The law of local causality says that what a thing is now causes what it does now. This is foundational to Bayesian inference about the world, because it allows us to look at what things are doing and make inferences about what they are.
The law of non-contradiction further restricts the terms we can use to describe "what they are": if a description requires a contradiction, it's a signal that we need to get creative and invent new concepts, and new terms to go with them (this gets muddled because sometimes we just paste the old terms on top of the new concepts.)
So there: how science works, in two paragraphs!
With regard to oil, part of the "doing" is the future price. Oil for June delivery (either Brent or WTI, both have the same pattern) is in the mid-90s USD, up from mid-60's before this stupid war. Call it a 50% increase, although it wobbles around. But I'm not building a model here, I'm generating a plausibility argument.
Another part of the "doing" is traffic through the Strait of Hormuz, which is within rounding error of zero. I use Marine Traffic to keep an eye on things, but there are a bunch of other similar sites that all tell the same story: nothing is moving. The effective traffic is zero.
While it varies with who you listen to and how you count it, this is taking about 14 million barrels of oil a day off the market, out of a total of about 100 million barrels a day global demand. Call it 15% for roundness.
How do we expect that to work itself out through the economy?
This turns out to be a hard problem:: the relationship between supply, demand, and price for oil is not well-understood. This paper from the Bank of Canada (2016) points out:Understanding what causes movements in oil prices is important for forecasting and for assessing the consequences for the global economy. Empirically distinguishing among competing sources of fluctuations, however, has proved to be a particularly difficult task...
Doing so requires identifying supply and demand shocks. A stark illustration of this identification problem is depicted in figure 1, which shows the scatter plot between monthly surprises in oil prices and oil production implied by simple univariate AR(1) regressions. The dots clearly show that oil prices and oil production are uncorrelated. [emphasis added]
The lack of correlation between prices and production could be the outcome of very different oil market configurations and, by implication, could be explained by very different combinations of supply and demand shocks. For instance... the supply curve could be inelastic while the demand curve could be very elastic. As a result, fluctuations in oil prices and oil production would be decoupled, with prices driven uniquely by demand shocks and production driven uniquely by supply shocks. A market characterized by a very elastic oil supply curve and a very inelastic demand curve... would also lead to a decoupling of movements in oil prices and oil production. In between, as depicted by the red solid lines, lies an oil market with a downward-sloping demand curve and an upward-sloping supply curve, which would imply that demand and supply shocks jointly affect oil prices and production. Those market configurations, which we picked among many for illustrative purposes, are equally consistent with the data but, as we will show, have very different implications not only for the causes, but also for the consequences of oil prices fluctuations.
The paper concludes that it's probably about 50/50 demand and supply variations driving prices: More precisely, our identification targets a supply elasticity of about 0.10 and a demand elasticity of about −0.10.
That is, when oil prices go up by 10% oil producers are incentivised to increase supply by about 1% ( a ratio of 0.1) and when oil prices go up by 10% oil consumers are incentivised to reduce consumption by about 1% (ratio of -0.1)
These numbers are based on fluctuations in supply and demand of +/- 1% or so, not the 10 to 15 percent supply shock we're currently experiencing, so take them with a grain of salt.
Notice that elasticities refer to how supply and demand change in response to prices, not how prices change in response to supply and demand. This has always struck me as weird: it's like economists think that money is the important thing, and not the actual stuff people make and use.
Money is a tool for shifting demand through time. It neither produces nor consumes, supplies nor demands. It's as if engineers spent all their time studying cogs and gears, and none of it looking at the motive power or the end-use of the machine in question. Are cogs and gears important? Sure. But a theory of machines that gave them pride of place and treated motive power and end-use as afterthoughts would be fairly limited.
As may be, prices have gone up by about 50% in response to this exogenous (outside "the economy") supply shock, which implies a reduction in demand of about 5% and an increase in supply of about 5%, still leaving the market short 5 million barrels a day.
That assumes these numbers from the +/-1% range apply to +/-10%, which is doubtful. It also assumes that it is physically possible to increase supply, about which there are some doubts. It's not like there is a lot of extra capacity hanging around, especially outside the Middle East, which is where it needs to be (probably) to have an effect.
If we want to be super-naive about stuff--and really, who doesn't?--we could say that with a price increase of 70% supply and demand shifts would bring the market back into balance.
Could that happen?
And what happens if it doesn't?
My prediction a couple of weeks ago was that it couldn't happen, and we'd see oil prices go through the roof and supply would simply disappear from some peripheral markets, because we would reach the boundary of acceptable adaptation, where oil demand did not drop no matter what the price was.
One thing that happens when irreducible demand outstrips maximum supply is that price volatility goes through the roof. We saw this happen to natural gas prices in the early 2000s, with prices bouncing around by as much as a factor of three over a matter of weeks.
I was expecting that to happen with oil prices, and sooner rather than later. It still might, if people can't adapt to the limited supply the world is now facing.
But early predictions often underestimate adaptive behaviours, which is why not a single one of the Club of Rome's dire predictions from 1973 have yet to come true half a century later, and in fact almost every one of them has gone the the opposite direction: less poverty, fewer people going hungry, longer lifespans, all that because people adapt to change.
One of my favourite expert predictions comes from Ahmad Murad, a senior executive of the Kuwait Oil Company, who predicted in 1991 that it might take five to seven years to extinguish the oil well fires started by Iraqi forces retreating from Kuwait. They were all out in less than seven months.
So now that oil prices have been more-or-less stable for a month despite the Strait still being closed and no progress on the mostly-imaginary "talks" between the US, Israel, and Iran, I've been wondering if I'm not simply wrong.
Unlike everyone else, I'm wrong all the time, sometimes by a great deal. Dunno how other folks manage it, but this is why I write things like this: to think through ideas in my slow and fumbling way.
The big force pushing the world away from the "yes, we have no oil" scenario is that everyone thinks it's coming, which means we're getting ahead of it.
Supply-wise, sitting on spare production capacity six months from now could reasonably be expected to be a gold mine if constrained supply leads to hyper-volatility. If you have oil to sell when prices are $300 a barrel, you win. Better get those old wells producing, just in case!
There are also non-Hormuz exits from the Middle East. Apparently there is a truck convoy running from Iran through Syria that is expected to reach 500 trucks a day from the current 150-200 level (as of April 1st), which amounts to about 0.1 million barrels of oil per day. It's not much, but if you get ten operations like that running it starts to cover 1% of global supply. Little ships can move large armies.
On the demand side, in the most prescient move of my life, I bought an EV on February 27th, and a lot of other people are going to do the same Real Soon Now.
In 2025 EVs cut oil demand by 2.3 million barrels per day, and that number is projected to double by 2030... but here's the thing: every single estimate of the rate of EV and other renewable/electro-tech adoption for the past 25 years has been low by factors of two, or more (except mine: I modelled my estimates of solar power growth on the rate of rural electrification in the early 20th century, which was just stupidly fast, and have been pretty accurate, but right now even my numbers are getting buried by reality.)
So given that experts are estimating a doubling in five years, I'm pretty confident in predicting it'll happen in a year to 18 months. I'm putting that "18 months" in as a caveat, because I don't see how it can possibly happen in a year, but I also know that what I or anyone else can possibly imagine is completely unrelated to what will actually happen, so honestly my bet is: one year. Which is another 2.3 million barrels of oil per day that don't need to be produced.
Technologies have learning curves. I'm old enough to have driven automobiles with chokes, distributors, and clutches. Fuel injection, electronic ignition, and automatic transmissions all took over in the '80s and a certain kind of automotive technology development simply came to a stop because we finally had solutions that were good enough. This is typical of human problems: there is a limit after which further development becomes uninteresting. Air travel hit this in the '70s, with the Concorde demonstrating the cost of supersonic travel wasn't worth it.
Hitting this limit in automobiles meant that when you sat down to design a new car, you weren't asking a whole raft of questions about how to best select from a large group of sub-optimal solutions to a bunch of persistent problems. You could take for granted that certain aspects of the basic platform were fixed.
My sense is that EV, solar, and battery technologies have all reached that kind of phase transition. It might not even be the last one. But LiFePO cells, in particular, have been just dramatic in the way they've swept all before them in a bunch of applications, eliminating the need for cobalt and nickel, which were two of the persistent irritants in the electrification of everything.
We've also learned that battery life is much longer than originally projected--my EV is ten years old, with no signs of deterioration--and recycling of batteries is better than 95% efficient, so people who think that that's an interesting thing to bring up every time batteries are mentioned are going to have to find something else to talk about. Clean air, maybe.
When a phase transition like this happens, costs go down, quality goes up, and the companies on the leading edge make truly astonishing amounts of money. This is one of the things that capitalism is actually good at, although it needs the hand of the state to make it happen at all (because without state intervention in free markets corporations can't exist) and to make it happen humanely (because have you ever met a capitalist?)
Also with regard to demand, fuel prices could easily push more people to put their foot down and insist on working from home due to excessive commuting costs, and companies might even let them do it. It's not like we don't know how.
Ground transport of one kind or another accounts for about 35% of all crude oil demand, with automobiles (passenger road transportation) coming in at 25% and trucking making up the balance. The rest is split between petrochemicals (15%), aviation (7%), maritime transport (6%), power generation (4%), and many other smaller industries and applications.
In the above I've focused on transportation because the Law of Optimization says "Reducing a big demand by a little is easier than reducing a small demand by a lot," and transport is where the big demand it. It's also where we happen to have the best technological positioning to really cut into it.
SO... this meditation was motivated by wondering why oil prices aren't going higher despite the Strait staying closed. Is it market irrationality, or am I missing something?
I think it pretty likely I was missing something: there is a plausible scenario where oil prices remain high but not astronomical, EV and other electrification reduces demand to meet supply, and the world moves on even while the Strait remains closed.
This could end up being the biggest favour Putin has done the West: by artificially constraining supply by pushing Trump to invade Iran, he has plausibly accelerated the electric revolution.
Talk about an own goal.
