People display this bias when they gather or remember information selectively, or when they interpret it in a biased way. The effect is stronger for emotionally charged issues and for deeply entrenched beliefs. Confirmation bias is a variation of the more general tendency of apophenia. 
First some definitions, and comments, some taken from the article, "Crude oil the next five years":
  • backwardation describes a market where spot and near-dated futures trade above longer-dated futures [higher prices today vs lower prices tomorrow]. The term structure is downward-sloping. An upward sloping term structure is called contango. Sometimes, part of the curve trades in contango and part of it in backwardation. In an environment of low inventories, the market typically trades in backwardation, and vice-versa
  • contango describes a commodity price curve where spot prices and near-dated futures trade below longer-dated futures [higher prices tomorrow than today]. The opposite is called backwardation, where spot prices and near-dated futures trade above longer-dated future.
  • From wiki: A market is said to be in contango when the forward price of a futures contract is above the expected future spot price. Normal backwardation, which is essentially the opposite of contango, occurs when the forward price of a futures contract is below the expected future spot price. 
  • so, I guess that means, in a contango situation, I have a contract to sell my grain six months from now for $5/bushel and now my financial advisor is telling me that forecasts suggest that when I take payment for my grain in six months, I will get a price higher than the spot price (whoo-hoo) -- maybe I should leverage my grain in the silo if in six months we are still in a contango situation -- I could buy cheaper grain on the spot market using my stored grain as collateral and use that cheaper grain to meet the contract...
  • so,  I guess that means, in a normal backwardation situation, I have a contract to sell my grain six months from now for $5/bushel and now my financial advisor is telling me that forecasts suggest that when I take payment for my grain in six months, I will get a price lower than the spot price -- maybe I should take profits now (if I can find a buyer who needs grain now) and then hope I can re-fill my silo later ... but it will likely cost more later which would be a problem if I  had to buy ... but if I produce it myself ...
  • From Investopedia:
  • Traders with access to physical oil and storage can make substantial profits in a contango market.
    Other traders may seek to profit on a storage shortage by placing a spread trade betting on the contango structure of the market to increase.
    Contango means that the spot price of oil is lower than future contracts for oil. A futures contract is a legal agreement to buy or sell a physical commodity at some point in the future. The spot market is the current cash trading price for that commodity.
    For example, assume that the spot price of oil is $60 a barrel. The future price of oil two months from now is trading around $65. This represents a contango futures term structure.
    At some point, the futures price will converge to the spot price, whether the futures price is above or below the spot price. In this situation, a trader who controls physical barrels of oil and has access to storage can easily lock in a profit.
    Going back to the example, the trader will sell a futures contract for delivery two months out at $65. By locking in that profit at the higher price, and then sitting on the physical oil for a couple of months, a trader can realize substantial gains. One futures contract of oil represents 1,000 physical barrels. On a full-size oil futures contract, that would represent a profit of around $5,000 for merely storing the oil for a couple of months.
  • I have great trouble keeping these two terms straight, even though it's incredibly simple in theory. I think folks like me misunderstand the concept because we don't actually have any contracts; we are simply comparing today's price with the expected futures price, and something tells me that is not the correct way to "imagine" or "understand" the terms.
  • a reader suggested it's easy to keep the terms straight: contango is the normal way of things with commodities that will deplete over time ... one would except prices to go up for commodities that deplete in the future...
Lots of data is presented. Trying to sort through this long article to get the writer's bottom line is difficult, but let's try.

First of all, and I'm thrilled with this. The writer starts off with global inventories of oil based on "days of supply." I've always felt this is the best metric for estimating whether there is a glut or a shortage of oil -- days of supply. When you go to the linked article, pay attention to the x-axis for "days of supply": for global inventories, between 2011 and March, 2018, the range has been from 38 days to 44 days.

From that graph, the writer says "global inventories have pushed the crude oil price curve into backwardation and spot prices sharply higher. [That suggests to me that going forward, we should see an increase in global inventories.]

From the article: the sharp decline in oil inventories over the past two years led to a massive shift in time spreads. In early 2016, the Brent curve was in steep contango. Prompt month prices traded $15 below the 5-year forward. As of today, Brent is trading $15 above the 5-year forward.

That tells me the amount of oil coming to market five years from now is expected to increase significantly.

But the very next statement:
The longer-dated price remained practically unchanged for the past two years. Hence, the entire move in the spot price was due to the shift in the curve, which was driven by inventory decline.
Importantly, the change in the spot price does not imply that the market somehow changed its view on how much it costs to produce oil. Longer-dated prices, which are set by the marginal cost of future supply, are still below $60/bbl. This means that the market sill believes that $55 - $60/bbl gives enough incentive to producers to make the necessary investments to meet future demand. The spot price rally, thus, was simply due to the decline in inventories.
And the trend has now shifted: inventories are increasing once again (see the first graph in the linked article).

Historically:
  • OPEC had large amounts of spare capacity; able to bring on-line in a matter of months -- sometime just weeks -- when a shortfall occurred, for example during the first Gulf war
  • non-Opec producers have almost always produced at maximum capacity
  • global major oil companies have no incentive to keep any capacity idle, unless operating costs exceed the price of oil (2008 - 2009; and again in 2014 - 2015)
  • that means, that over the short to medium term (five years), non-OPEC production follows a set path and is almost completely price-inelastic (price-inelastic: consumers buy about the same amount regardless of price, within limits)
The shale revolution:
  • changed that historical picture somewhat
  • producers still produce at capacity, but there is much more price elasticity (I think this is quite interesting)
This is key to the entire argument and one I often think about:
But while shale producers can ramp up production much faster than conventional non-OPEC producers, it would still take years to compensate for large shortfalls.
Read that again:
But while shale producers can ramp up production much faster than conventional non-OPEC producers, it would still take years to compensate for large shortfalls.
Before we go on, one comment and two questions to ponder regarding that last statement:
  • the comment: I don't buy that statement that "it would still take years to compensate for large shortfalls"
    • first of all, I can't think of anything that would result in large shortfalls that will last years (see below)
    • analysts consistently underestimate how fast US shale oil producers can respond to prices
  • two questions:
    • what is the definition of "large" shortfalls? and, 
    • what might cause a "large" shortfall (war -- regional/global); normal shipping channels upended (narrow straits closed; policy changes -- lower sulfur fuel for tankers); but, not much else -- and neither would last years  
And note: there can be a considerable time-lag between changes in price and changes in production; by the time US shale finally peaked in April, 2015, spot prices had been falling for almost year; by the time production bottomed in September, 2016, prices had been recovering for almost a year as well.

Now the discussion shifts:
OPEC's attempt to balance the market; OPEC sits on massive amounts of spare capacity; and, this led to a period of severe under-investment by non-OPEC producers
[This "severe underinvestment by non-OPEC producers" is something one of my readers frequently reminds me.]
The writer then gets to the nub: how much spare capacity does OPEC currently have?
  • On paper: 2.8 million bbls/day. Go to the linked article to see this discussion.
  • The writer's opinion: realistically, OPEC spare capacity is closer to 1.5 million bopd
There is increasing doubt that OPEC has much spare capacity. Time will tell whether the oil sector has under-invested over the past decade; they have invested much less than analysts think is required, but that doesn't mean the analysts are correct.

History:
  • the 1990's: a decade of very, very low prices for oil. Why? OPEC's spare capacity in the early 1990s, 10 million bopd
  • compare that to 2 million bopd today
  • Then, see discussion how shale production plays into this, at the linked article.
New: peak oil supply fears have been replaced with peak oil demand concerns. Look at this
The outlook from a huge shift in the transportation sector away from oil makes it difficult for producers to sanction a project with a 30-year life span. These projects cost billions of dollars, some even tens of billions, and could potentially become worthless. Instead, even the major oil producers push increasingly into the shale space.
Bottom line for conventional, 30-year projects: CAPEX is drying up (has already dried up, some would argue).

And then this, "the Art Berman" argument:
This will only be felt in a few years from now. Every new project that came only over the past few years, and all projects coming on-line for the next few years, were sanctioned a long time ago - some even prior to the financial crisis. Yet, despite the massive CAPEX spending prior to 2015, the new projects coming on-line hardly make a dent. In 2017, we saw only 0.2mb/d of non-OPEC production growth ex-shale (see Exhibit 7). And there is not any improvement in sight. 2018 and 2019 will look similar, and beyond that, non-OPEC ex-shale output will outright decline.
Again, the "Art Berman" argument:
On net, we expect non-OPEC supply to grow by 0.2mb/d in this year and next; after that, growth will slow down to less than 0.1mb/d by 2020. From 2021 onwards, non-OPEC supply will begin to outright decline.
Importantly, the chances for an upside surprise to this forecast are extremely slim, regardless of how prices develop over the next 5 years, because this was all set in motion many years ago.
Practically none of the large projects that are sanctioned now will come on-line before 2022. But if no new projects are sanctioned today, then non-OPEC ex-shale production risks falling off a cliff in 5 years from now (2023).
Then there is a great discussion in response to this question: with shale oil, does it really matter that there is no investment in large conventional projects?
The fact is, shale producers have shown an astonishing ability to grow production and an even more astonishing resilience to low prices. At its peak, US oil output grew by 1.5 million b/d, almost all from shale producers.
This rapid production growth eventually led to a price collapse in 2014, from $110/bbl to as low as $30/bbl. As a result, US production growth slowed to 1mb/d in 2015 and declined by 0.4mb/d in 2016. [Say what? The Saudis announced a surge in production in late 2014; began to take effect; was to last two years; but lasted into early 2017.]
But with prices now at over $60/bbl in 2018, US production is again growing at around 1.2 mb/d year over year. While production growth has slightly leveled off over the past months, we believe production will accelerate to around 1.5mb/d for the remainder of the year.
The writer asks: Can US shale oil sustainably grow at the peak rate of 1.5 million b/d? One of the arguments we often hear is that shale gas producers have clearly shown that the technology allows them to scale up production at will. However, shale oil and shale gas differ in key aspects.

The writer's opinion:
We estimate that with a continued production growth of 1.5 million b/d per annum, decline rates will reach 4 million b/d by 2022.
That means, in order to maintain production growth of 1.5 million b/d in 2022, drillers would have to bring 5.5 million b/d of new supply online that year - more than the entire current output.
The writer then gets into the refinery dilemma -- heavy vs light oil -- and the Canadian oil issue. The bottom line, according to the writer: very little US crude oil will be exported in the out years.

The writer's last line before going into "bringing it all together."
Overall, there are a number of limiting factors which will pose great challenges for US crude oil production growth over the next couple years. We do not think that US production can grow at 1.5 million b/d or even higher beyond 2019. However, we think that is what is currently priced into longer-dated prices.
Conclusions:
  • non-OPEC supply to grow by 1.7 million bopd over the next two years
  • demand to grow by 1.5 million bopd over the same next two years
However, in 2017, the global oil balance was in deficit by 0.5 million b/d. Commercial inventories are now at the low end of the range, meaning that stocks can't decline much further before refineries run into trouble. Consequently, in order to bring the market back to balance, global supply has to grow by 2 million b/d in 2018. Hence, OPEC will have to increase its production by 0.3mb/d this year, which means rather than building more spare capacity, OEPC will have to draw on some of its spare capacity  
The writer says:
We think the market is currently unaware of this. In our view, the prevailing market view is that OPEC needs to keep at least its current production curtailments in place for the 2018-2019 period to avoid a renewed inventory build.
The problem is exacerbated, in our view, by the fact that some OPEC producers are in outright decline. Production from Venezuela, for example, has been declining by 0.4mb/d year over year over the past six months, and the declines are accelerating and are currently closer to 0.6mb/d. This means that core OPEC producers (Saudi Arabia, Kuwait, Qatar, UAE) will have to bring back a substantial amount of spare capacity just to offset Venezuela.
And then a long discussion on OPEC.

Finally, prices -- this is hard to understand if you read quickly because of the way the paragraph is written:
Hence, from a fundamental perspective, we expect inventories will remain at these low levels throughout 2018 and 2019. This implies that the market will remain in backwardation. We also think there is very little downside for longer-dated prices from here. In fact, we think the next move in oil will be that longer-dated prices will start to move higher, and we explain that in more detail below. With a floor at the back end of the curve and strong time spreads, we therefore don't think spot prices (Brent) have much downside from here, and the risk is skewed to the upside.
Bottom, bottom line:
Our time spread model implies that roughly half of the 1-60 month time spread in Brent is driven by the extreme spec position. In other words, with a neutral spec position, our model would predict Brent spot prices to be roughly 10% lower. We think a sell-off due to an unwinding of the record spec net length would offer a great entry opportunity, as fundamentally, the market looks strong for the next 18 months.
Again, bringing it all together  (this article never seems to end):
The period from 2020-2022 will be marked by declining non-OPEC ex-shale production and a slowdown in US shale oil production. This means that OPEC will not just have to bring back all its spare capacity, it must build new capacity and it has to build it fast. We believe that this will prove challenging, and there is a very high risk that in a few years the market will have to be balanced by demand destruction again, which will be achieved through higher spot prices.
However, we think before that happens, longer-dated prices will have to start moving higher in order to trigger large-scale investments in future conventional non-OPEC production. At USD55/bbl, this is clearly not happening. A longer-dated price of USD55/bbl signals that the market believes that shale oil producers will deliver all future production growth. As we pointed out above, that is simply not possible. We need major investments in conventional projects, and we need new investments from OPEC for 2022 and beyond as well.
Hence, regardless of where spot prices are going in the near term, we think the next big move has to be in longer-dated prices. We think a price increase of at least 35% is needed to get enough investments to eventually catch up to demand. That is with no cost inflation. The longer the back end of the curve remains depressed, the worse the problem will get in a few years. 
****************************************
My Reply To The Writer At The Linked Site (Above)

So much to say, but I limited it to this much:
Excellent, excellent article.
At the end of the day, there are two unknowns: a) whether Canada be able to supply the heavy oil US refineries need; and, b) whether or not US shale lives up to Harold Hamm's expectations. In the big scheme of things, the drawdown of global oil inventory is a myth: number of days of supply (2011 - 2018) has ranged from 38 to 44 days; most likely will stay within that range. US supply is around 30 days of supply; well above the 19 - 20 days that was the historical norm years ago. 
Whether the excess global supply is 0.5 million bbls/day or 2.5 million bbls/day, makes little difference. Even at 0.5 million bbls/day at the end of one year, there's another 200 million bbls of excess global inventory, and this will go one year after year after year. All things being equal, adding 200 million bbls/day, at the end of one year, the inventory will increase from 38 days of supply to 40 days of supply.
Libya is no longer relevant; hasn't been relevant for years. At 0.5 million bbls/day production when global demand is 100 million bbls daily, Libya's output is irrelevant.It will be interesting to see if Venezuela's output drops to less than one million bbls/day. The tea leaves suggest that will happen. 
Goldmoney is very smart, when it comes to predicting the price of oil, to simply say the price of oil is likely to increase, without getting any more specific. It's a fool's errand to predict the price of oil. With the current glut of oil, and the likely persistence of that glut, I find it amazing the price of oil has increased to the extent it has. If Goldmoney's thesis is correct, oil bulls are going to do very, very well.
So much more could be written; I simply don't have time. 

No comments:

Post a Comment