Canada has spent the past few years telling us that high-cost debt is a problem.

Ottawa calls some of it predatory lending. The federal government lowered the criminal interest rate to 35 percent APR in 2025. Provinces regulate payday lenders. British Columbia and Alberta have even created an official regulatory category called “high-cost credit.”

Fair enough.

There is only one rather large problem.

Canada regulates entry into expensive credit far more seriously than it considers the exit.

A borrower can end up with a perfectly legal installment loan carrying an interest rate in the low 30s. Their circumstances can then improve. They can have a job, income, and the ability to service a conventional loan. They may actively want to replace that debt with cheaper financing.

That should be a good thing.

But throw an unresolved tax balance into the picture, and the machinery can start working backward.

CRA may know about the tax debt. The taxpayer may be cooperating. CRA may have granted time to get things in order. There may be no garnishment, no lien, no seizure, and no enforcement action at all.

A mainstream lender can still look at the CRA exposure and decide it wants nothing to do with the refinancing.

The expensive lender, meanwhile, is quite happy to continue collecting.

So the borrower keeps paying 30-odd percent rather than 10 or 12 percent. Their cash flow remains weaker. Their tax debt takes longer to clear. More of their income disappears into interest. Financial stress continues.

And the federal government waits for its money.

It is difficult to think of a more self-defeating arrangement.

Canada has built a second credit system

Canada quite rightly takes pride in the stability of its banks.

But beneath the mainstream banking system lies another credit market.

It lends to people the banks will not or no longer lend to. There is nothing inherently wrong with that. Riskier borrowers cost more to serve and default more often. Subprime lending, by itself, is not synonymous with predatory lending.

But the numbers matter.

British Columbia regulates credit above 32 percent APR as high-cost credit. Alberta does much the same. These are not necessarily $300 advances until payday. The category can include installment loans, lines of credit, vehicle-related lending, and other products that can remain on a household balance sheet for a long time.

The federal government itself has acknowledged that high-cost installment loans became a fast-growing, widely held form of debt and that borrowers could become caught in refinancing cycles.

Before Ottawa changed the law, some installment loans were advertised at rates approaching the old criminal threshold. The new ceiling of 35 percent is considerably better than what came before.

But “better than 48 percent” is an extraordinarily low bar for a developed economy.

Consider what 34 percent actually means.

Put $50,000 into expensive credit at roughly 34 percent, and the opening annualized interest burden is around $17,000.

Put the same $50,000 at 12 percent, and it is around $6,000.

That difference, roughly $11,000 a year before considering amortization, has to come from somewhere.

It comes from household income.

It comes from savings that never happen, purchases that are postponed, retirement contributions that are missed, taxes that take longer to pay, and sometimes from another loan.

People talk about interest rates as percentages. Households experience them as missing money.

That is why the real public policy question cannot simply be, “How high a rate should a lender be allowed to charge?”

It should also be, “How does a borrower get back out?”

Canada has surprisingly little to say about that.

Other countries made different choices

There is no perfect international model, and Canada should resist the usual temptation to find one attractive statistic from Europe and pretend an entire financial system can be imported with it.

The United States, for example, can be considerably worse. Consumer credit regulation varies enormously by state, and payday lending in some parts of the country remains extraordinarily expensive.

But several other wealthy countries have taken a much harder look at the economics of high-cost borrowing.

The Netherlands currently caps the maximum cost of consumer credit at 14 percent annually.

Japan went through its own ugly period of consumer finance excess. Its reforms eventually brought maximum rates generally into the 15 to 20 percent range, depending on loan size, and introduced limits on how much consumers can borrow from moneylenders relative to income.

Britain took another route. Rather than pretending expensive short-term credit would disappear, the Financial Conduct Authority imposed rules on its structure. High-cost short-term lenders face a daily price cap, limits on default charges, and, crucially, a total cost cap. A borrower cannot ultimately be charged more in interest and fees than the original amount borrowed.

The FCA has since said those reforms saved borrowers substantial amounts of money and reduced debt problems.

Australia has also imposed affordability restrictions around some forms of small, expensive credit.

Canada is not uniquely permissive. That would be an easy claim to knock down.

What is striking is that Canada still tolerates long-term consumer credit at rates that would look extraordinary in several peer economies, while simultaneously declaring that predatory lending is a national problem.

And then we make escaping that credit harder than it needs to be.

CRA makes the contradiction worse

Tax authorities need power.

Nobody seriously proposes letting taxpayers simply decide whether paying taxes is convenient.

CRA needs the ability to collect unpaid taxes, garnish funds where necessary, register claims, and act against people who refuse to engage.

But there is a world of difference between someone hiding from the tax authority and someone working with it.

Our financial system does a poor job of recognizing that difference.

Suppose a taxpayer owes CRA money and is actively dealing with it. CRA has agreed to give them time. No enforcement is underway.

The taxpayer also has expensive private debt.

Refinancing that debt would immediately improve their cash flow and increase the amount available to pay CRA.

This is exactly the kind of transaction a rational creditor should want to happen.

Yet the existence of the CRA exposure can frighten away the cheaper lender.

The irony becomes almost comic when you read CRA’s own guidance. In considering payment arrangements, the agency can expect taxpayers to look at ways of rearranging their finances or borrowing to meet their obligations.

Borrow, says one part of the system.

Not with that CRA issue, says another.

The high-cost lender then continues to collect.

There is a name for this sort of thing in ordinary life.

A Catch-22.

The IRS has already worked this out

The most uncomfortable comparison is not the Netherlands or Japan.

It is the United States.

The IRS is not famous for being soft.

Yet American tax administration explicitly recognizes that preserving the government’s creditor position at all costs can sometimes be economically stupid.

The IRS can subordinate a federal tax lien.

In practical terms, it can allow another lender to move ahead of the government when it makes financial sense.

Its own guidance gives the obvious example: a taxpayer refinances at a lower interest rate, creating enough additional monthly cash flow to increase payments toward the tax liability.

Read that carefully.

The IRS has formally contemplated exactly the situation that exposes the weakness in Canada’s approach.

A tax authority can improve its recovery by helping the taxpayer lower the cost of other debt.

The United States does not forgive the tax. It does not surrender its right to collect.

It simply recognizes arithmetic.

The IRS can even withdraw a public notice of a tax lien in some circumstances when doing so will help the taxpayer pay the government more quickly.

That is what sophisticated creditors do.

They care about recovering money, rather than merely possessing the strongest piece of paper in the room.

Canada should shamelessly steal the idea.

This is also a productivity problem.

Canada is currently obsessed with productivity.

With good reason.

The Bank of Canada has used unusually strong language about our productivity performance. Canadian output per worker has lagged. Business investment has disappointed. Living standards have struggled to keep pace with those south of the border.

We discuss the problem as though productivity exists only in factories, software companies, and spreadsheets maintained by economists.

It also exists inside people’s heads.

Financial stress is distracting.

The Financial Consumer Agency of Canada has published research saying precisely that. Employees worried about money report difficulty focusing at work. Financial problems consume working hours. People check balances, move payments, call creditors, worry about which bill is next, and lose sleep thinking about money.

Anyone who has ever been under financial pressure knows this without needing a government study.

Debt follows you to work.

Someone worrying whether a 34 percent loan payment will clear on Friday is not necessarily giving an employer their best thinking on Thursday afternoon.

Obviously, high-cost debt does not explain Canada’s productivity crisis. That would be absurd.

But it is equally absurd to spend billions of dollars worrying about productivity while tolerating avoidable financial structures that consume workers’ time, attention, and resilience.

The Bank of Canada wants more output from every hour worked.

Perhaps we should stop designing systems that fill some of those hours with anxiety about debt.

Then there is the money that never gets spent.

The productivity effect is only half the story.

High-cost debt also absorbs disposable income.

An extra $10,000 or $11,000 a year going to interest is $10,000 or $11,000 that cannot be used for something else.

It cannot buy a new washing machine, pay a contractor, or be spent in a restaurant.

It cannot fund a course, contribute to an RRSP, support a small business, pay for a family trip, or build an emergency fund.

And yes, the lender does something with the interest it receives. This is not an argument that interest somehow disappears from GDP.

The point is vulnerability and opportunity cost.

A heavily indebted household has far less room to absorb a shock. When something goes wrong, spending gets cut quickly because so much income is already spoken for.

The Bank of Canada has repeatedly warned about the wider economic consequences of heavily indebted households.

The consequences eventually reach government as well.

Ottawa collects GST when Canadians consume taxable goods and services. Most provinces collect sales taxes of one form or another. Governments collect corporate taxes from the companies households spend money with. They collect personal income taxes from the people those companies employ.

Less economic activity means less tax revenue.

Finance Canada says this plainly in its own fiscal modeling. When economic growth falls, personal income tax, corporate income tax, and consumption-tax revenues suffer.

Again, nobody can calculate a neat number and say, “Canada lost $X billion because people have expensive personal loans.”

Economies do not work that neatly.

But the direction is not mysterious.

If government policy unnecessarily forces households to spend cash flow on avoidable interest rather than productive activity or tax repayment, somebody should ask whether the system is working.

The provinces are part of the problem

Canada’s federal structure makes all of this messier:

1. Ottawa sets the criminal interest framework.

2. Provinces regulate much of consumer lending and consumer protection.

3. Payday lending operates largely through provincial regimes.

4. British Columbia and

5. Alberta regulate high-cost credit.

6. Other provinces have their own rules.

7. CRA operates under a separate set of powers.

8. Banks are federally regulated but make individual underwriting decisions.

9. Credit unions sit in another part of the structure.

10. Insolvency law comes back to Ottawa.

11. Everyone regulates their own slice.

12. The household gets the whole pie.

13. That is the institutional failure.

A provincial government can license a high-cost lender. Ottawa can describe predatory lending as a problem. CRA can give a taxpayer time to deal with a tax balance. A mainstream lender can nevertheless decide that the CRA issue makes refinancing too uncomfortable.

Every institution can say it followed its rules.

The consumer remains trapped at 30-odd percent.

This is a very Canadian kind of policy failure. Nobody is necessarily doing anything obviously wrong inside their own box.

The boxes do not fit together.

And eventually, some people break.

Canada has formal solutions when debt becomes unmanageable:

1. Consumer proposals.

2. Bankruptcy.

3. Licensed Insolvency Trustees.

By the time someone reaches that point, however, the economic damage is already well underway.

Creditors can take losses. The borrower suffers years of damaged access to credit. Consumption is constrained. Stress rises. A CRA receivable that might once have been paid in full may become part of an insolvency proceeding.

That is hardly a triumph of tax collection.

Governments should care deeply about preventing viable taxpayers from becoming insolvent.

It is cheaper.

It preserves tax revenue.

It preserves creditworthiness.

It preserves consumption.

It preserves productivity.

Most importantly, it recognizes the obvious fact that someone who cannot refinance a manageable problem can eventually end up with an unmanageable one.

Stop lecturing borrowers and build an exit

Canadian consumer policy still has a paternalistic streak:

1. Read the disclosures.

2. Make a budget.

3. Understand your APR.

4. Shop around.

5. Borrow responsibly.

Fine.

But if someone is already paying 32, 33, or 34 percent, telling them that high-interest debt is expensive is not consumer protection.

They damn well know.

The question is whether they can escape.

Canada needs to start measuring success by how many consumers graduate out of high-cost credit, rather than simply by how carefully lenders disclose their price.

CRA should create a meaningful managed-compliance status for taxpayers who cooperate, file what is required, comply with an agreed arrangement, and are not subject to enforcement.

That status should carry weight in the credit market.

Canada should also examine an IRS-style subordination regime where refinancing demonstrably improves CRA’s recovery.

The conditions could be strict. CRA could insist that part of refinancing proceeds go directly to the Receiver General. The taxpayer could be required to remain current with filings. The lender could be regulated. The transaction could be assessed individually.

None of that requires forgiving a dollar of tax.

It requires the government to behave like a rational creditor.

The provinces should also ask a more uncomfortable question about the high-cost lenders they regulate:

1. What is the graduation rate?

2. How many borrowers actually move back into mainstream credit?

3. How long does the average customer remain in high-cost lending?

4. How many loans are refinanced into another high-cost loan?

5. How much principal is repaid compared with interest and ancillary charges?

6. If governments do not know, perhaps they should.

Because licensing a debt trap does not make it any less of a trap.

Follow the dollar

Canada has a habit of discussing these issues in silos:

1. Productivity goes to the Bank of Canada and economists.

2. Tax collection goes to the CRA.

3. Consumer protection goes to the provinces.

4. Banking goes to Ottawa.

5. High-cost lending goes to another regulator.

6. Mental health goes to health ministries.

7. Insolvency goes somewhere else.

8. The person carrying the debt experiences all of them at once.

So follow one dollar instead.

1. A Canadian earns it.

2. The government wants part of it in taxes.

3. The household needs part of it to live.

4. The economy would benefit if part of it were spent, saved, or invested.

5. Instead, an unnecessarily large portion goes to interest on a high-cost loan.

6. The borrower finds a cheaper lender.

7. The cheaper lender sees unresolved CRA exposure and walks away.

8. CRA waits.

9. The expensive lender keeps collecting.

10. The worker remains stressed.

11. The employer loses focus and productivity.

12. The household spends less.

13. The federal government collects its tax debt more slowly.

14. Federal and provincial governments collect less from the economic activity that never occurred.

15. If the situation eventually ends in insolvency, everyone takes another hit.

Yet every institution involved can insist that it followed the rules.

Canada is spending enormous political energy searching for productivity, affordability, and economic growth while leaving obvious friction embedded in the financial lives of its citizens.

We regulate predatory lending because we understand that expensive debt can trap people.

Then, astonishingly, we tolerate a system that can make it harder for those same people to leave.

That does not protect taxpayers.

It can punish them.

It does not necessarily protect the public purse.

It can delay repayment to it.

And it is difficult to call a system prudent when the taxpayer, the employer, and the government can all lose while the most expensive creditor in the room keeps getting paid.

That is not consumer protection; it is bad economics.

Further reading

• Government of Canada, Budget 2024: Doing More to Crack Down on Predatory Lending. Read the federal government’s discussion of high-cost instalment lending

• Financial Consumer Agency of Canada, Understanding Payday Loan Use and Perspectives. Read FCAC’s 2025 research on high-cost credit and payday lending

• Government of British Columbia, High-cost credit products. See British Columbia’s high-cost credit rules

• Government of Alberta, High-cost credit regulation. See Alberta’s definition and regulation of high-cost credit

• Bank of Canada, The Productivity Problem. Read the Bank of Canada’s discussion of Canada’s productivity weakness

• Financial Consumer Agency of Canada, Why Your Employees’ Financial Well-being Matters. Read FCAC’s material on financial stress and workplace productivity

• Internal Revenue Service, Publication 784, Certificate of Subordination of Federal Tax Lien. Read the IRS guidance, including the lower-interest refinancing example

• UK Financial Conduct Authority, High-cost short-term credit. Read the FCA’s current high-cost credit rules and assessment

• Netherlands Authority for the Financial Markets, Maximum cost of consumer credit. See the Netherlands’ current maximum consumer-credit rate

Years ago, during treatment at Combat Stress, someone gave me a piece of advice I have never forgotten.

“Don’t be a steel bar. Be a rubber ball”.

I cannot remember the precise conversation around it. I remember the sentence.

At the time, I was not especially impressed.

I had spent eleven years as a Royal Navy officer. Steel sounded reassuring. Our ships, which were also our homes, were made from it.

Nobody had ever unveiled a new warship by comparing it favourably to a child’s toy.

But they were not talking about ships. They were trying to explain mental resilience.

A steel bar can withstand enormous pressure while remaining straight and apparently unchanged. Once that pressure exceeds its limit, however, it bends permanently or snaps.

A rubber ball changes shape. It absorbs the impact. It may fly off in an unexpected direction, but it survives.

The lesson was not that I needed to become softer or weaker.

It was that strength and rigidity are not the same thing.

Read the sea

The Navy had already taught me a version of this, although I had not understood it in quite the same way.

You can leave harbour with a chart, a forecast, a set of orders and a carefully prepared plan. Then the weather changes. Equipment fails. The situation shifts by the hour. Another vessel is not where it was expected to be.

The sea does not care how long you spent preparing the briefing. It has not read the minutes and will not be attending the follow-up meeting.

A competent officer does not stand on the bridge insisting that the original course must remain correct because several senior people approved it.

You reassess. You protect the ship and the people aboard it. You preserve the mission, but you may have to change the route.

Sometimes you turn back.

At sea, that is not automatically seen as weakness. It may be good seamanship.

This becomes strangely controversial once we move ashore.

Human beings become attached to plans, particularly plans we helped to create. We defend the forecast because we wrote it. We protect the strategy because we presented it. We cling to the future we imagined because acknowledging that it has changed feels like admitting we have lost something.

Sometimes something has been lost.

Refusing to acknowledge it does not bring it back.

I have spent much of my life being the person who keeps things moving. In the Navy, in companies and in relationships, competence became part of my identity.

When something stopped working, my instinct was rarely to put it down.

It was to work harder. Longer hours. More energy. More of myself.

That can look like resilience from the outside. For a while, it can even produce results.

It can also hide the fact that a system no longer works.

Reading the market

I have spent much of my career in shipyards, ports, infrastructure, energy and transportation. These are physical businesses. Weather matters: equipment, labour, regulation and customer behaviour matter.

This can be irritating for people who have become emotionally attached to a model.

Models are useful. Plans are necessary. Forecasts create discipline. But they are descriptions of what we think may happen, not blood oaths.

The mistake is not assuming that it will later prove wrong. That is unavoidable.

The mistake is defending it after reality has disproved it.

Most new businesses do not survive indefinitely. In one large American cohort, almost half were no longer operating after five years and nearly two-thirds after ten.

Those figures do not explain why each business closed. Some will have been sold, merged, or deliberately wound down. Still, they should make every founder humble.

Starting something is not proof that the world needs it, wants it or particularly cares.

A company can generate enormous activity without discovering whether anyone genuinely wants what it is selling.

Large companies are no safer from this mistake. Sometimes their size makes it harder to see.

Information travels through layers of people whose careers may depend upon the message remaining reassuring. By the time reality reaches the boardroom, it has often been edited, formatted and placed beside a green arrow.

BlackBerry is a familiar Canadian example.

Its devices were secure, reliable and beloved by business users. The physical keyboard became part of the product’s identity. As late as 2010, BlackBerry still held more than 40 percent of the American smartphone market and close to 20 percent globally.

The company understood email, security and the needs of its established customers brilliantly.

What it took longer to recognize was that the smartphone was becoming something else. It was no longer principally a telephone with excellent email. It was becoming a platform around which people would organize much of their digital lives.

BlackBerry kept strengthening the fortress while the market moved beyond its walls.

It survived only after accepting that its future lay in software, cybersecurity and embedded systems rather than handsets.

The Swiss watch industry offers an even more interesting example because it recovered without abandoning what made it distinctive.

By the beginning of the 1970s, Swiss watchmaking was built around generations of mechanical skill and an international reputation for precision. Then quartz watches became cheaper, easier to mass-produce and more accurate.

Swiss researchers had helped develop early quartz technology, so this was not simply a failure to notice an invention.

The deeper failure was to understand what that invention would do to the market.

Between 1970 and the middle of the 1980s, the number of Swiss watch companies reportedly fell from more than 1,600 to fewer than 600. Employment dropped from approximately 90,000 people to 33,000.

The industry had not forgotten how to make excellent watches.

It had temporarily lost sight of what kind of excellence customers would still be willing to pay for.

Its recovery came through adaptation at both ends of the market. Swatch made quartz watches inexpensive, colourful and fashionable. Luxury manufacturers increasingly presented mechanical watches as objects of craftsmanship, heritage and engineering rather than merely the most efficient way to tell the time.

They stopped trying to beat quartz at being quartz.

They changed their understanding of the product while preserving something essential about themselves.

That is the distinction I missed when I first heard the rubber-ball advice.

Adaptation is not becoming whatever the market, the boardroom or another person demands.

It is deciding what must remain and what can change.

Knowing when to turn back

Business culture celebrates persistence.

Keep going. Push through. Believe when nobody else does.

Most of those stories are told by people for whom persistence eventually worked.

We hear less from the equally intelligent people who continued pouring money, health and years of their lives into ventures whose underlying conditions had changed, or had never been favourable in the first place.

There is a difference between difficulty and futility.

Most worthwhile ventures encounter resistance. Customers take longer to arrive. Capital becomes scarce. Technology fails. Competitors react. The economy turns.

Headwinds alone are not a reason to stop.

But headwinds are information.

A business may discover that the market is too small, the economics do not work, the regulatory burden is disproportionate, or the timing is simply wrong.

A founder may also realize that the personal cost has become greater than any plausible reward.

Walking away can be rational. It preserves capital, and capital is not only money. It is also attention, health and the ability to begin again.

Projects inside large companies acquire the same emotional protection.

Money has already been spent. A team has been formed. A strategy has been announced. Stopping would make the original decision look wrong.

So another quarter is funded. Another consultant is hired. Another explanation is written.

The project continues not because the evidence supports it, but because stopping would require somebody to admit that the evidence no longer does.

The money already spent cannot be recovered by spending more. The years already invested cannot be reclaimed by surrendering the next five.

A sailor does not continue into worsening weather merely because the ship has already travelled a long way.

Distance travelled is not proof that the destination remains reachable.

Sometimes good seamanship means turning around while there is still fuel, daylight and room to manoeuvre.

There can be dignity in a well-managed ending.

It means telling people the truth, meeting your obligations and salvaging whatever remains useful. It also means learning from the venture without letting it drain the energy needed for whatever comes next.

Not every plan deserves another year of someone’s life.

The future that actually arrives

Relationships have plans too.

We imagine where we will live, what our careers will become and what kind of life we will build together. We make assumptions about money, timing and who will carry which responsibilities.

Then life happens.

A career change. Money becomes tighter. A parent becomes unwell. A business struggles. A child needs more attention. Someone’s health deteriorates.

The future that once felt certain becomes a series of difficult conversations held when neither person is at their best.

It is easy to love the future two people designed together.

It is far more difficult and arguably builds a stronger, more resilient bond when circumstances change, and we are forced, as a team, to redesign it together.

Intent does not erase impact. But impact does not erase intent either.

The difficulty begins when change becomes one person’s job.

One person becomes the planner, provider, problem-solver, and the one expected to resolve each crisis. The other waits for the future to arrive.

That arrangement can survive for quite a long time, particularly when the person carrying it is competent.

That said, a partnership cannot remain healthy if the person carrying the weight dismisses every expression of fear as ingratitude.

Both people have to be willing to look at the life that has actually arrived.

This is not what we expected.

What is still possible?

What can each of us contribute?

Those questions may not sound romantic, but they can be more loving than another promise about a future neither person knows how to reach.

Sometimes two people can redraw the plan. Sometimes the headwinds pass. Sometimes care, humility and changed behaviour allow something damaged to recover.

Sometimes they do not.

There are also times when continuing consumes the energy both people need to recover. Accepting that does not make what they shared meaningless.

What now?

In my last essay, I wrote about granite: the need for a centre strong enough to withstand pressure without mistaking every tremor for the end of the world.

Granite and rubber may sound like contradictory materials.

I no longer think they are.

Our values may need to be granite.

Our methods cannot be.

That is a balance I have often struggled to find.

I have bent when I should have stood firm. I have stood firm when wisdom required movement. I have mistaken rigidity for conviction and exhaustion for resilience.

No instrument on the bridge tells us precisely when perseverance becomes denial. No alarm sounds when loyalty becomes self-erasure.

We have to pay attention to the growing distance between what we say and what we know.

Children often manage this more naturally than adults.

They can be deeply disappointed that something has not happened, then become absorbed by what happens next. Adults may spend months litigating the case for what should have happened.

A child asks a simpler question:

“What are we doing now?”

My son is eight.

I want him to have principles and to tell the truth when life does not follow the plan. I also want him to change his mind without feeling humiliated, to apologise without believing he has become smaller and to encounter a future he did not expect without becoming frightened of it.

I want him to have a centre without building a prison around it.

Years ago, I thought Combat Stress was teaching me how to return to the person I had been before difficult things happened.

I am no longer sure we ever return in quite that way.

We read the sea, the market and the people we love. We change course when we can and turn back when we must.

I am learning that the aim is not to remain unchanged.

It is to preserve enough of myself to sail again.

I was talking with my son recently about truth.

Not truth in the grand philosophical sense. We were not sitting beside a mountain monastery while a bell rang in the distance.

We were talking about what happens when someone says something that does not feel right, and the adults around you either do not notice or do not want to deal with it.

He is at the age where children begin to understand that adults are not always reasonable.

That is a difficult discovery.

When you are young, you assume adults know what they are doing. They are bigger. They have cars, passwords and bank accounts. They decide when bedtime is. Surely, somewhere along the way, somebody gave them the answers.

Then you slowly realize that adults can be frightened, insecure, selfish, confused and completely wrong.

I told him that one of the most useful things I have learned, embarrassingly late in life, is that arguing with an irrational person rarely makes them rational.

It usually just makes both of you tired.

Imagine sitting in a negotiation.

I say:

“Two plus two equals four.”

Someone across the table says:

“No. Two plus two equals seven.”

There are two paths available to me.

The first is the path I followed for most of my life.

I pull out the evidence. I produce a deck. I show the arithmetic. I explain the history of numbers. I commissioned an independent review. I find three experts who agree that the answer is four.

Then I prepare for the follow-up meeting because someone has raised concerns about whether the number two was properly consulted.

I am joking.

Mostly.

The danger is that I can become so determined to prove that two plus two equals four that I allow someone else’s irrationality to control my energy, my time and eventually my judgment.

I leave my own centre.

I enter their confusion.

And once I am inside it, I begin reacting to them rather than thinking for myself.

There is another path.

I can stay calm and say:

“Interesting. Can you explain how you got to seven?”

My son laughed (even an eight-year-old can see what happens next.)

If the answer is based on evidence, we can examine it.

Perhaps I missed something.

The problem may be different from the one I thought we were solving.

Being centred does not mean assuming I am always right. It means being secure enough to test my own thinking without surrendering it.

I have believed it too, which can help your audience feel understood and less defensive about their own certainty.

I have believed it too.

I have been certain when I should have been curious. I have defended positions because I had invested too much of myself in them. I have mistaken resistance for evidence that I was brave, and confidence for evidence that I was correct.

Sometimes I was right.

Sometimes I was in a circular pattern of self-conviction.

Humility begins by admitting that possibility, which can make your audience feel more trusting and open to growth.

There are, however, also moments when an answer is built on anger, status, insecurity or wishful thinking.

The killer line, and I mean “killer” affectionately, is:

“I see your point. You should study that option. I am open to discussing it again when you have the facts and data.”

Ouch.

No shouting.

No dramatic exit.

No fifty-slide presentation entitled Why Four Remains the Preferred Strategic Outcome.

Just a question, a boundary, and a return to the centre-this practice helps maintain emotional balance and fosters healthier relationships.

The addiction to being understood

For years, I believed that if I could explain something clearly enough, people would understand, and if they understood, they would make a rational decision. If they did not understand, I assumed I had not explained it properly.

So I tried again.

And again.

I changed the language. I added evidence. I was expecting objections. I made the argument simpler, then more detailed, then simpler again.

This can look like persistence.

Sometimes it is.

It can also become an addiction to being understood, which hampers growth; recognizing this helps you develop humility and effective communication skills.

The uncomfortable truth is that some people do understand.

They disagree.

Others understand but do not want to acknowledge what the facts would require of them.

And some are searching for an explanation that allows them to preserve their comfort, power or existing position.

That is their choice.

It does not have to become my emergency.

This has been one of the hardest lessons of my life.

Being correct does not mean you have to exhaust yourself proving it.

You’ll be able to say the facts.

Please feel free to ask the question.

You can remain open to new evidence.

Then you can stop.

That is not surrender.

It is discipline.

Nothing left to prove

During the same conversation, my son and I watched a video online, attributed to Anthony Hopkins, reflecting on life at 87.

Something was arresting about hearing an older man speak about detachment.

Not the brittle detachment of someone who has stopped caring, it is the type that comes when you finally understand how little of life can be controlled by force.

What I heard was both stoic and simple: stop handing your peace to people who have not earned the right to hold it. Stop measuring yourself through reactions you cannot govern. Let people misunderstand you when being understood would require you to abandon yourself.

At 87, those ideas sound different.

They are no longer productivity advice.

They feel like an accounting.

How much of a life can be spent proving, fixing, explaining and carrying before a person finally asks what all that effort was for?

The video includes a line from T. S. Eliot’s The Love Song of J. Alfred Prufrock:

“I have seen the moment of my greatness flicker.”

There is something painful in that line.

The recognition that time has passed while we were hesitating, explaining, seeking permission or waiting for someone else to understand us.

But watching the video with my son, I heard something else in it, too: a flicker is not necessarily an ending.

Perhaps greatness was not intended to be the title, the position, the rescue or the moment when everyone finally admitted that you were right.

Perhaps it is the moment you stop giving your life away in the effort to prove that it mattered.

My son found the video inspiring.

So did I, although perhaps for a different reason.

He was hearing a way to live, while I was hearing a description of the life I had not always known how to live.

That may be one of the strange gifts of talking honestly with your children.

You begin by trying to give them the lesson.

Halfway through, you realize you are still learning it yourself.

The danger of certainty

There is another danger hidden inside all of this.

Certainty.

Conviction can be tested. Courage can admit fear.

The dangerous kind of certainty turns disagreement into disloyalty and correction into humiliation.

I have seen what that does to organizations. At first, it can look like strength:

The leader is decisive. The direction is clear. The room moves quickly.

Then something changes.

People stop bringing bad news.

They learn which facts are welcome and which ones are not.

They begin sanitizing their thoughts before they speak.

The smartest people become silent while the most compliant become louder.

Eventually, everyone knows that two plus two equals seven.

So they stop discussing the arithmetic and begin discussing how enthusiastically they support the answer.

That is how companies lose touch with reality.

Not all at once, but one unasked question at a time.

Hubris is expensive because it can make people feel like they’re in defeat.

A humble leader can say:

“I may be wrong”.

An insecure leader hears those words as a surrender of status.

So the mistake remains. Another decision is built on top of it, then another.

Soon, the organization is protecting the person who made the first mistake rather than confronting the mistake itself.

Families can drift into the same pattern.

A parent becomes unable to apologize, or a partner must always be right.

A child learns that honesty creates conflict, while silence preserves peace.

Questions begin to feel dangerous.

Affection becomes conditional on agreement.

The family may still look intact from the outside, but the people inside it no longer feel safe enough to tell the truth.

I do not say this from some enlightened position; ego and hubris are, from my point of view, the human condition.

I deeply understand the feeling of needing to defend a decision because admitting doubt would feel like losing authority. I know how quickly a person can confuse being challenged with being disrespected.

That is why humility matters.

Humility is not thinking poorly of yourself.

It is still teachable.

It is holding a strong view without making your identity depend on it.

It is saying:

“This is what I believe, based on the evidence I have. Show me what I may have missed.”

That creates a different kind of wealth.

The wealth of people who trust you enough to tell you the truth.

The wealth of a child who is not afraid to ask a question.

The wealth of a team that catches the mistake before it becomes a crisis.

The wealth of a relationship where love does not equate to passivity and silence.

The wealth of changing your mind without feeling that you have lost yourself.

Real wealth is not the certainty that you are right.

It is having enough security within yourself to remain open.

You cannot lay bricks on sand

Our conversation moved to Maslow’s hierarchy of needs.

I showed my son the familiar pyramid. Food and shelter are near the bottom. Safety. Belonging. Esteem. Then, somewhere near the top, the possibility of becoming fully yourself.

I explained that human beings do not always move from one level to the next. Life is chaotic.

But the basic idea matters.

You cannot keep building indefinitely on foundations that are collapsing.

“You can’t lay bricks on sand,” I said.

He laughed again.

“That makes sense.”

Of course it does.

A child understands instinctively what adults spend years trying to avoid.

If the foundation moves every day, everything built above it becomes unstable.

You can still build. Human beings are remarkably resilient.

But you will spend enormous energy keeping the walls upright.

For much of my life, I became exceptionally good at holding up walls.

At work.

In relationships.

In institutions.

In moments of crisis.

I could enter a chaotic situation, absorb the pressure, identify the problem and begin fixing it.

People valued that.

I valued it too.

Being the fixer gave me purpose. It made me useful. It made me need.

There is a quiet seduction in being needed; being needed can be deceitful and feel very similar to being loved.

The fixer’s bargain

Every organization has fixers.

They are the people who stay late because the process doesn’t work.

They correct someone else’s mistake before anyone notices.

They translate confusing instructions into something the team can operationalize and then execute.

They absorb the pressure created by poor decisions.

They protect junior people.

They fill gaps that should never have existed.

Organizations celebrate these people.

They call them resilient.

Dependable.

Mission-focused.

A safe pair of hands.

Sometimes they even give them an award before asking them to fix three more things.

But there is a hidden devil’s bargain.

The organization can keep avoiding its problems because the fixer keeps compensating for them.

The more competent the fixer becomes, the longer the system can pretend to function.

Competence becomes a substitute for dysfunction.

Eventually, the fixer begins to disappear inside the role.

You stop asking what you need.

You stop noticing how tired you are.

You tell yourself the next crisis will be the last one.

You believe that if you can stabilize everyone else, there will finally be time to stabilize yourself.

Unfortunately, the reward for rescuing a system is usually another system to rescue.

I learned this very late.

Perhaps I am still learning it.

Humility also requires me to acknowledge that fixing everything was not a harmless act.

Every crisis I absorbed was one that the system did not have to confront.

Every person I worked around was protected, at least temporarily, from the consequences of their own choices.

Sometimes I preserved the mission.

Sometimes I helped the dysfunction survive.

While it is extremely difficult to admit, it is also freeing.

If I helped sustain the negative pattern, then I should not be powerless to stop repeating it. Right?

The virtues that can consume us

My early professional life was shaped by military officer training.

You learn courage.

Loyalty.

Duty.

Service.

Endurance.

You learn that the team comes before the individual.

You learn to keep going when you are tired, frightened or uncertain.

These are serious virtues, and I remain grateful for them.

However, few people take the time to teach us what happens when those virtues lose their boundaries.

Courage can become unnecessary exposure.

Loyalty can become complicity.

Endurance can become silent deterioration.

Service can become self-erasure.

Self-sacrifice can become a habit long after the emergency has passed.

You become so accustomed to carrying weight that you stop asking whether the weight is yours to carry.

A leader who never considers their own needs is not infinitely strong; they are drawing down a finite reserve while pretending it is renewable.

Eventually, the cheque arrives at the table.

Sometimes in exhaustion.

Sometimes in anger.

Sometimes in illness.

Sometimes, in the sublime realization that you have spent years living according to everyone else’s needs and no longer know what your own life feels like.

Self-care is often discussed in language that makes “serious” people recoil.

It can sound indulgent or superficial.

Buy a candle. Take a bath. Download an app.

Self-care is the disciplined protection of the human being who must make the decisions.

It is stewardship, agency.

Destroying yourself for the mission does not automatically serve the mission.

Sometimes it simply leaves the mission with fewer capable people.

Staying inside your own circle

I explained the locus of control to my son in simple terms.

There are things we control.

There are things we influence.

And some things remain outside us, regardless of how desperately we want them to change.

• We control whether we tell the truth.

• We control the questions we ask.

• We control how we treat people.

• We control whether we act from courage or fear.

• We control how much of ourselves we continue giving to a situation.

• We do not control whether another person listens.

• We do not control whether they become self-aware.

• We do not control whether they choose truth over comfort.

• We do not control whether they understand us.

That last one can be extraordinarily painful, especially when you care deeply about said person.

Staying centred does not mean becoming cold.

It does not mean withdrawing from everyone or pretending not to care.

It means caring without abandoning yourself.

It means refusing to make someone else’s emotions, confusion or instability the organizing principle of your life.

You can love someone and still recognize that their choices are theirs.

You can be loyal and still have boundaries.

You can forgive and still leave.

You can understand why someone behaves as they do without being harmed by it.

Compassion does not require participation.

Fear sometimes wears noble clothes.

We also talked about why people remain in situations that make them unhappy.

• Sometimes it is love.

• Sometimes it is loyalty.

• Sometimes it is a genuine commitment to working through something difficult.

Those are great things.

But sometimes what looks like loyalty is fear.

• Fear of being alone.

• Fear of financial insecurity.

• Fear of judgment.

• Fear of starting again.

• Fear that the unknown will be worse than the pain we already understand.

Human beings often prefer familiar suffering to unfamiliar freedom.

We create noble explanations for staying.

We say we are being patient, responsible, and committed. Strong.

Sometimes we are.

Sometimes we are afraid.

I have called in for fear loyalty, and I have called the inability to let go commitment.

I have stayed because I did not know who I would be if nobody needed me to fix anything.

The answer is not to hate ourselves for the years we spent afraid, which can create a prison in our own minds.

The answer is to notice, to learn and then course-correct.

I understand why I made that choice with the information, wounds and needs I had at the time.

Now I know more, and now I can choose differently.

Regret can teach us to improve; self-hatred rarely does.

Happiness is fleeting, and that’s ok

At some point, my son and I began talking about happiness.

I told him I no longer believe happiness is a logical, immutable state.

Happiness is wonderful; however, like all other emotions, it is fleeting.

You can wake up happy, become frustrated shortly after, laugh at lunch with your friends, feel anxious in the afternoon and experience complete peace while brushing your teeth.

I believe that contentment is very different from happiness:

Happiness is a passing cloud.

Contentment is the sky above.

Being centred does not mean that you never become angry, frightened, excited or sad.

You feel it.

You listen to what it is telling you.

Then, eventually, you return.

Back to the centre.

Back to the granite.

Back to the part of you that remains steady while the cloud moves on.

That centre, that boulder, takes time to form.

It is formed through truth, boundaries, self-respect, honest relationships and work that has meaning.

The courage to admit when something is hurting you.

The humility to admit when you are wrong.

The wisdom to understand that you cannot fix everything.

However, even a boulder in the middle of a field is not invulnerable.

It is not perfection.

It is a foundation that is strong enough to survive change without mistaking every tremor for the end of the world.

What I wanted my son to understand

I did not want to teach my son how to win arguments.

I wanted to teach him that he does not have to enter every argument he is invited into.

I wanted him to know that asking a question in a measured and calm manner can be infinitely more powerful than raising his voice.

I wanted him to understand that adults can be wrong, and he is not responsible for fixing them.

He can observe.

He can ask.

He can trust his feelings while remaining curious about the facts.

He can love people without carrying their entire emotional world.

He can be kind without surrendering himself.

And I wanted him to understand something I wish I had learned much earlier:

Your value does not depend on how much pain you can endure for others.

There will always be people who insist that two plus two equals seven.

You may be tempted to stay all night explaining the arithmetic.

Sometimes the most rational response is simpler.

Ask how they reached seven.

Listen.

Remain humble enough to discover that you may be wrong.

And when the evidence is absent, allow yourself to stop arguing.

Then go home.

You have foundations to build.

Thames Water is, at time of writing, on its twelfth consent request to its super senior creditors. Twelfth. The company that supplies water to a quarter of England is being kept alive month to month by waiver letters while a creditor consortium hammers out a recapitalisation that would write off billions in debt and lock dividends until 2035.

How did it get here? Not through the villains you would expect. Forget nationalised sloth, and forget the theory that private ownership is rotten by design. It got here because between 2006 and 2017 its owners loaded a monopoly water utility with debt, drew out roughly £2.7 billion in dividends, and ran gearing above 80 percent on an asset whose entire economic logic is boring, regulated, inflation-linked cash flow spread over a century.

Picking up the thread

Last time out, I argued that we spent a generation aiming our best technology and our most ambitious capital at capturing attention, while housing, energy, healthcare and the plain cost of living went begging. The fix, I mentioned, starts with where we point our effort.

Suppose we win that argument. Suppose serious money turns away from the attention economy and shows up at the fundamentals, at pipes and ports and grids and care homes. What then?

Then the shape of the money starts to matter as much as its destination. The attention economy and the infrastructure wrecks of the last two decades are the same illness at different scales: capital optimising for its own metabolism rather than for the thing it is bolted to. An engagement algorithm strip-mines your attention because the business model demands it. A seven-year fund strip-mines a water network because the return target demands it. Pointing capital back at the real economy is necessary, and on its own it fixes very little. The money also has to fit.

The lazy debate

Every time an infrastructure asset fails in private hands, the same argument restarts. One side calls privatisation the original sin and demands renationalisation. The other side points at Network Rail, or any number of state-run utilities, and asks whether the public sector’s record is really the one you want to defend. Both sides are arguing about the label on the shareholder register.

That label is the least interesting thing about an owner. Governments have starved assets of maintenance capital for decades, because the capex kept losing to hospitals inside the budget cycle. Private owners have poured superb lifecycle investment into assets for decades, because their liabilities ran out fifty years and boring was precisely what they had paid for. Ownership form, in other words, tells you remarkably little. Fit tells you almost everything: whether the shape of the money suits the asset it is bolted to. And that has next to nothing to do with the name on the share certificate.

That is the whole argument. Match the capital to the asset, and the label on the register turns into a footnote.

Four dimensions of the match

There are four dimensions you can measure, and score, before a deal closes.

Duration.

A water network, a port, a transmission grid runs on a clock measured in generations. Point five-to-seven-year fund capital at it and the fund is obliged, by its own wiring, to be eyeing the exit multiple before the first regulatory period is even out. Point a pension fund at the same asset and the exit is not a concern at all, because for a book matching sixty-year liabilities there is no exit to plan. Identical asset, and the two owners behave nothing alike.

Return expectation.

A regulated asset base earns a regulated return, mid-single-digits real on a good day. A fund underwritten to fifteen or twenty percent gross cannot reach that by running the asset better.

The only road to the number is the balance sheet: debt piled up at holdco level, securitisation, dividend recaps, asset sales dressed up as portfolio optimisation. The maths does the forcing. Nobody at Macquarie woke up one morning wanting to break Thames Water.

Operating competence.

Infrastructure is an operating business wearing the clothes of a financial asset. Somebody still has to dredge the channel, replace the mains, sit across the table from the union, keep the drydock cycle turning. Govern that through a quarterly reporting pack and the physical asset will keep surprising you, usually with a bill attached. Put former operators on the board and in the asset team, though, and you tend to see the trouble coming three budget cycles out, while it is still cheap to deal with.

Tolerance.

This is the dimension nobody prices. Regulated infrastructure comes bundled with politics, public accountability, a hostile press, labour complexity and the occasional select committee summons. Treat all of that as friction for the advisers to handle and it will eat you alive. Understand it instead as the weather you sail in, the way a fishing fleet reads a sky, and you price it in and get on with the work.

Score well on all four and private ownership works, often better than the public version. Discipline shows up. Investment gets made on engineering logic instead of the electoral calendar, and decisions move at commercial speed rather than committee speed.

Score badly and the failure runs to a script. Debt where the thing needed resilience. A reporting pack where it needed someone who could smell trouble three cycles out. Cash walking out on a dividend schedule while the mains age on a slower one. And underneath all of it the exit clock, ticking against an asset that only ever rewards an owner who stays.

The obvious objection

Here is the problem with the thesis as stated: matched capital risks being unfalsifiable. If any failure can be redescribed after the fact as a mismatch, the framework explains everything and predicts nothing. Thames failed, therefore the capital was mismatched. Circular, and useless.

The answer is that the mismatch is visible before the failure, in the deal documents, if anyone bothers to look. Four tests you can run before the outcome:

Hold period against asset life.

Fund life under ten years against an asset with a fifty-year investment cycle. You do not need to wait and see.

Underwritten return against regulated return.

If the gap between the IRR in the investment committee paper and the allowed return on the regulated asset base runs to more than a few hundred basis points, the difference will be manufactured on the balance sheet. There is nowhere else for it to come from.

Distribution policy against capex cycle.

Dividends front-loaded against a maintenance profile that is back-loaded. If the money leaves faster than the asset ages, someone else is holding the renewal bill.

Where the debt sits.

Debt at opco that funds the investment is healthy. Debt stacked at holdco and midco, invisible to the regulator, sitting there only to gear up the equity, is the tell. Ofwat regulated Thames Water Utilities Limited but the problem lived three entities up the structure, in Kemble.

Score a transaction on those four before close and you are making a prediction you can be wrong about, which is the only kind worth anything. I would happily run the test against any concession signed in the next twelve months and check my working in 2031.

A short casebook

Run the tests backwards over the last twenty years and the pattern holds with uncomfortable consistency.

Chicago, 2008.

The city leased its parking meters to an investor consortium for seventy-five years in exchange for about $1.16 billion up front, money that was largely gone plugging a budget hole within a few years. The city’s own inspector general later concluded the concession was worth far more than was paid. Seventy-five years of pricing power on street parking, sold to close one budget cycle. Test one fails in both directions at once: the capital’s horizon was long, the seller’s was two years. Mismatch cuts both ways, and the grantor can be the one holding the wrong end of it.

Southern Cross, 2011.

Britain’s largest care home operator was assembled under private equity ownership, its freeholds sold and leased back with upward-only rent escalators, the operating company floated in 2006. When local authority fees flattened after 2008, occupancy softened and the rents kept climbing on schedule, because that is what the documents said they would do. The company collapsed with roughly 31,000 elderly residents in its homes. Test three, failed in advance and in writing. Anyone who read the leases against the fee outlook could have called it.

The quiet counterfactual. Canadian pension funds and Australian superannuation capital have held airports, toll roads, ports and rail links across five continents for two decades, mostly without incident and mostly without headlines, which is the entire point.

Boring is the product. These owners pass all four tests by construction: liabilities measured in decades, return targets a regulated asset can comfortably meet, operating teams on staff, and a stubborn patience with the political weather. That missing scandal is the design working, engineered in at the deal table decades ago.

Matched private capital exists, in size

So none of this is an argument against private capital. Some of the best infrastructure owners in the world are private. Patient money with in-house operating teams, single-digit return targets and no exit clock owns a meaningful share of the world’s essential assets, quietly compounding.

Even private equity is not disqualified by category. Open-ended core infrastructure vehicles, continuation structures with genuine permanence, firms that have built real operating platforms rather than financial holding structures: these can pass the four tests, and some do. The thing that fails is narrower and more specific: a closed-end, 2-and-20 buyout structure pointed at a regulated monopoly, dead on tests one and two before the ink is dry.

What regulators should screen for

Which brings us to the practical conclusion. Regulators and concession grantors spend enormous effort on ownership labels, fit and proper person tests, and change of control provisions. They spend remarkably little on the shape of the capital. The screening should invert.

At award or change of control, test the four dimensions explicitly. Cap holdco gearing, or at the very least require structural transparency all the way up the stack. Lock distributions to delivery of the capex programme. Require demonstrated operating capability, not an operations and maintenance subcontract waved about as a fig leaf. Match the concession length to the capital’s real duration rather than its marketing materials.

And on the grantor’s side, remember Chicago: a government selling a century of cash flow to fix this year’s budget is running its own duration mismatch, and no amount of diligence on the buyer will fix that.

The bitter irony of the Thames Water endgame is that the rescue terms now on the table, dividend locks to 2035, gearing forced down to the lowest in the sector, creditor equity locked up through the regulatory period, are exactly the matching mechanisms that should have been conditions of ownership in 2006. The regulator is retrofitting a matched capital structure at the point of maximum distress, twenty years and one near-insolvency late.

Capital has no character of its own. It gets shaped. Duration, return target, operating depth, appetite for the public realm: all of it is a design choice. Which loops back to where the last piece left off. Once the money does turn toward the fundamentals, someone has to check its shape at the door, because misdirected capital wastes a decade and mismatched capital can break an asset. For twenty years we have argued about who should own infrastructure. The duller and more useful question is what the money was built to do. Ask it at the start, and Thames Water is a footnote instead of a case study.

If this was useful, send it to someone who sits on an investment committee, a regulator’s board, or a city council that is about to sign a concession. They are the ones who get to ask the question while it still costs nothing.

Further reading

The case for the prosecution

- Brett Christophers, Our Lives in Their Portfolios: Why Asset Managers Own the World. The most complete map of who actually owns the pipes, and the sharpest version of the argument this article pushes back against.
- James Meek, Private Island: Why Britain Now Belongs to Someone Else.
- Brendan Ballou, *Plunder: Private Equity’s Plan to Pillage America. A prosecutor’s brief, care homes and all, useful precisely because it refuses to distinguish matched from mismatched capital.

Inside the machine

- Sachin Khajuria, Two and Twenty. The defence, written from inside, and worth reading before dismissing.
- David Carey and John Morris, King of Capital. How Blackstone actually works, including why the fund structure drives the behaviour.
- Joyce Moullakis and Chris Wright, The Millionaires’ Factory. The Macquarie story from the Australian side, essential background to the Thames chapter.
- Ludovic Phalippou, Private Equity Laid Bare. The academic who checks the industry’s homework on returns, which bears directly on test two.

The assets themselves

- Bent Flyvbjerg and Dan Gardner, How Big Things Get Done. Why infrastructure delivery fails, from the person with the largest project database on earth.
- John Kay, Other People’s Money. Financialisation as intermediation gone feral, the wider frame this article sits inside.
- Dieter Helm’s water regulation papers (dieterhelm.co.uk, not a book, worth the detour). Two decades of warnings about gearing and the regulatory asset base, most of them ignored on schedule.

The primary record

- The National Audit Office and Ofwat publications on the financial resilience of the water sector. Dry as the subject demands, and where the dividend and gearing numbers live.
- Chicago Inspector General’s 2009 report on the parking meter concession. Fourteen pages that should be laminated and handed to every mayor considering an asset lease.

I am done. Done with the doomscroll, done with the little hits of despair between emails, done with reading forty-seven takes on why everything is a mess before I have finished my breakfast. It was bringing me down. Worse, it was boring, too many people are pointing at the ship going down, fewer people bother with the shoring up the hull.

So this one is about repairing the ship we call Earth.

The weird part

We are soaked in technology, and the basic deal of a life has somehow gotten worse.

I can summon a car with my thumb, point my phone at a Portuguese menu and read it in English, video-call someone on the far side of the planet, generate a business plan before lunch, watch every film ever made from the sofa, buy shares in the time it takes to sneeze. Miraculous, all of it.

And yet a home feels further away than it did for my parents. Groceries have gone strange. Childcare costs about the same as a second mortgage, except the mortgage at least builds equity. Retirement has quietly become a rumour. Do everything right, follow the boring competent middle-class script to the letter, and it still does not buy the security it once did.

Nobody believes the inflation number

This is why people laugh at the official figures. Tell someone inflation is three percent. Watch them look at their rent, their insurance renewal, their grocery bill, their tax code and their mortgage rate, and say, with total sincerity, respectfully, have you lost your damn mind.

They are not wrong. The consumer price index measures something real, it just does not measure the thing people feel, which is the cost of getting and keeping a stable life in the place they happen to live.

Look at what got cheap: Information, streaming, gadgets, choice, the ability to have a thing delivered before you have finished wanting it. Abundant, and getting more so by the month.

Now look at what got brutally expensive. Housing. Healthcare. Childcare. Education. Insurance. Energy. Getting from A to B.

What we did is make the toys cheap and the foundations really really expensive.

The good kind of cheap

And here, against my instincts, I turn hopeful.

Two things made life expensive. One is the money question, who ended up holding it, the other is a targeting error. We took the best machines ever built and aimed them at attention, advertising, finance, and getting a burrito to a door in nine minutes. Some of that is useful. Some of it is a laugh.

A fair chunk is an expensive piece of technology for making people lonely and slightly worse at concentrating.

Fine. Now imagine aiming that same firepower at the cost of being alive.

There is a good kind of cheap that has nothing to do with cutting corners. Refrigeration made food safe to keep. Electrification rewired the house. A steel box, the shipping container, took the cost of moving anything anywhere and knocked it flat. Sanitation has saved more lives than every general in history managed to spend, and got no ticker tape parade. Then vaccines and antibiotics, which turned what used to be a death sentence into a prescription at the local doctor.

By the time any of these finished arriving, they had stopped feeling like technology. They just felt like normal life, which is the highest praise the stuff can earn.

That is the cheap worth chasing, and it is a far more complex provlem than teaching a phone to nag you about your calendar.

Now, the actual solutions

So what would it look like to aim technology at affordability instead of engagement? A partial list, because I find lists calming.

Start with housing, the one that ruins people. We still build homes about the way we did in 1975, one hammer at a time, gated by permits that outlast the pouring of the concrete. Factory-built and modular homes can cut the cost and the wait, if the rules allow it. In an expensive city the house is the cheap part, what you are really buying is the land, the permission, and the years.

A piece of software that turns a two-year permit into a two-week one would do more for a young family than every granite countertop ever installed.

Then energy, which is tricky, because it hides inside the price of everything else. Cheap clean power is a discount on food, heating, transport and factories all at once, since all of them run on it. Solar and storage have already dropped further and faster than the analysts forecast. What is left in the way is grids, connections and paperwork.

I am aware that is the least thrilling sentence in this essay. It also happens to describe one of the largest bills on the continent.

Healthcare barely needs a robot. A frightening amount of medicine is people filling in forms about medicine, so start there. Notes that write themselves. Diagnosis that comes back the same afternoon. The dull preventive work that stops a problem before it takes up a well needed hospital bed.

Government is the one I can speak to, having worked inside it and grown the grey hairs to show for it. The state is very often the reason a thing is slow and expensive.

A public sector that had been properly digitized would hand every one of us, left and right, a few thousand hours of our lives back off the phone queue. No politician runs on it. Everyone privately wants it.

Food and transport are the same story told twice, so tell it once: build in a bloody contingency! Sorry but a drought on one continent should not show up as panic pricing on yours, that’s just bad planning. And please, forget the flying cars. The prize nobody puts on a magazine cover is duller and far larger. Buses that come. Trains that run. Ports that clear. Freight that crosses a country without a dozen middlemen each clipping a coupon. The unglamorous work of moving things is where prices are quietly won and lost.

And schooling. One good teacher, explaining one thing well, at the scale of the internet, so that a clear explanation stops depending on your postcode.

The catch, and it is a big one

Cheaper essentials are necessary, on their own they fix nothing. There is a failure mode here and it is the obvious one; if all this lovely cheapness arrives while ownership of themachines producing it stays bottled up in a few hundred hands, the savings never reach you. They get skimmed on the way down. You end up with a cheaper product and a thinner slice of the world that makes it.

Technology without ownership is simply wealth extraction and concentration.

So the other half of the answer is ownership. Wages that track productivity, pensions and public wealth funds that hand ordinary people a real stake in the upside.

Employee ownership, co-operatives, housing models where a person builds equity instead of paying down a landlord’s. Wider access to the assets that do the producing, so that when the system gets more efficient, there is true stakeholder alignment.

We are often tricked into forgetting that the real power base in the world should be the consumer and the tax payers (ie the citizens of the world), not the companies, not the government.

Get that half right and the cheapness becomes freedom. Get it wrong and it becomes a subscription with no cancel button, which is arguably worse than the problem we started with.

What I have decided

Somewhere between the doomscroll and my morning walk, I made up my mind.

The future does not have to be the sad montage. Aging populations, unaffordable homes, a warming planet, furious politics, and a small circle of people who own the rest of us.

That montage is available, and it has excellent production values. I have just stopped believing it is the only edit.

There is another version, where we decide on purpose that the point of all this technology is not to make life faster, stickier and more profitable for people who already have plenty. The point is to make a decent life easier to afford for everyone else.

That is a future worth building. It is also, and I cannot stress this enough, a far better use of a phone than the doomscroll.

Further reading

If the piece did its job, here is where to go next. I have kept it to writers who think we can fix this, grouped to match the sections above:

Why the essentials got expensive

• The Cost Disease (2012), William Baumol. The whole argument in one book: the work that resists automation, care, teaching, a string quartet, gets dearer precisely because everything else got cheaper. If you read one thing on this list, read this.

• “Considerations on Cost Disease” (2017), Scott Alexander. The internet’s clearest short version, awkward charts included. Free, and done before your coffee is cold.

The miracles we stopped noticing

• The Rise and Fall of American Growth (2016), Robert Gordon. The definitive account of the great cost collapses, plumbing, electrification, refrigeration, and the uncomfortable case that they were a one-time windfall. Long. Worth it.

• Energy and Civilization (2017), Vaclav Smil. Everything runs on energy, and Smil will not let you pretend otherwise. Essential for anyone who likes their optimism sourced.

Point the machines at the essentials

• Abundance (2025), Ezra Klein and Derek Thompson. The current statement of the build case, more homes, more power, more of the physical stuff, and the rules that strangle all three. The book this article is arguing alongside.

• “It’s Time to Build” (2020), Marc Andreessen. Short, a little smug, and the essay that put the word “build” back in circulation.

• Construction Physics (newsletter), Brian Potter. Why a house costs what it does, and why the factory has not eaten homebuilding yet. The best thing going on the housing half of my argument.

• Build, Baby, Build (2024), Bryan Caplan. The case for tearing up housing rules, drawn as a comic, so it goes down without effort.

Cheap clean power

• How the World Really Works (2022), Vaclav Smil. The realist’s field guide to energy and materials. Read it, lose some illusions, stay hopeful anyway.

• Not the End of the World (2024), Hannah Ritchie. The data case that clean and cheap is winnable. The exact antidote to the doomscroll.

The catch: who owns the machine

• Technofeudalism (2023), Yanis Varoufakis. His name for extraction wearing a nicer logo, a handful of platforms owning the rails while the rest of us pay to stand on them.

• The Capitalist Manifesto (1958), Louis Kelso. The original argument for putting capital in ordinary hands, decades before anyone called it predistribution.

• “A Social Wealth Fund for America,” Matthew Bruenig (People’s Policy Project). The public-ownership version, one fund, everyone a shareholder. Short and specific.

And because someone always asks about the flying cars: Where Is My Flying Car? (2021), J. Storrs Hall, which is really a history of why the ambitious future stalled, cars or no cars.

On Saturday Iran declared the Strait of Hormuz shut again, less than two days after the signing, blaming Israeli strikes in Lebanon and American bad faith on a truce whose ink had not dried. Washington disputes it the way it disputes everything now, Central Command swearing the traffic runs clean and the Vice President on television bragging about a record sixteen million barrels. Iran says the venue is closed. The promoter says the show goes on.

Twice now since April. Closed, reopened, closed, the news doing the one thing the Pixies trained a whole generation to expect. Loud, quiet, loud. The April ceasefire died in ten days. This one made it two.

I grew up an hour up the coast from Seattle, which in the nineties meant the whole thing came through the walls whether you asked for it or not. CFOX taped off the radio with the DJ talking over the intros, a Matthew Good single wedged between catalog Nirvana and whatever Seattle had coughed up that week. The decade handed kids like me two anthems and both were about being a loser, one from Beck and one from Radiohead. What the anthems left out is that the kid in the corner, the one with the flannel and no plan, the one with nothing riding on the release, knows a lot more than you think! He can tell the song is derivative before anyone holding a contract will say it out loud. Tetlock spent a career proving the same thing with data, amateurs on base rates beating captured experts, which is a professor’s way of describing every fourteen-year-old who called a sellout before the band did.

So play the deal as a record. The band is the negotiators and the officials. The label is a calendar, four months to a midterm, and the label does not want an album. It wants one hit, a single called Peace in heavy rotation through November. A single only has to chart. Nobody at the label cares whether the strait is open. They care whether the word peace printed.

And the band knows. The Vice President is on television counting the economic pressure Washington still holds. The President threatened American tolls in the same sentence he promised there would be none. They read the chart fine. Then they cut the Unplugged version for the cameras, all candles and acoustic guitar, while the electric set tears through the actual room, the mined strait and the spoiler nobody could bind into the deal. That spoiler is Israel, the one party with the power to wreck the thing, never in the room and never a signature on it. It means to keep hitting Hezbollah whether or not that sinks the talks.

The genuinely fooled were the cheap seats, the market and the press that bought the single and printed the word peace. They had no contract to protect and no excuse.

Everybody watches the frontman. The frontman is the handshake on the resort lawn and the headline that the strait reopened toll-free. The bass line is what holds the song up, and here it is four notes it has been playing the whole time, an oversupplied oil market under a strait full of mines. The re-closure is the chorus coming back. Closure, ceasefire, closure. The Pixies built that structure and Nirvana sold ten million copies of it, and Iran has covered it twice since April without changing a note. The room cannot hear the bass line because the room is invested in the single going platinum, and investment is what kills your ear.

Before this turns into a victory lap, which is the move that drops a column into a guy screaming at a festival, I got the biggest thing wrong. I blew the oil price. I said so in Volume Six and I will say it again. I was in the repricing crowd waiting for a permanent key change, a darker tuning that held, and the price spiked to a hundred and sixteen in March and slid right back to the eighties with the strait still shut. I read the scene and walked out with the major-label remaster anyway.

So there is no genius in this story, and that is the actual problem. If a guy who cannot even call the price of crude can hear the structure from a basement, the failure upstream was never about brains. The people in the room are not stupid. They are paid to look away..

Do I give up. A little. Not on the work. I cannot stop doing the work. On the smaller hope that getting it right and getting heard sit in the same building. They do not. The people who get heard are paid not to be right, and the deal they signed last week is already coming apart on the clause I flagged on Wednesday.

I am in a basement up the coast, already hearing the next closure tune up.

Reading list

Iran reportedly closes Strait of Hormuz again, casting shadow over nuclear talks, CNBC, 20 June 2026. The contested re-closure in one place: Iran declares the strait shut while Washington insists the traffic is fine and the Vice President brags about a record sixteen million barrels. Read it for the distance between what gets announced and what is true.

Iran says Strait of Hormuz is closed, citing Israeli attacks in Lebanon, NBC and MS NOW, 20 June 2026. Central Command’s denial, plus the detail that matters most, that US intelligence expects Netanyahu to keep hitting Hezbollah even if it sinks the talks. The spoiler, confirmed.

East of Suez, Vol VII: Sixty days, no charge, Unpacking Complexity, 17 June 2026. The volume that flagged the Lebanon clause as the wish nobody signed, two days before it broke.

East of Suez, Vol VI: The reopening, Unpacking Complexity, 17 June 2026. Where I called the oil price wrong, on the record, so you do not have to take my word for it now.

A read-aloud of the fourteen points is finally public. The toll vanishes for sixty days then comes back, and the war in Lebanon was declared over by the two governments that do not have troops there.

Last week I told you this volume would read the memorandum against two questions, and only two. Does it make Israel leave Lebanon? Does it kill the toll? I also promised that if the text turned out to be nothing, I would say so and spare you a column. It is out, after a fashion, and it answers both questions.

First, the caveat, because it changes how hard you can lean on any of this. There is no printed, signed text. What circulated on Wednesday was a senior administration official reading the fourteen points down a phone line to reporters, alongside an earlier draft that leaked and that the White House has since disowned.

The signing happens Friday at a resort above Lake Lucerne, and the words on that page may not match the words read out this week. So I am working from a transcript, the parts of it that hold steady across CNN, CBS, TIME and the Lebanese and Israeli wires, and I will tell you where the ground is soft.

With that said. Read paragraph five.

The promotional rate

Paragraph five is the toll clause, and it is a small masterpiece of the genre. Iran will use its “best efforts” for the safe passage of commercial vessels, “with no charge for 60 days only”, from the Gulf to the Sea of Oman and back. Traffic starts at once. Then, allowing for demining and the removal of technical and military obstacles by Iran, it is to be running properly within thirty days. Ergo, Iran will sit down with Oman, and with the other Gulf states, to define “the future administration and maritime services” of the Strait of Hormuz.

Walk that through slowly.

The toll-free reopening that Trump authorised on Truth Social, that Macron demanded in his video, that von der Leyen called an essential condition, is real, unfortunately it also has an expiry date Sixty days, no charge. After that, the price of passage becomes whatever falls out of a negotiation that Iran convenes, about an administration that Iran helped invent, for a waterway whose far shore Iran owns. The Larak Island toll I described in Volume One has not been abolished by this deal. It has been put on an introductory offer, with the full-price menu deferred to a committee Tehran chairs.

Everything inside that paragraph leans Iran’s way. The demining is to be done by Iran, on its best efforts, the softest verb in any contract. The thirty-day clock is Iran’s to run. And the mines themselves are the mines that, as of this week, the US military still has not confirmed Iran ever laid. So Iran keeps the clock and the meter, and signs a document that thanks it for the favour.

The reversion is now in the text

Volume Six argued that the oil price had already mean-reverted while the strait sat closed, and that the spike was a spike rather than a new regime. Paragraph ten puts that argument into the treaty. The US Treasury issues waivers the moment the thing is signed, authorising Iran to export crude, products and derivatives, and the waivers reach the banking, the insurance and the transport that move them. More Iranian barrels, immediately, into a market that was long before the war started.

You do not need a model to see where the price goes when the second-largest holder of spare production in OPEC is handed an export licence into an oversupplied market.

The molecule the treaty forgot

Now do something the cheering is not doing. Search all fourteen points for the word Qatar. For LNG. For Ras Laffan. They are not there. The reconstruction money in paragraph six, at least three hundred billion dollars of it, is general, aimed at Iran, and routed largely through regional partners and private capital rather than the US taxpayer. None of it un-bombs a liquefaction train, and you cannot down-blend a wrecked one.

So the gas disruption that Volume Six called the durable half of this crisis survives the memorandum completely. The strait can reopen on Iran’s best efforts and the molecule with no bypass still does not move, because its problem was never the water. Two molecules, two clocks, and this treaty wound only one of them.

The guarantee nobody signed

Paragraph one is where the document stops being clever and starts being dangerous. The United States and Iran declare an immediate and permanent end to military operations “on all fronts, including in Lebanon,” and pledge to ensure “the territorial integrity and sovereignty of Lebanon.” The final deal, it says, will confirm all of that.

Hold that text against the wires from the same morning. Israeli troops advancing toward Haddatha in the south. Fresh strikes through the day. The defence minister stating that the army stays in Lebanon “indefinitely.” Netanyahu declining to say a word about the deal at all until he gets a meeting with Trump. Reporting that Washington never showed Israel the text, against Trump’s own line that it did and that Israel “could do a better job” in Lebanon.

Put it plainly. The one army that is physically in Lebanon did not sign this, is not bound by a syllable of it, and spent the morning of its release moving deeper in.

So the United States has put its name to a guarantee of Lebanese sovereignty that it has no mechanism to honour.

The Lebanese wires also carried word that Israel and Lebanon are edging toward their own separate ceasefire, on a separate track, with the G7 pushing Hezbollah disarmament. If that holds, paragraph one gets rescued by a deal it had nothing to do with. Either way, the fate of the clause sits outside the document that contains it.

What each side actually walked away with

Strip the victory laps that everyone seems to be taking:

• Iran got its oil-export waivers on day one and a path to several hundred billion in reconstruction.

• Its nuclear programme is frozen rather than dismantled, since down-blended uranium is reversible chemistry and standing centrifuges are not destroyed centrifuges.

• Its missile programme and its proxies were kept off the negotiating agenda entirely. And it traded a sixty-day toll holiday for the right to build a permanent toll regime under a respectable name.

For a country whose Supreme Leader was killed in February, that is a remarkably full basket.

Washington got oil prices falling into an election year, a signing ceremony, a blockade it can clamp back on, and a document it can call peace. It also got Trump, on the very day the text was read out, warning that he will resume the bombing if he dislikes how the next sixty days go, which is not the language of a man who thinks the war is finished.

Israel got none of the things it started the war to get. Everything it went to war over sits roughly where it sat in February, frozen in place by a status-quo clause Iran negotiated for itself. And a US-Iran text now invokes Lebanese sovereignty while Israeli armour rolls through south Lebanon.

The sixty-day option

Everything in here runs on the same timer. The toll holiday, the standstill on nuclear status, the negotiating window, the oil waivers that precede any formal lifting of sanctions. Sixty days from Friday lands in the back half of August. What fills the space between an August expiry and a November US election is the variable few are pricing.

Think of it as an option, because that is what it is.

For sixty days Iran exports freely, banks the goodwill of a toll-free strait, negotiates the administration that will price the strait afterwards, and keeps its programme exactly where it is under the status-quo clause. If the final talks stall at day sixty-one, the holiday ends, the standstill lapses, and the strait’s “future administration” becomes whatever Tehran says it is. Friday signs a sixty-day option, Iran is long it, and the United States wrote it.

So, the two questions, answered:

Does the text make Israel leave Lebanon. No. It asks for it in two signatures that cannot deliver it.

Does it kill the toll. No. It suspends it for sixty days and hands Iran the pen for what comes next. The wire says the war is over. The text says it is paused, with a return ticket in paragraph five and a promise in paragraph one that the people who made it do not control.

Price the sixty days, not the memo.

Volume Eight when the window runs down: whether the toll comes back, whether a treaty and an occupying army can both hold Lebanon at once, and whether a single laden tanker ever clears Larak without a warship beside it.

Reading list

US releases official agreement with Iran. Read the 14-point text, CNN, 17 June 2026. The official read-out, plus the useful detail that a leaked draft differed from it and that the White House said the draft did not reflect the real memorandum. Read this first, and read it as a transcript rather than a signed page.

Read the 14 points of the agreement between Iran and the U.S., CBS News, 17 June 2026. The cleanest point-by-point of the version dictated to reporters, including the immediate oil exports and the thirty-day reopening.

Read the Full Text of the 14-Point Agreement Between the U.S. and Iran, TIME, 17 June 2026. Good on the gap between the outdated leaked draft and the official read-out, and on the down-blending language that appeared only in the latter.

Full text of US-Iran memorandum of understanding, Naharnet, 17 June 2026. Worth it for the honesty of its own disclaimer, that because the deal was read aloud the transcript may be inexact, and for the Lebanese wires running alongside it.

Full text of US-Iran MoU read aloud by senior administration official, i24NEWS, 17 June 2026. The fullest version of the later paragraphs, including the sequencing clause that ties everything to continued implementation, and the oil and asset provisions.

Live updates: Trump-Iran deal, NBC News, June 2026. For the Israeli end: the defence minister’s “indefinitely,” Netanyahu seeking a meeting before he comments, and the Israeli strike that nearly derailed the announcement.

The Middle East and Global Energy Markets, IEA, June 2026. The gas numbers, again, because no clause in this memorandum touches them. Ras Laffan offline since 2 March, the LNG wave set back two years, and no bypass to fix it.

East of Suez, Vol VI: The reopening, Unpacking Complexity, 17 June 2026. The piece this one answers. The false-dawn call and the two questions, now adjudicated against the text that was supposed to settle them.

A deal signs in Geneva on Friday and oil just touched a three-month low. The gas market is not buying it.

“Ships of the World, start your engines.” That was the President on Sunday night, announcing that the United States and Iran had reached terms to end a war that started on the last day of February and to reopen the Strait of Hormuz. Fourteen points. A sixty-day ceasefire on every front, Lebanon included. The American naval blockade lifted. The strait, in the language of the memorandum, opened immediately.

The oil market read the same headline and sold.

Brent fell about four percent on the news and then kept falling, a fourth straight session lower, down into the low eighties and the weakest level since the first week of March. WTI broke eighty. Anyone short the reopening, and I gave that view plenty of room in these pages, spent the week watching the screen go red.

So let me do the thing the cascade pieces kept promising and split what actually changed from what the wire says changed.

The deal is smaller than it sounds

Read the fine print. The memorandum opens the strait on paper. The water is still mined. Britain and France spent part of the G7 talking about who clears them, which is not a conversation you have about a waterway that is open. A senior administration official promised a “significant increase” in traffic within two weeks while conceding in the same breath that mines take time and that different flags carry different appetites for sailing over them. The UAE’s own oil company has said full flows do not return until 2027 even on a fast deal.

A signature in Geneva is a fact about a piece of paper in a room in Switzerland. The mines are still in the water, and the only reopening that settles an invoice is the first laden VLCC that clears Larak Island without a warship alongside.

None of which makes the deal fake. Call it a starting gun, fired over a course still being swept for ordnance.

Why eighty dollars, with the strait shut

Roughly a fifth of the world’s seaborne oil moves through Hormuz, call it twenty million barrels a day. The IEA called the closure the largest supply disruption in the history of the oil market and meant it. Brent touched 116 on the ninth of March. Physical North Sea cargoes printed near 130, and at the peak some barrels changed hands around 150. Gulf producers shut in something close to ten million barrels a day because the tanks were full and the ships would not come.

And then, with the strait still closed, the price came back down. It is lower today than it was in March. That is the fact that should bother anyone who wrote, as I did, that the repricing was permanent.

The repricing of the front month was never going to hold, because the floor under the spike sat closer than the headlines implied. The market was oversupplied before the first missile flew, with OPEC+ unwinding cuts from April and the EIA still penciling Brent into the mid-sixties by 2027 on inventory builds of two to three million barrels a day. Iran never stopped exporting; it stopped everyone else, while its own escorted cargoes kept moving east to Chinese refiners restocking at a discount. And Washington reached for the tool the headlines never headline, a Treasury licence to lift roughly a hundred million barrels of stranded Russian crude onto the water, easing one sanction to plug the hole left by another.

That last move is the whole game in miniature. Nobody reopens a chokepoint by force on a deadline. You route around it. You run down the inventory you built for exactly this, then ease a rival’s sanction so his crude keeps your own voters in petrol.

The gas did not get the memo

The oil disruption was always going to mean-revert. The gas disruption was not.

You can pipe crude around Hormuz. Saudi Arabia has Petroline to Yanbu ant.d the UAE has the Habshan to Fujairah line, together perhaps three and a half to five and a half million barrels a day of genuine spare once you account for the fact that the binding constraint was never the pipe but the loading berths at the far end, which nobody sized for a siege. Imperfect, attacked, GPS-jammed, but real.

There is no such relief valve for liquefied natural gas. Not one bypass terminal exists on the Gulf coast. Ras Laffan, the largest LNG plant on the planet, has been offline since it was first hit on the second of March, and QatarEnergy is sitting on a force majeure. The IEA counts more than two billion cubic metres of lost supply every week and reckons the damage has pushed the entire global LNG growth wave back by at least two years, a cumulative hole of around 120 billion cubic metres out to 2030. Asian spot prices jumped forty percent at the open. Goldman put European gas at double its level on a one-month halt and above a hundred euros a megawatt hour past two months.

The strait, the transit problem, can clear in weeks. Bombed liquefaction trains are a multi-year repair. Hold both facts at once and the shape of the thing resolves: two molecules, two clocks. The deal that fixes the water does almost nothing for the gas. The cascade migrates

The oil winners were a peak, and peaks roll over. Russia’s Urals discount collapsed and its monthly take jumped while crude was bid; Frontline printed its best quarter in twenty years with VLCC day rates that touched the low four hundred thousands. Read those numbers again now that oil is at eighty and the ships are being told to start their engines. Tanker rent is a function of dislocation, and the dislocation is being negotiated away.

The gas winners are structural, and they are only getting started. United States Gulf Coast terminals, Sabine Pass and Corpus Christi and the rest, Australian supply out of the Pacific basin, and the unglamorous business of floating regasification and storage.

What Tehran actually bought

Iran’s deputy foreign minister calls the deal a victory. Its Supreme National Security Council insists the Americans must perform first. Discount the theatre and ask what holding the strait shut was still earning Tehran by June, because the answer is most of why the gun went off.

On oil, the glut meant the world could outwait Iran. Every week of closure that the inventory buffer absorbed was a week the economic weapon lost potency rather than gained it. On gas, the worst was already done and could not be undone on any timeline Iran controls; Ras Laffan does not un-burn because the IRGC keeps the lane closed. So the closure was strangling Iran’s own ports under the blockade while extracting a smaller and smaller premium from everyone else. Macron and the President both leaned hard on the same phrase, traffic “without tolls”. The Larak Island toll model I named in Volume One is precisely what this memorandum is drafted to kill. The toll-collector folded for the dullest reason in commerce: the meter stopped paying him enough to bother.

A note from the last time a strait closed

Permit me one jump sideways. When Egypt closed the Suez Canal in 1967, fourteen ships were trapped in the Great Bitter Lake and stayed there eight years, crews rotating, decks going yellow under the desert dust. The Yellow Fleet, they called it. While those hulls sat, the world did not wait for the canal. It built bigger tankers and sent them round the Cape, and by the time Suez reopened in 1975 the detour had become the system and the canal had lost a measure of the centrality it spent a century accruing.

Watch the same physics in the Gulf. If Qatari gas stays constrained into 2027 on the Ras Laffan timeline, the bypass pipelines, the American and Australian cargoes, the new FSRUs off Rotterdam and Tokyo Bay all harden from emergency into architecture. Iran’s chokehold accelerates the diversification that makes the choke worth less. A weapon held too long rusts in the hand.

What to watch, and what to ignore

The deal will hold or it will not, and the load-bearing risks are not in Geneva. They are an Israeli army that says it is not leaving the ground it took in Lebanon, a nuclear file the memorandum punts into a sixty-day window, sea mines whose count nobody will publish, and a successor son, Mojtaba, installed in the middle of a war and not yet tested by the peace.

So ignore the ribbon-cutting on Friday. The number that matters is the war-risk premium on the first VLCC that clears the strait unescorted, still running at something like eight times its pre-war level the last time anyone checked, and what the underwriter quotes for the ship behind it. Price the water, not the press release.

Reading list

2026 Iran war, Britannica, updated June 2026. The clean chronology from the 28 February strikes and the Khamenei assassination through the April ceasefire to the June memorandum. Start here if you have not followed the war week by week.

Israel/US-Iran conflict 2026: Reopening the Strait of Hormuz, House of Commons Library, June 2026. The soberest official accounting of how little “reopening” means in practice, including the UAE estimate that full flows wait until 2027 and the WTO read on a ninety-five to ninety-nine percent collapse in crude and LNG transits.

Trump, Iran agree to memorandum of understanding opening Strait of Hormuz, ABC News, June 2026. The fourteen points, the sixty-day ceasefire, the digital signatures, and the Friday ceremony. Note the official conceding that mines and risk tolerance gate the actual reopening.

US, Iran reach deal to extend ceasefire, open strait, Axios, June 2026. The tightest summary of what the deal does and what it leaves unresolved on the nuclear question.

Strait of Hormuz, IEA. The factsheet on bypass capacity, Petroline and the Habshan to Fujairah line, and why three and a half to five and a half million barrels a day is the honest spare number rather than the headline pipe rating.

Hormuz Bypass Infrastructure Was Sized for a Short Disruption. This Is Not That, Engineering News-Record, March 2026. The best piece on the real constraint: not the pipes but the loading berths at Yanbu and Fujairah, and Iran’s habit of striking the exit ramps rather than the strait itself.

The Middle East and Global Energy Markets, IEA, June 2026. The gas numbers that should have been the centre of this series from the start: Ras Laffan offline since 2 March, two billion cubic metres lost weekly, the global LNG wave delayed two years.

Hormuz Under Fire: LNG Disruption, Regional Exposure, and Energy Sovereignty in MENA, Arab Reform Initiative, March 2026. Why this is a logistics and flexibility shock as much as a price shock, and why no engineering fix exists on the LNG export side.

Short-Term Energy Outlook, global oil, US EIA. The forecast that explains the floor under the spike: inventories building through 2026 and 2027 and Brent drifting back toward the mid-sixties once flows normalise.

Strait of Hormuz live tracker, straits.live. For the readers who want the daily reading rather than the narrative: transits against the ninety-four-a-day baseline, the crisis pressure score, and the war-risk premium running near eight times normal.

The Repricing of Hormuz (East of Suez, Vol I), Unpacking Complexity, April 2026. The original framing piece, named the IRGC toll model and called the repricing permanent. This volume is where I check that claim against the tape and concede the half of it that did not hold.

East of Suez, Vol V: The cascade, Unpacking Complexity, May 2026. The wealth-transfer accounting. Read it next to this one and watch the beneficiary list split into an oil half that is closing and a gas half that is only opening.

Last time I wrote about the alibis people reach for when a trade goes wrong: blame the rain, the Chinese, the competition. Often what you blame is usually the most predictable part of the deal, and the forward curve was likely pointing you in the right direction but did not fit the narrative. Sometimes, however, a competitor builds the capacity, drives the price below cost, and waits for you to fail, because your failure is the objective. No forward curve warns you about a rival who will lose money for ten years to clear the field.

There is a particular kind of meeting I sat in more than once, early in my career, where a clever analyst would explain that a competitor was behaving irrationally. The competitor was selling below cost. It made no sense, the market would correct! The irrational player would run out of money and come to his senses, and we, the rational ones, would still be standing.

We were not always still standing.

And it took me years to understand why, which is that the analyst had made a category error so deep he could not see it. He assumed everyone at the table was playing the same game, the game with money as the score: some of the players were not, some were playing a different game entirely, one where money was not the score but the ammunition, and a loss on the income statement was simply the price of a seat that someone else would soon have to vacate. Sovereign actors are often more patient and can play for whatever exists after you are gone.

That second game, the one the West keeps mistaking for the first, runs through four markets, and the order matters because the order is the argument. By the end the same move has shown up four times, always made by the same kind of player, with the bill arriving last.

The move, in one sentence

Strip away the commodity and the move is always this. Build capacity faster than demand. Let the price collapse. Let the collapse bankrupt every producer who needs the price to survive. Then stand in the cleared field you now own. Zero is the munition, and the target is the monopoly on the far side of it, the rent you collect once you are the only one left who can supply the magnets, the cells, the refined lithium that the buyer’s economy now runs on.

The reason this is hard to spot is that every stage looks, in isolation, like ordinary market behaviour. Overcapacity looks like a planning mistake. A price collapse looks like a glut. Bankruptcies look like the cycle doing its work. The sequence only resolves when you notice the same player walking out of every one of them larger. A cycle returns you to where you started. This leaves one player holding the field, which is the older meaning of enclosure.

Minerals, already gone

China sells processed rare earths below the cost of digging them out of the ground. The German Marshall Fund notes plainly that Chinese sellers drop prices to reclaim share and drive out competitors, that mines shut from this predatory pricing, and that processed minerals are sold below extraction cost. The lithium story is the same arc with bigger numbers. Lithium prices surged eightfold in 2021 and 2022 and then fell more than eighty percent, while supply led by China, Indonesia and Congo poured into the market and crushed the price. Cobalt, nickel, graphite, the same shape.

And here is what sits on the far side of zero. China holds over sixty percent of global processing for cobalt, lithium and manganese, over seventy percent of graphite extraction, and for the magnet rare earths the share of refining is higher still. Roughly seventy-two percent of rare earth oxide output in a recent year. That position was bought. It was bought with two decades of margins nobody else was willing to forgo, and once it was complete, the price went up and the gloves came off. In October 2025 Beijing announced sweeping export controls on rare earths and then on the entire lithium-ion battery supply chain, covering cells, cathode precursors, anode materials, production equipment. The thing that was sold cheap to clear the field is now metered to whoever Beijing chooses.

Watch the Western response, because it shows the West still scoring the money game. The Pentagon signed a deal with MP Materials guaranteeing a price floor of one hundred and ten dollars a kilo for neodymium-praseodymium, paying the difference whenever the price falls below it. A government has concluded that the only way to keep a strategic mine breathing is to step outside the price mechanism and underwrite it directly, because an open auction sells the asset to whoever spent a decade pricing it to zero on purpose. The free market’s answer to strategic enclosure is to stop being a free market. That concession is the whole thesis, signed into a procurement contract.

Solar, mid-collapse

If minerals is the finished painting, solar is the artist still at the easel, and you can watch every brushstroke.

Polysilicon, the base material of a solar panel, fell from thirty-two dollars and seventy cents a kilo in February 2023 to four dollars and forty cents by May 2024. An eighty-six percent collapse. Prices sat below cost for more than a year. Chinese capacity reached more than twice global demand, utilisation fell to fifty or sixty percent, and the industry bled billions. A Western analyst files this as a disaster, a sector that destroyed itself through indiscipline. Anyone who has watched the move before files it as the middle act, going to plan.

And the consequences landed where they always land, on the producers who needed the price low. Norway’s Norwegian Crystals went bankrupt in late 2023. Switzerland’s Meyer Burger closed its German factory in 2024 and filed for US bankruptcy in 2025. Module prices fell below one renminbi per watt, which is to say below any level at which a European or American factory could draw breath, and the West’s solar manufacturing base, such as it was, suffocated. China now supplies around eighty percent of the world’s panels and ninety-three percent of its polysilicon.

Now the final act, the one the minerals story already finished and the one that shows zero was never the point. In December 2025, six of the largest Chinese producers, Tongwei, GCL, Daqo, Xinte and others, formed a joint venture to buy out and shut down excess capacity, backed by a fund reported around seven billion dollars. Polysilicon rose nearly seventy percent in a single month. The industry association started talking openly about survival of the fittest. The flood is followed by the cull, the cull by the consolidation, the consolidation by the re-rate. The price drops to zero to clear the room and climbs back once the room is empty, and the climb accrues to whoever is left holding the capacity. After a decade of below-cost selling that only a balance sheet indifferent to losses could sustain, that is China.

Oil, where one side isn’t a company

Oil is the case where the asymmetry is naked, because here the Western combatant is not a state and cannot pretend to be one.

The cost curve is the entire story and it is brutal in its simplicity. Saudi Arabia lifts a barrel for three to five dollars. Russia for ten to twenty. US shale needs around sixty-five to break even. Hold those three numbers next to each other and the 2014 price war stops being a mystery. OPEC refused to cut, drove crude from a hundred dollars to thirty, and aimed to bankrupt American shale. It worked, partially, hundreds of producers went under, and then shale adapted and came back, which taught Riyadh that one cull is not enough. So they are running it again, and this time they have said the quiet part aloud. As of early 2026 Saudi Arabia has abandoned its hundred-dollar target, pivoted to what its own ministers call volume over value, and Saudi state television has been running commentary on the lessons of the 1980s glut, when the Kingdom learned what happens to a swing producer who keeps cutting while everyone else cheats.

The framing in the trade press is the thesis stated without embarrassment. The winner of a price war, one brokerage wrote, is the producer who can survive longer at lower prices. That is the whole doctrine. Not the best producer, not the most efficient or the most innovative. The one who can bleed longest. And when the contest is a three-dollar sovereign barrel against a sixty-five-dollar private one, the question of who bleeds longest is not really a question. The US shale producer answers to shareholders who want returns this year. The Saudi barrel answers to a state that wants the oil market in 2040. Those are not the same time horizon, and the longer horizon wins every contest that is decided by endurance.

Then the bill comes

Now bring it home, because everything above is prologue to the thing the West is actually betting its future on/

The artificial intelligence build is the largest infrastructure commitment of our lifetimes. Alphabet, Amazon, Meta and Microsoft are expected to spend more than six hundred and fifty billion dollars in 2026 alone on AI capacity. The demand projections are vertical. Data centre electricity consumption is heading toward a thousand terawatt-hours, which would rank the sector, if it were a country, fifth in the world for power use, between Japan and Russia. A single AI task can draw up to a thousand times the electricity of an ordinary web search.

And it is colliding with a wall built from exactly the things China spent the last decade enclosing. As of April 2026, roughly half of all planned US data centre builds for the year are projected to slip or cancel, not for lack of money and not for lack of demand, but because the grid cannot power them. The bottleneck is physical, high-voltage transformers, switchgear, grid-tie batteries, components whose lead times have stretched from two years to five. Electrical equipment is under ten percent of a data centre’s cost and one hundred percent of its bottleneck. And to cope with the shortage, US operators have turned to China: imports of high-power transformers from China rose from fewer than fifteen hundred units in 2022 to more than eight thousand in 2025. The country racing to decouple from China for its most strategic technology is buying the picks and shovels for the gold rush from China, because China makes them, because making them at scale was another field quietly enclosed while Western boards were measuring quarterly margins.

The batteries that smooth AI loads, the critical minerals in the power electronics, the panels and the magnets, every input to the West’s defining growth story traces back to a supply chain someone bought at a loss on purpose. The International Energy Agency now reports, in its routine accounting, that trade restrictions are targeting the critical minerals for advanced power electronics and the batteries needed to smooth AI loads.

Of twelve gigawatts of 2026 US data centre capacity tracked across a hundred and forty projects, only five gigawatts is actually under construction. Eleven gigawatts sits announced with no physical progress, and a quarter of the projects have not disclosed how they intend to get power at all. This is the same disease as the LNG terminals and the dark fibre of 2001, capital committing to capacity on a forecast, except this time the capacity cannot be powered even if the demand shows up, because the inputs are controlled by a competitor who is under no obligation to sell them to you on a timeline that suits your shareholders.

The realist reading

Here is what ties the four together, and why the Western frame keeps failing.

In every case, the West read a market it was losing on price and reached for the tools you use when you lose on price. Tariffs. Subsidies. Anti-dumping duties. A price floor from the Pentagon. These are the moves of a player who believes the other player is also trying to make money and is simply doing it more cheaply, perhaps unfairly, and can be disciplined back into the money game with the right penalty. But the other player was never in the money game. Selling below cost for a decade is not a pricing strategy a profit-seeking firm can tolerate. It is a sovereign acquisition, denominated in dollars per tonne instead of dollars per share, and the asset being acquired is not market share, it is control of a chokepoint in someone else’s economy.

The pricing system has no field for any of this, and the blank is the core of the problem, not an edge case. A price measures the marginal trade. It carries no figure for strategic value, no line for the worth of being the last supplier standing when a rival’s defence industry needs magnets. The people who run the Western economy, the exchanges, the analysts, the boards rewarded for return on capital, are paid to optimise the number in front of them, and no one in that chain is paid to price the chokepoint. They track the round trip and miss the enclosure, the same way they track the spot price of oil and miss the structural premium. The number on the screen is real. It is pointed at the wrong war.

The deepest irony is that the West invented this move. The East India Company was a commercial entity wearing the powers of a state, a trading firm with an army, and it used market mechanics to achieve territorial control until the distinction between the two collapsed entirely. What we are watching now is the same fusion run in reverse. Then, a company acquired the functions of a state. Now, states have acquired the functions of a company, and they use the price mechanism the way the Company used its charter, as a tool of acquisition that looks, from a distance, like ordinary commerce. The enclosure of the English commons took fences and Acts of Parliament, and everyone it dispossessed could see the fence. This one takes a forward curve and a decade of patient losses, and it arrives in a number, dollars per tonne, that a Western finance desk is trained to file as good news. Cheap solar, cheap lithium, cheap oil, right up until the moment none of it is for sale.

I spent the first part of my career on the wrong side of that clever analyst’s table, learning slowly that the patient player losing money was not a fool to be waited out but a strategist to be feared. The lesson cost a lot to learn at the scale of a trading book. We are now learning it at the scale of a civilisation, and the tuition is the control of the inputs to the next century. The race to zero has a winner. It was never going to be the one keeping score in money.

A short, strange reading list

The books that taught me to see the move rather than the market, mostly from outside economics, which is the only vantage from which economics is legible.

Karl Polanyi, The Great Transformation (1944). The indispensable one. Polanyi’s argument is that the “self-regulating market” is not a natural state but a thing states construct and must constantly defend, and that land, labour and money are fictitious commodities that markets cannot actually govern without society pushing back. Read it and the Pentagon’s price floor stops looking like an aberration and starts looking like Polanyi’s prediction arriving on schedule. The market was always political. We just forgot.

Giovanni Arrighi, The Long Twentieth Century (1994). A history of capitalism as a succession of hegemonies, Genoa, the Dutch, the British, the Americans, each of which fused commercial and state power, dominated, and was displaced by a challenger who fused them differently. Arrighi gives you the long clock against which the China move is just the current turn of a very old wheel. Heavy going, worth it for the vertigo.

Daniel Yergin, The Prize (1990). The history of oil as the history of power, because they were never separable. If you only absorb one thing, absorb that every actor who treated oil as a commodity rather than a strategic asset eventually lost to one who knew the difference. The 1980s glut chapters are the direct ancestor of the Saudi 2026 pivot, and Riyadh, unlike the West, has clearly reread them.

Sun Tzu, The Art of War (c. 5th century BCE). Yes, the obvious one, but read specifically for the doctrine of winning without fighting, of shaping the field so the outcome is decided before the battle. Below-cost selling for a decade is the supreme example, the enemy is defeated by the terrain you spent years preparing, and the actual price war, when it comes, is a formality. Most quote this book. China appears to have run it.

Vaclav Smil, How the World Really Works (2022). The antidote to digital fantasy. Smil insists, with relentless physical accounting, that the modern world runs on materials and energy that cannot be wished into existence or coded around, steel, cement, ammonia, plastics, and the energy to make them. The data-centre wall is a Smil chapter written in real time, the discovery that you cannot ship a forecast through a transformer that takes five years to build.

John McPhee, The Control of Nature (1989). The strangest pick and the one I’d press on you hardest. Three essays on humans trying to hold back lava, the Mississippi, and the mountains above Los Angeles. It is ostensibly about engineering hubris, but what it is really about is the difference between a force you can negotiate with and a force that is simply patient and does not care about your timeline. Read it as a book about competing against a player who is not in a hurry. The river always wins, eventually, for the same reason the three-dollar barrel does.

Cathay Pacific lost sixteen billion dollars on a bet, and called it competition.

Between 2016 and 2019 Cathay’s fuel hedging program lost 16.24 billion dollars. Strip the hedges out and the airline would have cleared 18.4 billion in profit over the same four years. Ergo, the trading desk did not manage the risk of the business, it inverted it and turned a fortune into a loss. When management explained the years to shareholders, the word they reached for was competition: Stiff Asian competition, the other airlines were just so aggressive.

Across the same stretch, in the same fuel market, reading the same screens, Southwest Airlines saved three and a half billion dollars hedging. One quarter the hedging gains came to 291 million against a company profit of 34 million, which is to say the bet on fuel was, that quarter, roughly eight times the size of the airline underneath it. Same instrument. Same information. One read it and one gambled, and only one of them said so afterward.

From my perspective, one thing I learned is that the market almost always has good indications: the forward curve, the line that plots where a commodity is priced months and years out, is a bookmaker. It chalks up the odds in plain sight. The losses I am going to walk you through are not stories about people who could not get the odds. They are stories about people who got the odds, did not like them, bet the other way, and then, more frequently than not, blamed the thing that arrived on schedule.

The reason I am writing this now, rather than as a tidy history, is that the same bet is being placed today at a scale that makes Cathay look like a man feeding coins into a fruit machine. Hold that thought. First the pattern, because you need to see it three times before it becomes impossible to unsee.

The corn that anyone could see coming

Let’s start small and stupid.

VeraSun made ethanol. Ethanol is mostly corn. At the end of June 2008, with corn near eight dollars and the market gripped by a drought scare, VeraSun bought six months of it and locked the price. A few days later it started raining in Iowa.

There is the entire skill of the commodity business in one sentence: it rains in Iowa. The drought premium baked into that corn price was a bet that the rain would not come, and the rain comes, it is the Corn Belt, the place exists because the rain comes.

VeraSun did not buy corn. It bought the top of a weather panic and held it while the weather did the most ordinary thing weather does. The position helped destroy the company with a company forecast that ignored the market forecast.

The flag carrier that bet against the curve

Up the scale. Japan Airlines, a national flag carrier, went bankrupt, and among the triggers were wrong bets on the price of crude. Not a crash nobody saw. A position the airline chose, on the direction of oil, that went the other way.

What happened next tells you the industry knew exactly what it was looking at. After JAL went down, the other majors pulled back from hedging, reasoning that oil had run to two-year highs and would probably ease. They read the JAL body correctly, as a warning about betting on direction, and then placed the opposite directional bet, that prices would fall, and called that prudence.

A casino does not become a savings account because you moved your chips from red to black.

The five-billion-dollar mine that sold for ten million

Now the one with a town attached.

In 2011, iron ore around 190 dollars a tonne, Cleveland’s Cliffs Natural Resources paid roughly 4.9 billion dollars for the Bloom Lake mine in Quebec. Top of the market, top dollar, for a high-cost deposit sitting on the wrong side of the planet from its only serious customer, China. The freight disadvantage was geography. The price was a print anyone could read. And the force that would break that price was being announced out loud by the people about to apply it: Rio Tinto, BHP, Vale and Fortescue were spending billions expanding production, telling the market in plain language that a wall of new supply was coming, then shipping it even as Chinese demand cooled, because they wanted to earn out their own enormous investments.

The trap was visible from both sides at once. Demand, a question mark. Supply, a stated fact. Cliffs paid five billion into the gap.

Iron ore went from 190 to 40. Bloom Lake entered creditor protection in early 2015 and sold out of bankruptcy that December for 10.5 million dollars cash, plus 42 million in environmental liabilities the buyer had to swallow.

Call it a 99.8 percent loss on the purchase price in four years.

A parallel casualty, Sweden’s Northland Resources, had built its mine on a feasibility study assuming 110 cents and then blew through its own cost base by 425 million dollars before the price even turned; eight hundred-odd jobs went when it folded. A Labrador town watched its mine go cold and waited for a buyer who came late and small.

And the cause of death, the one in the chairman’s statement and the wire copy? The dramatic fall in iron ore prices. China slowed. As if the slowdown were a meteor rather than the most-watched economic story on the planet, as if the supply wave were weather.

The tell

Notice the shape, because here is where it stops being history.

The corn man blamed the rain. The airline blamed the competition, the oil price, the cycle. The miner blamed China. In every case the thing they blamed was real, it genuinely happened, and it was also the single most foreseeable element in the entire picture. Iowa rain. Chinese growth easing off a once-a-generation peak. Iron majors shipping the supply they had spent years and billions promising to ship.

They bet against the odds the bookmaker had chalked in plain sight, and when the favourite romped home they pointed at the favourite and said, who could possibly have known.

This is not a story about bad luck and it is not really a story about exotic finance. It is a story about a particular kind of lie, the one you tell yourself first and the shareholders second. The bet was the decision. The weather, the Chinese, the competition were the alibi, drafted in advance and read out on cue, because “the rain came” sounds like fate and “I bet against the rain in Iowa” sounds like what it was.

Why this matters more today, not less

Here is the objection, we have just lived through a year where the price of oil was set less by curves than by one man’s temper. In 2026, WTI ran from the mid-fifties in January to nearly 116 dollars in early March on the US-Iran conflict, then collapsed nine percent in a single session on a report that Washington was considering seizing the Strait of Hormuz, then climbed back toward a hundred by May. Brent touched 140 intraday at the peak, the highest since 2008. None of that was on any bookmaker’s board. No forward curve prices the afternoon a supreme leader dies or a president decides to send the fleet. So surely, you might say, the whole sermon collapses. If the future is a coin held by a politician, reading the curve is a mug’s game and the alibi is just the truth.

What the year proves is the distinction the whole argument turns on, and it is the distinction almost everyone gets wrong.

There are two different things you can be wrong about. You can be wrong about the timing of a discrete political shock, which is genuinely unknowable, no curve will ever hand you the date Iran moves, and being caught out by the date is not a sin, it is the human condition. And you can be wrong about the fragility that makes the shock matter, which is almost always visible for years. The chokepoint. The single supplier. The high-cost asset on the wrong side of the world. The unhedged fuel book. Hormuz did not become a 21-mile bottleneck for a fifth of the world’s seaborne oil in March 2026. It has been that bottleneck for fifty years. The timing was hidden. The fragility was on a map.

So the modern sin is not failing to predict the politics. It is leaving yourself wired to a fragility you have always been able to see, and then, when the known weakness is finally struck, reaching for the freshest and most flattering alibi on the shelf. And the alibi for this decade is already chosen. It is the word geopolitics.

By 2025 not one of the four largest US airlines ran an active fuel hedge, the case for going naked resting on government forecasts of moderate oil through 2027. Then Hormuz lit up. When the losses are booked, watch the language. It will not be “we stopped hedging a fuel-intensive business on the strength of a forecast”. It will be “unprecedented geopolitical volatility”. The volatility was not unprecedented. The exposure was a choice.

The largest bet ever placed on today’s price

Which brings me to the thing that actually prompted this, the bet so large it makes every grave above look like a warm-up.

Consider liquefied natural gas. An LNG export terminal takes four to five years to build between the moment the money is committed and the moment the first cargo sails. That lag is the whole point, so hold it in your mind. The decision to build is taken today, on today’s prices, and the plant earns its living half a decade later, in a market that today’s prices say nothing reliable about.

What did the industry do? After Russia’s invasion and a tight few years that pushed gas prices high and kept export terminals running at over 97 percent utilisation, the builders read the high price as a green light and floored it. 2025 was the largest year for LNG final investment decisions on record, more than a hundred billion cubic metres of new capacity sanctioned, over ninety percent of it in the United States. The wave has rolled straight into 2026. Add it all up and roughly 300 billion cubic metres a year of new export capacity is due online by 2030, a fifty percent increase in global supply.

Here is the part that should make the hair on your neck stand: the same agencies cheering the buildout are forecasting, in the same breath, that this very wave will create a supply glut by late 2026 and drive prices down for years.

The head of the IEA has said it plainly, the market is going from a seller’s market to a buyer’s market. Up to three-quarters of the new LNG has no fixed destination, so it will slosh into whatever market is weakest and push the price down further. The builders are building the glut. Each terminal sanctioned on today’s high price is a brick in the wall that will crush tomorrow’s price, and they are all laying bricks at once, on the same assumption, into the same wall.

When the cargoes arrive in 2029 and 2030 into the oversupply that was forecast the day the concrete was poured, the story will not be “we sanctioned a four-year project on a spot signal we were told would not last”. The story will be “an unexpected wave of global oversupply”, “demand from Asia disappointed”, “nobody could have foreseen the speed of the build”.

LNG is a more disciplined example. Look at the wilder one. The hyperscalers, the great cloud companies, are committing more than a trillion dollars across 2025 and 2026 to AI data centres, and the utilities behind them are planning some 1.4 trillion dollars of capital through the end of the decade to feed them power, the whole edifice resting on demand forecasts for a technology whose actual revenue is, to put it gently, still emerging.

We have watched this film.

In the late 1990s the telecom industry laid half a trillion dollars of fibre on a forecast of infinite internet traffic, the NASDAQ fell seventy-eight percent, the bankruptcies came in waves, and the fibre sat dark for a decade. Some of it is still dark. The capacity was real and eventually useful however the companies that built it on the forecast mostly died. When the AI capacity overshoots the demand, and the history of every capex supercycle says some of it will, the explanation will not be “we extrapolated a demand curve off two good years”. It will be a new and unimprovable word for the rain.

What to do with this before the next quarter closes

I am not telling you the future is knowable. I have spent this whole piece insisting the opposite, that the timing of the shock is a coin you cannot call. I am suggesting that the fragility is almost always visible, that the forward market is usually less wrong than the person who has already decided what he needs it to say, and that the gap between those two is where the money goes to die.

So a small and uncomfortable exercise. Take the position you are most confident about, the long-dated one, the one where the market is quoting you a future price you have privately decided is wrong. Now write down, today, the sentence you would say if it went against you. If that sentence already exists, if it already has a villain in it, the rain or the Chinese or the competition or the geopolitics or the unexpected glut, then you are not holding a conviction. You are holding a bet, and you have started drafting the alibi before the dice have stopped rolling.

Courage is to write the alibi down now, while it can still embarrass you into changing the position. The cowardly one is to wait, and read it out later, as news.

A short, strange reading list

Not the finance canon. These are the books that taught me to see the alibi forming, mostly from outside markets entirely, which is the only place you can see markets clearly.

Charles Mackay, Extraordinary Popular Delusions and the Madness of Crowds (1841). The oldest one and still the best on the central fact: people do not go mad one at a time and recover one at a time, they go mad in herds and recover slowly and singly. Read the South Sea chapter and then re-read the LNG numbers above. Same animal, better lighting.

Garrett Hardin, “The Tragedy of the Commons” (1968). Twelve pages. The LNG wave is a commons problem wearing a balance sheet: each builder’s terminal is individually rational and the sum is collectively ruinous, and no one of them can stop without simply handing the loss to the others. Hardin tells you why “they should have known” and “they could not have acted differently” are both true at once, which is the part most market writing misses.

Philip Tetlock, Superforecasting (2015). The empirical demolition of the confident expert. Tetlock’s forecasters win by separating the timing they cannot know from the base rates they can, which is the timing-versus-fragility distinction in this piece, proven with data rather than asserted with anecdotes. The chapter on why pundits are worse than dart-throwing chimps should be laminated and stuck to every boardroom door.

Ernest Hemingway, The Sun Also Rises (1926). Yes, the novel. For one exchange, on how a man went bankrupt: “Two ways. Gradually, then suddenly.” Every grave in this article is that line with a commodity attached. Read it for the rhythm of ruin, which finance textbooks cannot teach because they are written by people who have never felt it.

Marc Levinson, The Box (2006). A history of the shipping container, which sounds like the dullest book ever written and is in fact a thriller about how a single infrastructure bet, taken on a contested forecast, quietly rewired the entire world economy and bankrupted most of the people who bet on it first. The ones who were right about the future still mostly lost, because they were early. The best correction I know to the fantasy that seeing the future is the same as profiting from it.

Norman Dixon, On the Psychology of Military Incompetence (1976). Why capable, intelligent, well-resourced men make catastrophic decisions and then defend them to the grave. Written about generals, true of boards. The mechanism Dixon names, the way an organisation’s need to protect a prior commitment overrides the evidence in front of its eyes, is the engine under every alibi in this piece. If you only read one, read this one, and read it as a mirror.d.

What 90 days has actually cost.

The short why. Vol V is the bill the world is paying while the actors who could end the Hormuz crisis profit from extending it. Reroute costs run about $200 million a week. The cascade is much larger: fertilizer through 2027, a fifth straight year of German industrial contraction, a Japanese trade deficit growing tenfold, an Irish fiscal model leaning on three US companies, and a $580 million oil-futures trade fifteen minutes before Trump’s Iran reversal post. The longer it runs, the harder these numbers become to walk back.

Wall Street has named the trade

The cascade has a Wall Street acronym now. NACHO, “Not A Chance Hormuz Opens.” It started on trading desks in early May, inverted from last year’s TACO trade (”Trump Always Chickens Out”). Dennis Dick of TripleDTrading posted “The TACO trade is over. It’s now the NACHO trade. Not A Chance Hormuz Opens” on 4 May. Within ten days CNBC, Fortune, MarketWatch, and Krugman had all picked it up. Mark Hackett at Nationwide put the positioning split cleanly to MarketWatch: oil markets aren’t pricing a deal happening, the bond market isn’t pricing one at all, and equities are either pricing a deal happening or treating it as if it doesn’t matter. Vassili Gkionakis at Lombard Odier added that only the oil, shipping, insurance, and rates markets are fully embracing NACHO while broader risk assets stay sanguine and equity indices hit record highs. Nadia Martin Wiggen at Svelland Capital told Bloomberg in late March that the market simply was not believing that May-loading barrels would arrive. They mostly did not.

Krugman wrote on his Substack: “I never bought into the TACO meme. But NACHO looks right. Hormuz won’t open until the economic damage from its closure becomes much more severe”.

The war-risk insurance premium for transiting the strait peaked at 2.5% of hull value per voyage in March, against roughly 0.1% before the war. The premium has eased since but still sits at around eight times pre-war levels per eToro data. US gasoline is at $4.52 a gallon per AAA, against just under $3 before the war. NACHO is what the trader desks call the duration thesis. Vol V is the analytical case for what that duration is actually costing.

The disruption is multiple bills running at once

The direct measurable cost of the Hormuz disruption runs around $200 million per week in additional fuel, insurance, and bunker spend across the global container and tanker fleet, per Maersk, CMA CGM, MSC, and Hapag-Lloyd carrier disclosures. That is the smallest line item.

Hormuz transit is down 95% since 28 February. On 10 March the strait recorded just two outbound crossings and zero inbound movements, against a pre-crisis seven-day average of 3.29 crossings per day. The casualty list includes 17 merchant ships damaged, seven abandoned, two captured, twelve seafarers killed or missing. On 4 March QatarEnergy declared force majeure on all LNG shipments after Iranian attacks on its Ras Laffan facilities, removing 20% of global LNG supply from the market overnight. The Singapore VLSFO cargo benchmark surged 76% to $988.5 per metric tonne in mid-March per Argus Media as bunker fuel flows rerouted.

The Shanghai Containerized Freight Index has reached levels not seen since the COVID-era peak in late 2021. Asia-to-US West Coast container rates have moved from $1,800 to $2,200 per forty-foot equivalent to over $4,500. Spot container rates on major routes are up roughly 150% since the closure began. Maersk reported a $153 million Q4 2025 loss in its Ocean division, its first quarterly loss in years, and its 2026 guidance now spans a range from $1.5 billion loss to $1.0 billion profit. That guidance band is the financial expression of how little anyone knows about how long this lasts.

The IEA has called the disruption the largest oil supply event in the history of the market. Cumulative supply losses exceeded 360 million barrels in March and 440 million in April. The agency released 400 million barrels from emergency reserves to defend the spot price. The Vol III chokepoints chart noted that around $10 trillion of trade passes through six maritime lanes every year. One of those lanes has been at 5% of normal volume for nearly three months, and a second (Suez plus Red Sea) has been compromised since the Houthis resumed attacks on 28 February. The shipping system has never operated with two of its primary corridors simultaneously broken in the modern era. We are doing it now.

The cascade arithmetic

Direct GDP exposure to a sector tells you almost nothing about real fragility. Germany’s auto cluster is roughly 6% of GDP narrowly. Add the supplier network, the chemicals cluster, and the energy-intensive industrial base and the figure runs to 12 to 15% of total industrial output. Layer in the currency, banking, fiscal and housing channels and a sustained shock can pull the GDP impact to one and a half to three times the direct number, eighteen months later, in the labour market data. Moody’s analytics chief Mark Zandi put US recession odds approaching 50% on the cascade in early April; his read sits at the cautious end. FGE NexantECA’s chairman emeritus Fereidun Fesharaki has been calling oil at $150 to $200 a barrel in any scenario where the strait stays near-closed for an extended period.

The same math runs through every concentrated economy. Korea’s manufacturing is 24% of GDP versus 13% across the OECD, layered on 85% energy-import dependence. Japan’s main cushion has been BOJ balance sheet expansion, which has its own limits. Each of these economies carries the same compounding structure between direct sector exposure and the cascade channels.

The economies most exposed to sustained Hormuz disruption are the same economies with the thinnest buffers and the most concentrated industrial structures. The wealth transferred out of their consumers and industrial bases goes to commodity producers (Russia, Saudi Arabia, US shale), to vessel operators (Frontline, Euronav, DHT, Nordic American Tankers), to refiners with sanctioned-crude access (Reliance), and to defence contractors running sustained-operations procurement. Anyone who has watched a fuel hedge unwind across a sustained price regime, or sat in a port operator’s office while insurance reprices on the same docks the previous week’s rates were written against, knows what this pattern looks like on the operating side. The cascade is not random damage. It is a wealth transfer with a known beneficiary list, and that list does not overlap with the people footing the bill.

Germany: the case study that no longer needs qualifiers

Peter Leibinger, president of the Federation of German Industries (BDI), said on 20 April that German industrial production has fallen every year since 2022 and the BDI no longer expects a recovery in 2026, only stagnation. The federation warned of a fifth straight year of contraction if shipping disruptions persist. That is the head of German industry’s main representative body, on the record, in April, before the Vahidi-Naqvi backchannel surfaced and before the 25 May CENTCOM strikes confirmed that Washington and Tehran are still in escalation phase. Capacity utilisation in German industry sits at 77.5%, well below historical norms around 84%.

The chemicals sector is where the cascade has already broken through. Markus Steilemann, CEO of Covestro and president of the European Chemical Industry Council (Cefic), warned in the Financial Times on 14 May that the Iran war’s compounding effect on naphtha and ethylene feedstock prices has pushed Europe’s chemicals industry beyond what the 2022 Russian gas crisis nearly delivered. Cefic’s latest survey indicates BASF, INEOS, Covestro, Lanxess, and Evonik are now running European plants at capacity utilisation levels of 62 to 68%, well below the 80% threshold at which heavy chemicals production becomes structurally unprofitable. BASF’s Ludwigshafen complex, the largest integrated chemicals site in the world, has been cutting permanent capacity since 2023, and Wacker Chemie and Air Liquide have moved into similar postures. European industrial gas prices remain roughly three to four times the US level and twice the Chinese level. The German Chemical Industry Association VCI, which represents about 2,300 firms, provided no 2026 forecast at its March press conference, citing extreme uncertainty.

The Bundesbank’s January 2026 monthly report identified energy-intensive sectors and motor vehicles as the primary drivers of the industrial decline. The ifo Joint Economic Forecast for Spring 2026 explicitly stated that the energy price shock is overshadowing the fiscal stimulus, with manufacturing gross value added, goods exports, and corporate investment all noticeably lower than the previous autumn read. Germany imports 70% of its energy on a net basis. Industrial fuel costs are the single largest variable cost line in German manufacturing competitiveness, and they are now stuck high for as long as the Middle East remains contested.

The supplier network is where the cascade really compounds. Bosch, Continental, ZF, and the thousands of Mittelstand suppliers behind them all operate on thin margins and assume continuity of energy and shipping cost. Each Hormuz week makes their working capital harder to manage. Chinese substitution in high-quality industrial goods has been advancing steadily since 2022, and the Hormuz disruption is accelerating the cost-competitiveness gap. Germany’s 2024 GDP contracted 0.5% and 2025 grew only 0.2%. If 2026 brings stagnation rather than recovery and 2027 is required to absorb the fertilizer and food cascade into household budgets, the German recession story stops being a 2024-2025 episode and becomes a permanent step-down of the European industrial base.

The Euro is a separate problem. Sustained German weakness pulls the Euro lower against the dollar, which inflates the cost of every imported industrial input, which deepens the cascade. The ECB has limited room to cut while energy is pushing headline inflation back up. Berlin’s announced infrastructure and defence spending should help in 2027, but the Bundesbank itself revised 2025 growth down and 2027 up, which is the central bank acknowledging that the near term is worse than the previous forecast.

Japan: the hidden balance sheet

Japan’s trade deficit shrank to 1.7 trillion yen in fiscal 2025 and could balloon to about 15 trillion yen in fiscal 2026 if Hormuz stays closed for an extended period, per an economist quoted by Nippon.com on 22 April. A near tenfold blowout, roughly $100 billion. In April, Japan’s crude oil imports plunged 64% year-on-year, the steepest drop since 1980. Tokyo is substituting US crude where it can, which helps the trade balance with America and worsens it elsewhere.

The exposure sits in three reinforcing layers. The BOJ balance sheet has been the cushion for over a decade and is now visibly approaching its operating limits. The yen-carry trade, one of the load-bearing flows in global finance, depends on Japanese current-account strength that a 15 trillion yen deficit would knock out. Behind both sits demographic decline, which leaves the household consumption buffer thinner than at any point since the war.

A trade deficit of that size weakens the JPY against the dollar. That forces the BOJ to choose between defending the yen and supporting the bond market. The choice stresses the global carry trade exactly as the AI capex cycle is demanding plentiful funding. Markets are not pricing this combination because each individual leg looks manageable on its own. The legs are correlated through Hormuz duration.

India and the cascade winners

Reliance Industries was Russia’s largest single buyer of crude in 2025, peaking at 746,000 barrels per day in June. Russian crude reached 40-45% of Reliance’s crude mix at peak per Rystad Energy. After Washington sanctioned Rosneft and Lukoil in October 2025, Reliance was forced to cut. Then in March 2026 the US Treasury granted a 30-day waiver allowing Indian refiners to continue Russian purchases for cargoes loaded before 5 March, and approximately 15 million barrels of Russian crude were on tankers in the Arabian Sea and Bay of Bengal ready for immediate uptake.

Reliance’s gross refining margin was $8.5 per barrel in FY25 versus Indian Oil at $4.8. Goldman Sachs projects Reliance EBITDA growth accelerating to 16% in FY26 from 2% in FY25, driven by what the bank calls “upcycle refining margins driven by favourable crude feedstock dynamics” alongside 1.3 million barrels per day of permanent global refining capacity closures over 2025 and 2026. Diesel crack spreads have moved from $20 per barrel to $35-42 per barrel since the closure, and the Reliance Jamnagar refinery has a 40 to 50% diesel yield. The maths is straightforward: cheap discounted Russian crude in, premium-priced refined product out to a Europe that cannot source from sanctioned Russia directly, and India captures the spread.

Reliance is not the only winner. Saudi Arabia and the UAE are pumping more oil through bypass infrastructure (the East-West pipeline to Yanbu, ADCOP to Fujairah, and the new Duqm capacity) at premium economics. US shale producers with WTI in the low $90s are extending Permian programmes. Reinsurers are repricing Gulf marine exposure on permanent terms. Defence contractors are running sustained-operations procurement schedules. The Atlantic Council’s Energy Sanctions Dashboard records Russian Urals crude trading at premiums of up to $10 over ICE Brent, with the G7 oil price cap at $44.10 a barrel rendered functionally irrelevant. Carnegie put Russia’s additional revenue at $8.5 billion a month.

The cascade winners produce commodities, own transport assets, refine sanctioned crude, or supply the security state. Everyone else is paying: industrial economies, consumer households, fiscally fragile states. Wealth flows from the second group to the first, week by week, for as long as Hormuz stays contested. The rational response of any actor in the first group is to extend the duration. Lanxess CEO Matthias Zachert told Reuters recently that the disruption is creating “temporarily more favourable market conditions for the European chemical industry,” which is a European industrial CEO admitting in plain language that his sector is partly a cascade winner. The same pattern repeats inside Iran (Vol IV is the full case) and at planetary scale here.

Front-running the policy

The wealth transfer runs through US insiders as well. On 24 March, Nobel laureate Paul Krugman published a Substack piece titled “Treason in the Futures Markets” documenting that roughly $580 million worth of oil futures changed hands in a single minute between 6:49 and 6:50 New York time on Monday 23 March, approximately 15 minutes before President Trump posted on Truth Social that the United States was in “productive conversations” with Iran. About 6,200 Brent and WTI futures contracts were sold in that window. S&P 500 futures spiked moments later. After Trump’s 7:04 announcement paused his ultimatum, oil sold off sharply and equities jumped, which is the outcome someone holding those positions would have wanted on both legs. The Financial Times documented the trade size.

Krugman called it treason because exploiting confidential national security information for personal profit is a category beyond standard insider trading: it broadcasts US government plans to foreign adversaries through the futures markets. Rory Johnston, oil market analyst, told Fortune that the pattern has been hard to ignore even without a smoking gun. This was not the first time. Large suspicious moves in the prediction market Polymarket preceded earlier US attacks on Iran and Venezuela. Iran’s parliament speaker Mohammad-Bagher Ghalibaf publicly denied that any negotiations were taking place that Monday, calling the claim “fakenews” used to “manipulate the financial and oil markets.”

The deeper implication is the part worth sitting with. If decisions about war and peace can be front-run by traders close to the White House, the incentive structure runs in both directions. The crisis benefits the actors holding the positions, and the actors holding the positions can influence the timing of the policy that moves the prices. Krugman raised the question plainly: are decisions about war and peace in part serving the corrupt insider trade calls? On 2 April he returned to the same vein, noting that the $4-a-gallon US gasoline price most commentators were treating as the headline of the crisis is “less than half” of the actual Hormuz hit once reroute costs, insurance premiums, fertilizer feedthrough, and the petrochemicals cascade are added in. The cascade has two payment streams. One routes through energy and shipping costs from industrial consumers and concentrated economies to Russia, the IRGC, and tanker owners. The other, smaller but politically more dangerous, routes from retail savers and unhedged producers to a small set of US insiders through front-run policy reversals on the futures clock.

The petrodollar test

The Hormuz disruption is also stress-testing the dollar’s role in oil settlement, which is the layer most cascade analysis misses entirely. Since March, Iran has been charging Strait of Hormuz transit tolls in Chinese yuan and stablecoins, not dollars. The Atlantic Council’s GeoEconomics Center documented the toll system on 30 March, with rates starting around $1 per barrel and payment routed through China’s CIPS clearing network or via mBridge, the digital-yuan settlement platform several Gulf central banks are now piloting. China is now Iran’s main oil customer, taking over 80% of seaborne Iranian crude. Indian refiners are settling Russian crude purchases in yuan and UAE dirhams, the first time a top-five oil importer has systematically bypassed the dollar at scale.

The petrodollar architecture, established by US-Saudi agreement in 1974, depends on global oil trade being denominated in dollars. Around 20% of global oil trade was already running in non-dollar currencies before the Hormuz crisis. The yuan toll at the chokepoint is what 20% looks like when it has access to a maritime extraction point. Each yuan-denominated transit through the strait is a small piece of evidence that the dollar is no longer the only settlement option for crude.

The dollar still clears 58% of global trade and 88% of foreign exchange volume. The architecture holds. But the erosion has stopped being hypothetical, and the Hormuz crisis is the closest the petrodollar has come to a structural stress test in five decades.

This is the cascade Vol V is least confident estimating, and the cascade with the largest tail risk if it accelerates.

The Irish cliff

The Irish Fiscal Advisory Council confirmed in February that 46% of Irish corporate tax in 2024 came from just three multinationals (understood to be Apple, Microsoft, and Eli Lilly), worth around €13 billion. The top two paid almost 40% on their own. 75% of all Irish corporation tax is paid by US multinationals. Ireland’s total corporate tax in 2025 was €33 billion, more than 31% of total tax receipts. Pharma represents 42.4% of Irish exports. The IFAC labeled the Irish government’s plan to cut the corporate tax savings buffer from 32% to 15% in 2026 as “risky”.

Brad Setser at the Council on Foreign Relations published “The Luck of the Irish” on 18 May, arguing the structural advantage Ireland has enjoyed since Apple became an Irish tax resident in 2014 is now exposed to a US fiscal regime that needs revenue. The Hormuz crisis raises the probability of that regime shift. Sustained elevated US fuel and food prices feed inflation, the Federal Reserve loses room to cut, fiscal pressure on Washington intensifies, and pharmaceutical profit-shifting becomes the politically obvious revenue target. Eli Lilly’s weight-loss and diabetes drug pricing is already under US political pressure for other reasons.

If even one of the three top Irish payers shifts profit booking back to the US, Ireland’s corporate tax take could fall by €5 to €10 billion in a single fiscal year. The current Irish government has not built the buffer to absorb that. The 32% to 15% savings ratio reduction means day-to-day spending now rests on revenue flows that the IFAC itself calls “largely unrelated to economic activity in Ireland.” A sudden contraction in those flows would force austerity into a workforce already at near-full employment in pharma and tech services. The downstream effects on Dublin property, Irish banking, and the Euro periphery would compound from there.

The Irish cliff is unlikely in any 12-month window, but it sits as the worst correlated tail risk in the developed-world fiscal data set, and the Hormuz crisis raises its probability by a measurable amount.

What governments are saying, and not doing

The state-actor record clarifies the duration thesis from another angle. On 21 March, twenty-two countries issued a joint statement condemning Iran’s attacks on commercial vessels and energy facilities and calling for immediate de-escalation. The signatories included the United Arab Emirates, Bahrain, the United Kingdom, Germany, France, Japan, South Korea, and Australia, covering roughly two-thirds of OECD GDP and most of the energy importers most exposed to the disruption. Three months later, none of those signatories has imposed material costs on Iran beyond what was already in place. The collective external response has been a press release.

Iran’s negotiating posture is the inverse. Iran’s lead negotiator told the press in early May that Tehran was “just getting started” with the Hormuz pressure, which is the public version of what Vol IV identifies as the IRGC’s rational pricing of the extraction. Foreign Minister Abbas Araghchi paired that with the warning that “return to war will feature many more surprises.” Parliament speaker Mohammad-Bagher Ghalibaf has periodically denied that the Doha negotiations exist at all. The state-actor signals from Tehran point one direction: extend.

Washington’s official position has been less coherent. Treasury Secretary Scott Bessent told reporters in mid-March that the media was “trying to make it into some crisis that it’s not”. Defence Secretary Pete Hegseth has since rolled out “Project Freedom”. a programme to escort stranded cargo ships out of Hormuz under US naval cover, which is the US government acknowledging operationally what NACHO has been pricing on the futures market. The Pentagon estimated at the start of the crisis that the Iran war would last up to six weeks. Day eighty-seven was yesterday. Robin Brooks at the Brookings Institution argued on his Substack on 13 March that the US could “implode Iran’s economy by shutting down its oil exports” as a forcing function on the strait, calling for Washington to give “the Ayatollahs a taste of their own medicine”. That option has been visible in the Trump administration’s threat menu without ever being executed.

The pattern across Washington, Brussels, Tokyo, and Riyadh is consistent: heavy rhetorical condemnation with occasional kinetic skirmishes, and no meaningful action that maps onto Tehran’s incentive structure. The Vol IV diagnosis predicts exactly this. The actors with operational ability to end the crisis sit inside Iran. The actors with the rhetorical authority to demand it from the outside have no instrument that reaches the IRGC. NACHO is confirmed from the state-actor angle: the institutions that could force a resolution are not doing so.

The inside-Iran half of the picture

Inside Iran, the people who can end the crisis (Brig. Gen. Ahmad Vahidi at IRGC, the security-council triumvirate around him, Mojtaba Khamenei’s narrow inner circle since 8 March) are the same people who benefit personally from sustaining it. The IRGC collects approximately $100 million per week from the Larak Island toll route directly into hard-currency accounts. The Foreign Ministry track in Doha is negotiating an MOU and a $24 billion asset release. Those flows would go to the central bank, bypassing the actors who actually operate the mines and the boats. Pakistan’s Interior Minister has flown to Tehran twice in the past two weeks to see Vahidi directly. The Doha process is the public ceremony; the real negotiation is in Tehran. Vol IV is the full case.

Vol V is the same pattern at planetary scale. The IRGC’s $100 million a week from Larak, Russia’s $8.5 billion a month from Urals premium per Carnegie, the record VLCC rates collected by Frontline and its peers, the diesel crack spreads captured by Reliance, the sustained-operations procurement booked by the defence contractors. None of those flows has any reason to want the crisis to end.

The cost is paid by German Mittelstand suppliers, Japanese household budgets, Korean exporters, US farmers facing 2027 fertilizer bills, the European industrial base, Egyptian Suez Canal revenue, and the Irish fiscal model. None of these payers has any operational tool to end the crisis. They have political voice, but the political voice runs into G7 governments whose deal-making counterparty is the Iranian Foreign Ministry, which itself does not control the actors profiting on the IRGC side.

The honest diagnosis is uncomfortable. The crisis ends when one of three things happens: the IRGC accepts an offer larger than the sustained extraction (unlikely while extraction continues), Washington applies kinetic pressure that exceeds the IRGC’s risk tolerance (the Camp David cabinet meeting tomorrow, Wednesday 27 May, will signal whether this is on the table), or the broader Iranian regime undergoes internal change that displaces the IRGC’s operational control (no current indicator of this). Without one of those three, the signed-but-unenforced equilibrium holds, and the cascade compounds for the duration.

What to watch over the next 90 days

German Q3 GDP and the Bundesbank’s autumn forecast revision. If Berlin acknowledges a 2026 contraction, the cascade is no longer a forecast.

Japan’s monthly trade balance and yen volatility. If the 15 trillion yen deficit forecast starts to materialise, the BOJ’s options narrow visibly and the yen carry becomes a global market story.

Reliance Q1 FY27 results and Indian refining margin trajectory. If GRM stays above $10 per barrel, the cascade-winner thesis is operational.

Irish corporate tax monthly receipts. The Department of Finance publishes these monthly. Any single-month drop greater than 15% versus the prior year is the early signal of profit-shift reversal.

US fertilizer purchase contracts for 2027. The pre-purchase advantage that buffered the 2026 farm economy will not exist for 2027. Mid-year contracts will reveal whether prices have stabilised or compounded.

EU industrial production indices from Eurostat. Germany leads but Italy, Poland, Czech Republic, and Slovakia (all German supply chain) will follow.

Four takeaways

The Hormuz disruption is multiple bills running concurrently, and the published estimates almost always count only the smallest one. Direct shipping cost is around $200 million per week, the cascade across industrial economies and food prices runs into tens of billions per quarter, and the fiscal cliff risks in Ireland and other concentrated economies represent uncapped tail risk on top. Anyone running their planning on a one-quarter oil-spike framing is mismeasuring the problem.

The cascade has known beneficiaries and known payers, and they do not overlap. Russia, the IRGC, tanker owners, sanctioned-crude refiners, and defence contractors are extracting wealth from German Mittelstand suppliers, Japanese households, Korean exporters, US farmers, European industry, and Irish fiscal authorities. The political consequences of this transfer will land before the financial consequences fully clear.

The wealth transfer reaches the US itself, alongside the foreign-to-foreign transfers. The $580 million Brent and WTI futures trade documented by the Financial Times on 23 March, fifteen minutes before Trump’s policy reversal post, is the visible part of a pattern that Krugman identified plainly: insider trading on national security decisions is a category of corruption that broadcasts US government intent to foreign adversaries while transferring wealth from retail savers and unhedged producers to a small set of people close to the policy clock.

The longer this runs, the harder the numbers become to walk back. Fertilizer prices will feed 2027 food bills regardless of when the strait reopens. The German manufacturing share ceded to Chinese substitution between 2024 and 2026 will not return on a ceasefire announcement, and Irish multinational tax routing, once politically targeted by a fiscally pressured Washington, does not quietly unwind back to where it sat in 2024. The G7 policy framework was built for crises that resolve in months. This one is being extended deliberately by the actors who profit from it.

Closer

Inside Iran, the question is who actually controls the negotiation (Vol IV addresses that). Outside Iran, the question is who pays the bill, and the answer is that the world’s industrial consumers and fiscally fragile states are paying it to extract a slow concession from an Iranian security apparatus that has no reason to give the concession. Vol VI will look at the geopolitical realignment that compounds on top of the economic cascade. The signal to watch in the next 30 days is whether Berlin or Tokyo or Seoul says any of this publicly, because the financial cost is now politically loud enough that one of them eventually has to.

NetBlocks confirmed on 26 May that partial internet access has been restored in Iran after 88 days of near-total blackout, which Trump immediately read as Tehran signalling it wants a deal. The same day, US forces struck a ground control facility in Bandar Abbas and intercepted four Iranian one-way attack drones near the strait. Hapag-Lloyd’s senior communications director Nils Haupt put the duration question bluntly in early April: “When the war is officially over, that does not mean the war is over for logistics. We will see hundreds of ships wanting to call at key ports in the Persian Gulf. The disruption will continue.” Camp David Wednesday is the next decision point on the kinetic side. The cascade arithmetic compounds regardless of what gets announced.

Fifth in the East of Suez series in Unpacking Complexity. Vol I established the five mechanisms of permanent repricing. Vol II identified Oman as the most likely tactical workaround. Vol III ranked four scenarios after the PGSA map and the GCC IMO letter. Vol IV named the IRGC veto, the Vahidi triumvirate, the Pakistani Naqvi-Vahidi backchannel, and the Iranian elite gold flight as the measurable confirmation signal. Vol V is the cascade: the structural cost of a sustained dual blockade running across the global economy.

Reading list, with the why

BDI President Peter Leibinger: German industrial production has fallen every year since 2022, no recovery expected in 2026 only stagnation, Global Banking and Finance citing the BDI annual statement at Hannover Messe, 20 April 2026. The most quotable line in the German cascade case. The head of industry’s main representative body putting the diagnosis on the record before the May escalation.

Joint Economic Forecast Spring 2026: Energy price shock overshadows fiscal stimulus, ifo Institute, 1 April 2026. The joint forecast of Germany’s leading economic research institutes explicitly attributing the downward manufacturing revision to the energy shock. Institutional confirmation of the cascade mechanism.

Bundesbank Monthly Report, Deutsche Bundesbank, March 2026. The German central bank’s own commentary identifying energy-intensive sectors and motor vehicles as the primary drivers of the January industrial decline.

Middle East Crisis Seen Affecting Japan Trade Balance, Nippon.com, 22 April 2026. The single most striking number in the Vol V data set: Japan trade deficit forecast to grow from 1.7 trillion yen to 15 trillion yen if Hormuz remains closed. Tenfold blowout, $100 billion order of magnitude.

Japan April exports rise 14.8% despite Middle East conflict, Reuters via Investing, May 2026. Includes the data point that Japan crude oil imports plunged 64% year-on-year in April, the steepest drop since 1980.

More concentration, more risk: three firms account for almost half of Ireland’s corporation tax revenues, Irish Fiscal Advisory Council, 19 February 2026. Brian Cronin’s analysis documenting that 46% of Irish corporate tax in 2024 came from three companies. The source for the Irish cliff numbers.

The Luck of the Irish, Brad Setser at the Council on Foreign Relations, 18 May 2026. The structural read on why Ireland’s MNC tax routing is exposed to a US fiscal regime that needs revenue. The most credible articulation of the Irish cliff thesis.

Fertilizer prices surge as Strait of Hormuz disruptions tighten supplies, World Bank, May 2026. The World Bank’s documentation that urea rose 80% from February to April to $850 per ton (highest since April 2022), with DAP up over 10% in April.

How fertilizer policies could exacerbate Hormuz price shocks, IFPRI with NDSU model, 23 May 2026. The model that puts trade flows at 50% of normal through August 2026 and 75% through early 2027 even on the baseline.

Fertilizer Cost Increases Resulting from the Iran Conflict, University of Illinois farmdoc daily, May 2026. The US farm economy perspective: 2026 was pre-purchased, the full impact lands in 2027. Food price cascade timing.

Reliance’s refining margins projected to stay strong compared to PSU peers, Business Standard, September 2025 with FY26 projections. Goldman Sachs projection of Reliance EBITDA growth accelerating to 16% in FY26 from 2% in FY25.

RIL Secures Russian Crude Amid Margin Surge and Waiver, Whalesbook citing US Treasury waiver, 6 March 2026. The 30-day Treasury waiver and the 15 million barrels of Russian crude on tankers ready for swift Indian acquisition.

Inside Tehran’s toll booth, Atlantic Council GeoEconomics Center by Alisha Chhangani, 30 March 2026 (updated 1 April with CIPS data). Documents the yuan-denominated Hormuz toll system, the CIPS clearing channel, and the mBridge digital-yuan settlement platform several Gulf central banks are piloting. The single best source on the petrodollar stress test running through the crisis.

Treason in the Futures Markets, Paul Krugman on Substack, 24 March 2026. The $580 million oil-futures trade documented by the Financial Times fifteen minutes before Trump’s Iran policy reversal. The most uncomfortable single piece of evidence that US insiders profit from sustained crisis. Read this one if you read nothing else.

How the Iran War Just Broke European Chemicals, European Business Magazine, 14 May 2026. Markus Steilemann (Covestro CEO, Cefic president) in the Financial Times saying the Iran war has pushed European chemicals beyond what the 2022 Russian gas crisis nearly delivered. Cefic capacity utilisation at 62-68% versus 80% structural threshold.

BASF Raises Prices Up to 30% as Iran War Disrupts European Chemical Supply Chains, SupplyStatus, 18 March 2026. Documents the BASF 30% price increase, the Huntsman €200 per tonne gas surcharge, and the VCI annual press conference where Wolfgang Grosse Entrup refused to provide a 2026 forecast.

Strait of Hormuz Crisis 2026: Full Timeline and Ocean Freight Impact, Sea Vantage, April 2026. Contains the 10 March traffic data (two outbound crossings, zero inbound, versus pre-crisis average of 3.29 per day), the Mussafah-2 tugboat strike, and the Hapag-Lloyd Nils Haupt closer-quote about logistics continuing after any ceasefire.

Hormuz disruption reshapes bunker shipping fuel flows, Arab Gulf Business Insight, April 2026. The Singapore VLSFO benchmark 76% surge to $988.5 per tonne. The bunker market data that makes the shipping reroute cost legible at the operational level.

East of Suez, Vol IV: Who’s the counterparty?, Unpacking Complexity, 27 May 2026. Names the IRGC veto, the Vahidi triumvirate, the Pakistani Naqvi-Vahidi backchannel, and the Iranian elite gold flight as confirmation signal. Vol V is the cascade running on top of the counterparty problem Vol IV identifies.

‘Not a Chance Hormuz Opens’: How Wall Street’s new NACHO trade bets on a prolonged oil shock, CNBC, 8 May 2026. The clearest write-up of the NACHO trade, with eToro market analyst Zavier Wong on insurance pricing (war-risk premium still around eight times pre-war levels) and Vassili Gkionakis at Lombard Odier on the equity-versus-rates positioning split.

Wall Street piles into ‘NACHO’ bet on looming oil shortages, Fortune, 8 May 2026. Includes the Krugman endorsement (”I never bought into the TACO meme. But NACHO looks right”) and Mark Hackett of Nationwide on the oil-bond-equity divergence.

The Repricing of Hormuz (East of Suez Vol I), Unpacking Complexity, April 2026. The original framing piece. Identifies the IRGC as the Larak Island toll-extractor and lists the IRGC hardliners as the most likely deal-breaker on any concession. Vol V is the cost paid by everyone else while that prediction holds.

The short why. Vol I called this in April. Six weeks later the IRGC has put the prediction in writing, a Pakistani backchannel has confirmed the operational track, and Iranian elites are converting cash to gold while oil traders sell war premium. The Doha delegation can sign the MOU. Whether it gets implemented is a separate question, controlled by separate actors with private incentives to keep the crisis running.

Vol I opened by noting that Iran’s foreign minister and deputy foreign minister gave contradictory public positions on the same afternoon, deliberately, so the regime could walk back either. Vol I also named the IRGC as the Larak Island toll-extractor and listed “a succession inside Tehran produces a winner who can sell this concession without being overthrown by the hardliners in the IRGC” as one of five conditions for durable resolution. The 10-20% probability range for durable resolution rested on that condition. Vol IV adds what was missing in April: the names, the Pakistani backchannel, and the elite money behaviour that turns the structural prediction into an observable equilibrium.

Iran is now functionally two actors with two negotiating positions. The diplomatic track lives in Doha and runs through three people: Foreign Minister Abbas Araghchi, Parliament Speaker Mohammad Bagher Ghalibaf, and Central Bank Governor Abdolnaser Hemmati. They are negotiating sanctions relief, asset releases, mine clearance, and Hormuz reopening. The IRGC track operates through the Pasdaran-linked press, principally Fars News Agency and Tasnim, with senior IRGC commander Mohammad Ali Jafari as the public voice. They are saying the strait stays under Iranian management regardless of what gets signed.

The question is which Iran the MOU actually binds.

What the IRGC track is saying in public

On 11 May, IRGC senior commander Mohammad Ali Jafari said there would be no further negotiations until “the war is ended on all fronts, sanctions are lifted, frozen funds are released, war damages are compensated, and Iran’s sovereignty over the Strait of Hormuz is recognized.” The quote ran in French via LaPresse News citing Fars. It listed five conditions, the last of which is the entire issue.

On 24 May, Fars News Agency wrote that Trump’s claim the strait will return to its pre-war status is false because “based on the latest text exchanged, in the event of a potential agreement, the Strait of Hormuz will continue to be under Iran’s management.” That is the IRGC’s main media channel publicly disclosing a substantive position from inside the negotiations, three days before the US-Iran delegations met in Doha.

Tasnim, also IRGC-linked, ran a parallel piece saying the strait “will not return to pre-war status” under any agreement.

These are not maximalist statements from minor figures. Fars and Tasnim are the IRGC’s media voice, and Jafari is one of its longest-serving commanders. When this constellation says the strait stays under Iranian management, the actors who control the mines, the boats, and the Pasdaran Gulf Security Authority are saying it. The Doha delegation does not control any of those things.

What the Doha track is saying in public

The Iranian delegation arrived in Qatar on 25 May. The composition was Araghchi, Ghalibaf, and Hemmati: three principals from three institutional tracks, with no IRGC representation. The composition is the message.

Foreign ministry spokesman Esmail Baghaei told the Tehran press on Monday: “We do not charge tolls.” Services are provided in the strait, and those services “require charging fees.” This is the same fee with a UNCLOS Article 38 defence drafted by foreign ministry lawyers around it. Baghaei also said Iran was working with Oman on a joint mechanism for safe passage. Muscat has neither confirmed nor denied the framing, which is itself a position.

The same press conference produced a line that ran across Iranian social media and through several Western outlets: “The agreement is both very far and very close.” That is a foreign ministry spokesman acknowledging, in public, that his own delegation does not control the variable that determines the outcome.

The Doha track and the IRGC track produce different outcomes if either is binding: a working agreement, or a signature without operational delivery.

Who actually runs Iran

Ali Khamenei was assassinated on 28 February. An Interim Leadership Council held authority for one week before Mojtaba Khamenei was selected as Supreme Leader on 8 March, under direct IRGC pressure that bypassed the constitutional process. He has not appeared in public for extended periods. Analysts at Atalayar describe him as “a puppet of the IRGC, or a non-entity.” The TIME profile of Iran’s new leadership group, by SWP’s Hamidreza Azizi, makes the same point with more diplomacy: the Supreme Leader is “no longer the ultimate source of authority.”

The actual decision-making triumvirate, per Gustavo Aristegui’s analysis in Atalayar: Brig. Gen. Ahmad Vahidi (IRGC Commander-in-Chief), Mohammad Bagher Zolghadr (Secretary of the Supreme National Security Council), and Mohsen Rezaei (military adviser to the Supreme Leader). All three are IRGC hardliners.

The Institute for the Study of War, in its 21 May report cited by Al Arabiya and Euronews, states the position bluntly: “Vahidi and members of his inner circle have likely consolidated control over not only Iran’s military response in the conflict but also Iran’s negotiations policy.”

President Masoud Pezeshkian, elected in 2024 on a reform platform, wants a ceasefire. Iran International reported on 21 April that the IRGC has blocked his presidential appointments and assumed control over key state functions. He is the elected executive of a republic in which the executive no longer controls the security apparatus.

Vahidi himself is the structural problem. Born 1958 in Shiraz, founding commander of the Quds Force, former Defence Minister under Ahmadinejad, former Interior Minister under Raisi. Wanted by Interpol over the 1994 AMIA bombing in Argentina. Sanctioned by the US and EU. The IBTimes UK profile notes that he has “reinforced hardline positions, resisting concessions on Iran’s nuclear programme and long-range missiles while directing IRGC activities around key waterways.”

The Vahidi-Naqvi backchannel

Pakistani Interior Minister Mohsin Naqvi visited Tehran twice in less than a week in mid-May. The Express Tribune, Dawn, Pakistan Today, and Aaj English TV all confirmed via Pakistani official sources that Naqvi met Vahidi. The interior ministry stated the meeting “focused on efforts to revive stalled talks between the United States and Iran.” PTV News released video footage of Naqvi greeting Vahidi.

Iran’s state-run IRIB denied the meeting and claimed the circulating photographs were from 2024. Euronews reported the denial.

Both can be true in the technical sense. The meeting happened, and the Iranian state does not want it acknowledged because admitting that Vahidi receives US messages directly through Pakistan destroys the legal fiction that Pezeshkian-Mojtaba are running the negotiations. Dawn reports that Pakistani Chief of Defence Forces Field Marshal Asim Munir is expected to follow Naqvi to Tehran for an analogous meeting. ISNA, the Iranian Students News Agency, frames Munir’s coming visit as an effort to “bridge the gaps and help reach a formal declaration of understanding.”

That is Washington bypassing the Doha foreign ministry track and routing directly to the IRGC chief through Pakistani military-to-military channels. The Doha process is the public ceremony; the real conversation is in Tehran.

The April 17 precedent

After the Islamabad talks collapsed in April, Araghchi posted on X on 17 April that the Strait of Hormuz was “completely open.” He also signalled flexibility on nuclear enrichment and on Iran’s support for regional proxies. Hardliners overruled him publicly within hours. The strait stayed closed, the flexibility was withdrawn, and Araghchi and Ghalibaf returned to Tehran to sharp criticism for offering concessions. The template was set in April. The foreign minister negotiates and the IRGC vetoes.

The May 25 CENTCOM strikes confirm Washington has read the room

The US strikes Monday targeted two IRGC boats laying mines in the strait, a SAM site at Bandar Abbas, and reported launch sites near Sirik and Jask. The targets were the mine-layers and the air defence covering them, not foreign ministry offices or anything Pezeshkian touches. CENTCOM is degrading the operational capacity of the actor that controls the mines while the State Department continues talking to the actor that does not.

Trump posted the same day that negotiations were “proceeding nicely.” Both statements are coherent if you accept the counterparty problem. The strikes are pressure on the IRGC, the talks are with the foreign ministry, and the question is whether sustained kinetic pressure can make Vahidi accept terms his publicly stated position refuses. The IRGC has not yet shown signs of accepting.

Follow the money: who benefits from sustained crisis

Mines were laid this week. CENTCOM struck the mine-layers on 25 May. The mining will continue. The question worth asking is who actually profits from the strait staying broken, because the people who profit are also the people who can keep it broken.

The external beneficiaries are who you would expect, at a scale you might not.

Russia is the single biggest winner outside the Gulf. Russian Urals crude went from a $25-per-barrel discount to Brent in late February to a $10 premium over Brent by mid-April per the Atlantic Council’s Energy Sanctions Dashboard. Carnegie Endowment estimated the additional revenue at $8.5 billion per month, with $5 billion of that flowing into state coffers. Russia’s daily oil and LNG revenue ran 14% above the February baseline through the conflict period per the Centre for Research on Energy and Clean Air. Indian and Chinese refiners that had been cutting Russian purchases in late 2025 switched back. The G7 oil price cap, lowered to $44.10 per barrel in February, has been completely overrun by physical premiums. Sanctions instruments are no longer the binding constraint on Russian revenue, and the Kremlin can fund another year of the Ukraine war from the Hormuz dislocation alone.

Tanker owners are the next-largest beneficiary. Frontline reported its strongest adjusted quarterly earnings in more than 20 years for Q1 2026: $559 million in profit on $714 million in revenue. VLCC spot rates on the Middle East-to-China route breached $423,000 per day in late March, a level with no precedent in data going back to 2005. Frontline, Nordic American Tankers, and DHT Holdings are up roughly 60% year-to-date. The crisis transferred profit from the cargo owners to the vessel operators, who collect rent from longer routes and tighter capacity.

Saudi Arabia and the broader GCC are pumping more oil through bypass infrastructure at higher prices. The East-West pipeline to Yanbu, the ADCOP pipeline to Fujairah, and accelerated Duqm capacity are now strategic assets generating premium economics. US shale producers, with WTI in the low $90s, are extending Permian drilling programmes. Reinsurers are repricing Gulf marine exposure on permanent terms per Vol I’s 15-25 bps estimate. Defence contractors are running sustained-operations procurement. Indian refining margins are inflated by access to discounted Russian crude and product premiums into Europe.

The Iranian beneficiaries are the structural answer to the IRGC veto.

Vol I noted the Larak Island toll route was generating $100 million-plus per week into IRGC hard-currency accounts. That revenue stream did not exist in December 2025. It exists now because the crisis exists, and it goes directly into IRGC accounts that bypass the central bank entirely. The PGSA service fees announced in May add another revenue layer on the same structural basis.

Vahidi himself is the beneficiary of his own command. He went from deputy commander in December 2025 to acting commander on 1 March to consolidated control over negotiations policy by mid-May per the ISW assessment. Each week of sustained crisis strengthens his bureaucratic position. Mojtaba Khamenei, with no religious authority of his own, depends on external threat to legitimize his clerical-IRGC alliance. Without a war, the constitutional questions about his selection become live. War freezes the political dynamics in his favour.

Then there are the elite financial signals from inside Iran. The Tehran Stock Exchange reopened on 19 May after an 80-day war shutdown. TEDPIX sat at 3.76 million points, down from its 4.5 million peak at the start of 2026, and energy and steel companies hit by US-Israeli strikes were excluded from the trading session per Al Jazeera. The regime is hiding damage behind a confidence ceremony. The rial closed at 1.32 million to the dollar on 25 May, down 3,047% year-on-year. Inflation runs at 50% officially, food at 112%, cooking oil over 200%. Average monthly income contracted to around $200 in early 2026, before the war.

Now the lateral signal. Iranian gold coins, the traditional inflation hedge and elite asset, are trading at widening premiums to their metal content. The Quarter Coin jumped 16.15% in a single session on 8 May. The bubble premium reflects fear-driven buying, and Iranians are converting cash to hard assets even as the regime signals a peace deal. Set this against Brent crude at $97, down more than 10% in a week. Global commodity traders are pricing in a reopening and a successful deal. Iranian elites with insider visibility are pricing in something else. Oil markets are deep and full of professionals reading the same news. The Iranian gold market is shallow, restricted to Iranians, and reflects what people see from inside the regime. One of these groups is wrong.

The synthesis.

The people with operational power to end the crisis (Vahidi, the triumvirate, Mojtaba’s inner circle) are also the people whose personal economic and political position depends on the crisis continuing. The MOU asks them to give up the Larak toll revenue, their bureaucratic primacy, their external-threat political cover, and the elite asset conversion window in exchange for an asset release that benefits the central bank and the foreign ministry. The Larak revenue flows to the IRGC; the asset release goes to the central bank. The Doha delegation cannot deliver what the IRGC will not surrender, because the IRGC is rationally extracting from the crisis at a pace that exceeds what any deal can offer.

This is the structural answer Vol I’s fifth condition for durable resolution implied without quite making explicit. The IRGC veto reflects rational organizational economics. The signed-but-unenforced outcome is the equilibrium of who profits from what.

Confidence on this section: High on the external beneficiary numbers (Russian revenue per Carnegie/CREA, tanker earnings per Frontline filings, VLCC rates per Paradox Intelligence) which are documented. Medium-High on the synthesis that IRGC private incentives explain the veto, because the math is straightforward but the counterfactual cannot be tested. Falsification: if Vahidi or Fars publicly endorses MOU terms that surrender the Larak revenue stream, the structural incentive argument is wrong.

What this does to the four scenarios

Scenario one (Oman’s quiet workaround) drops further to Low-Medium. Muscat is not the binding actor. If the IRGC continues mining, the workaround does not deliver no matter what Oman does.

Scenario two (Iranian escalation) holds at Medium. The escalation already happened in slow motion: PGSA fees, vessel boardings, ongoing mine-laying. It is just labelled as a stable regime now.

Scenario three (China brokers a face-saving fudge) stays at Medium-High. The UK-drafted UNSC resolution still faces Russian and Chinese resistance, and Q1 China-Iran trade is down 50% year-on-year. Beijing has reasons to want stability.

Scenario four (Trump erratic) moves up to Medium with widening variance. Monday’s strikes plus the “proceeding nicely” post is the pattern. Wednesday’s Camp David cabinet meeting is the decision point.

New scenario five: signed-but-unenforced deal. The Doha delegation signs the MOU, sanctions relief and asset releases happen, but the IRGC continues PGSA operations and the mines stay in place. Iran banks the diplomatic win without the operational cost. Probability: Medium and rising. This is the structurally most likely outcome given who is at the table and what the IRGC has publicly committed to.

What to watch over the next seven days

Camp David Wednesday is the immediate inflection point. The outputs will signal whether Washington accepts the foreign ministry as the binding counterparty or insists on IRGC endorsement before signing.

The IRGC’s public posture is the second watch item. Any Vahidi statement, Fars editorial, or Jafari follow-up walking back the “sovereignty over Hormuz” demand. Most likely outcome is continued maximalist framing.

The mines are third. CENTCOM reports of fresh mine-laying or IRGC boats in the strait through the weekend. If the mine-laying continues, the IRGC veto is operational. If it stops, the IRGC may have moved.

Oman fourth. Whether Muscat publicly accepts or rejects the Iran-Oman joint mechanism framing. The base case is silence, which functions as a refusal.

Frozen assets fifth. Whether the $12 billion first tranche moves before any mine clearance occurs. If the money flows before the mines come up, Iran has taken its primary asset without delivering its primary concession.

Lastly the rial and the gold premium. If both move toward stability over the next week, the Iranian elites are pricing in the deal. If gold keeps rising and the rial keeps falling, the elites are pricing in scenario five.

Four takeaways

The counterparty problem is the problem. Vol I named it in April. Most Western analysis still treats Iran as a unitary negotiating actor, and that gap between Vol I’s framing and the consensus is what makes the May evidence so useful: Jafari, Fars, and the Naqvi-Vahidi meeting turn an analytical prediction into operational fact. The deal can be signed by the foreign ministry and ignored by the IRGC, and the IRGC has publicly committed to ignoring the bits that bind it.

The IRGC has been telling everyone its position in plain text: Jafari on 11 May, Fars on 24 May, Tasnim throughout. Anyone reading only the Western coverage of the Doha track is reading half the regime. The other half is publicly committed to a different deal.

The IRGC veto is rational economics, not ideological zeal. Larak toll revenue flows to IRGC accounts at over $100 million per week. The asset release flows to the central bank. Bureaucratic primacy strengthens with sustained crisis. Political cover for Mojtaba depends on external threat. The MOU asks the actors with veto power to surrender private revenue and organizational position for the benefit of their domestic rivals. The structurally rational response is what we are watching in real time: public maximalism, operational continuation, signed-but-unenforced as the equilibrium.

The CENTCOM strikes hit the IRGC while leaving the Doha negotiators alone. Washington is signalling that it understands the structure. Whether sustained kinetic pressure can change the IRGC calculation is the open question. The Iranian elite gold flight suggests insiders think kinetic pressure cannot work; oil traders think it can. Pick a side and mark to market in 30 days.

Closer

We are 88 days into the 2026 Strait of Hormuz crisis. The diplomatic and operational tracks have diverged publicly. The Doha delegation is signing one MOU while the IRGC is publicly committed to a different version of the same document. Wednesday at Camp David is the next decision point. After that, either Washington accepts a signed-but-unenforced outcome, or the kinetic dimension widens, or the IRGC quietly moves.

Vol V is whichever of those resolves first.

Vol VI, already in drafting, takes the longer view: what a sustained dual blockade running into 2027 actually costs the global economy. Russia is making $8.5 billion per month in additional revenue. Frontline just posted its best quarter in 20 years. Fertilizer prices are forecast to stay elevated through 2027 per NDSU even under a hypothetical immediate full reopening. European industrial competitiveness is taking another structural step down. Tens of billions are flowing from retail and unhedged consumers to commodity professionals and Russian state coffers. The numbers compound. The longer this runs, the harder they become to walk back, and most G7 policy frameworks were not built for a Hormuz that stays partially closed for a year. Stay subscribed if you want that one.

Fourth in the East of Suez series in Unpacking Complexity. Vol I established the five mechanisms of permanent repricing and identified the dual-track regime hedging, the IRGC as Larak Island toll-extractor, and the IRGC hardliners as the most likely deal-breaker (one of five conditions for durable resolution). Vol II named the operational regime that emerged from the IRGC track and identified Oman as the most likely tactical workaround. Vol III responded to the 20 May PGSA map and the 23 May GCC IMO letter and ranked four scenarios. Vol IV gives the actors Vol I named structurally their operational identities (Vahidi, the triumvirate, Mojtaba’s weakness), identifies the Pakistani Naqvi-Vahidi backchannel, and adds the Iranian elite gold flight as a measurable confirmation signal.

Pushback welcome from Iran specialists, IRGC analysts, Pakistani diplomatic correspondents, and readers with maritime insurance, commodity trading, or Iranian capital markets expertise. The counterparty argument needs network testing and I am open to being persuaded the foreign ministry track is binding.

Reading list, with the why

Iran, Pasdaran: “Tehran has made no commitment on nuclear,” LaPresse News, 24 May 2026. The Fars quote in French, citing the Pasdaran-linked agency directly. This is the IRGC putting its position into the press three days before the Doha delegation arrived in Qatar.

Iran, Pasdaran: “No negotiations until war is ended on all fronts,” LaPresse News, 11 May 2026. Jafari’s five-condition statement. The fifth condition is the entire dispute.

IRGC seizes control of Iran state functions, blocks president, report says, Fox News citing Iran International, 21 April 2026. The structural claim that the IRGC has blocked Pezeshkian’s appointments and assumed control of key state functions.

The New Leaders Calling the Shots in Iran, TIME, by Hamidreza Azizi at SWP, 6 May 2026. Profiles Vahidi and Ghalibaf as the actual power figures and notes the Supreme Leader is “no longer the ultimate source of authority.” A reputable Iran specialist saying the structural quiet part out loud.

Iran’s Internal Struggle: Factions, Power and the Paths to Peace, Atalayar by Gustavo Aristegui, 6 April 2026. The triumvirate framing (Vahidi, Zolghadr, Rezaei). A Spanish geopolitical analyst with no stake in the US-Iran story going further than US sources will.

Naqvi meets IRGC chief in Tehran during second visit in less than a week, The Express Tribune, 21 May 2026. The Pakistani-confirmed Naqvi-Vahidi meeting that Iranian state media denied. PTV News video footage attached. The single most important reporting on the actual negotiating channel.

Pakistan races to prevent US-Iran escalation, Dawn, 21 May 2026. The detail that matters: Pakistani mediation is “complicated by quiet strains,” and Iran is uncomfortable with Pakistan’s coordination with Gulf Arab states and Washington. Also notes that Field Marshal Asim Munir is expected to follow Naqvi for a parallel meeting.

Who is Ahmad Vahidi, Interpol-wanted IRGC general, Euronews, 21 May 2026. The ISW quote on Vahidi consolidating control over both military and negotiating tracks. Also documents the Iranian state denial of the Pakistan meeting, which is itself information.

IRGC chief Ahmad Vahidi emerges as key power broker, Al Arabiya, 21 May 2026. Saudi reporting on Vahidi. Notable because Riyadh is the regional actor with most to lose from continued IRGC dominance, and the Saudi coverage is closer to the operational reality than Western framings.

Controlled reopening ends Iran’s lengthy stock market shutdown, Al Jazeera, 20 May 2026. The TSE reopened after 80 days but war-damaged companies did not trade. Confirmation the regime is hiding damage. Critical for the adjacent indicators argument.

Iran’s currency falls to new low as US blockade, sanctions impact trade, Al Jazeera, 29 April 2026. China-Iran trade Q1 down 50% YoY. March down 80%. The numbers behind the rial collapse. The China-as-customer relationship is broken, which changes Beijing’s incentives.

Iran delegation in Qatar seeks deal on frozen assets and Hormuz blockade, Euronews, 25 May 2026. Full Doha delegation composition: Araghchi, Ghalibaf, Hemmati. Confirms Qatar replaced Pakistan as primary mediator on the formal track while Pakistan kept the backchannel.

Brent crude oil futures stay near $97, Trading Economics, 26 May 2026. The benchmark the global oil market is using to price in a deal. The counterweight to the Iranian elite gold flight signal. One of them is wrong.

What the Russian Energy Sector Stands to Gain From War in the Middle East, Carnegie Endowment, March 2026. The $8.5B per month Russian revenue increase, Urals discount collapse from $25 to $15 between 27 February and 6 March. The single best documentation of who is winning outside the Gulf.

Frontline Sees Strongest Quarter in Two Decades as Hormuz Crisis Reshapes Tanker Trade, gCaptain, May 2026. Frontline Q1 profit of $559M, the highest in 20+ years. VLCC TCE at $103,500 per day average. The Hormuz crisis converted into earnings on actual filings.

Energy Sanctions Dashboard, Atlantic Council, updated 21 May 2026. Russia trading at premiums of up to $10 over ICE Brent. The G7 oil price cap rendered functionally irrelevant by physical market premiums. Confirms the sanctions regime has been overrun by the Iran-induced market dislocation.

Tanker Freight Rates Hit Records as Hormuz Traffic Collapses, Paradox Intelligence Research, April 2026. VLCC day rates at $423,000 on Middle East to China route in late March. Documented all-time records, no precedent in data going back to 2005. The tanker rent extraction in hard numbers.

The Repricing of Hormuz (East of Suez Vol I), Unpacking Complexity, 17 April 2026. The original framing piece. Opens with the regime running two contradictory public positions on the same afternoon, identifies the five mechanisms of permanent repricing, names the IRGC as the Larak Island toll-extractor, and lists the IRGC hardliners as the most likely deal-breaker. Vol IV is the operational confirmation of Vol I’s structural prediction.

East of Suez, Vol III, Unpacking Complexity, 23 May 2026. The previous instalment. Vol IV updates the four-scenario ranking with May evidence and adds a fifth scenario (signed-but-unenforced).

Iran published a map claiming supervisory jurisdiction over the inbound shipping lane near Oman’s Musandam Peninsula. Five GCC states formally objected. Oman did not sign.

Vol II went out on 19 May. It argued that Iran’s Persian Gulf Strait Authority was building a two-tier maritime regime over the Strait of Hormuz, that Oman’s geographic control of the inbound lane was the load-bearing constraint on Iran’s reach, and that Tokyo’s silent acceptance of Iranian-coordinated transit was the leading indicator of the postwar maritime order ending. The PGSA map landed the next day.

The map (20 May)

On Wednesday 20 May, the day after Vol II went out, the PGSA published a graphic and accompanying statement on X defining what it calls a “controlled maritime zone.” The zone is bounded on the east by a line running from Kuh-e Mubarak in Iran across to south of Fujairah in the UAE. The western boundary runs from the tip of Qeshm Island to Umm al-Quwain. If you draw those lines on a chart, the eastern line passes south of the Musandam Peninsula. The PGSA zone now claims supervisory jurisdiction over what the IMO traffic separation scheme has called Omani inbound waters since 1979.

Iran’s preferred framing is supervisory. The legal language stays inside what a careful lawyer can call regulatory cooperation. The operational language says vessels need to coordinate with PGSA, supply cargo manifests, crew nationalities, vessel origins, destinations, and previous flag registrations. The Institute for the Study of War assessed on 22 May that the map represents “a territorial claim on the sovereign territory of another country”.

The 20 May map is the move Vol II said Iran would probably not make. Vol II’s reasoning was that Iran needed Oman onside, and that Iran’s institutional path would stop at the IMO traffic separation scheme. Both parts were partially wrong. Iran has gone past the scheme and claimed jurisdiction over the inbound lane, while still talking to Muscat about cooperation. Iran is doing both at once.

The GCC pushback (23 May)

This week, the GCC five (Bahrain, Kuwait, Qatar, Saudi Arabia, and the UAE) sent a joint letter to the IMO warning commercial vessels against engaging with PGSA or transiting the waterway using Iran’s designated zone. Oman did not sign.

This is the first formal regional response. The GCC majority will not accept PGSA as a fait accompli. Oman is now publicly outside both the Iranian regime and the regional opposition, which is the diplomatic space the Qaboos-era playbook has occupied since the 1970s. Vol II said Muscat would decline to enforce Iran’s rules. The non-signature goes further.

The Habibi reading, partly overtaken

Mehrdad Habibi, quoted in Al Jazeera on 21 May, said the PGSA arrangement was “unprecedented” and argued that “there would not be such an outcome, unless there is a complete coordination between the GCC countries and Iran, with the approval of major international powers, such as China and the United States”.

That was Habibi’s structural claim. The PGSA regime could only function on tacit acceptance from the GCC, Beijing, and Washington. None had publicly objected at the time he spoke. The GCC has since objected. Beijing and Washington still have not. The US-China silence is what keeps the regime running.

Rubio is working NATO foreign ministers in Helsingborg this week on Hormuz reopening.

Updated scenarios

The four scenarios from Vol II hold in structure. The likelihood ratings have moved since publication.

Scenario one (Oman’s quiet workaround holds) firms back to High.

Oman’s non-signature on the GCC letter sits inside a public regional majority position, which gives Muscat more diplomatic cover than it had on 19 May. The workaround still requires Iran to not directly enforce against Omani-lane traffic. The diplomatic geometry now leans Oman’s way.

Scenario two (Iranian escalation, contained Gulf war) edges up from Medium to Medium-High.

The GCC pushback raises the cost to Iran of accepting de facto containment. If Tehran reads the letter as an open challenge, the institutional path becomes harder to defend domestically.

Scenario three (China brokers a face-saving fudge) drops from Medium to Medium-Low.

The GCC letter has demonstrated regional willingness to mount the open challenge the fudge would need to suppress. China can still broker. The political space has narrowed.

Scenario four (Trump erratic) holds at Low but with widening variance.

Rubio at NATO is the conventional play. Trump can override it at any time, and the longer the formal multilateral process runs without resolution, the more attractive the cross-cutting option becomes.

Probability mass has moved back toward scenarios one and two, away from scenarios three and four.

Britain, 1968, still

Vol II’s Britain 1968 parallel argued that the postwar maritime order was unraveling without a single announcement. The PGSA map is, technically, an announcement, the kind of bureaucratic document people only read in retrospect. The GCC IMO letter is the kind of paperwork that signals a regional position without committing to operational consequence. The 1968 Defence White Paper that formalised Britain’s East-of-Suez withdrawal had the same form, sitting in Hansard archives ignored while everyone debated Vietnam.

Britain took three years between 1968 and 1971 to do what Iran has done in 12 weeks. Both transitions ran on paperwork. The pattern only becomes obvious in retrospect. The GCC letter this week is the first piece of paperwork pushing back. Whether more follow is the open variable.

What to watch in the next 14 days

Oman first. A formal Omani statement either way, in support of the GCC letter or in continued silence, is the load-bearing signal. Continued silence is the most likely outcome.

China second. Wang Yi or the Chinese foreign ministry briefing on the map or the GCC letter is the analytical inflection point. If Beijing stays silent through next week, the regime keeps its US-China cover.

The GCC five third. Whether the letter is followed by operational moves, such as restricting GCC-flagged vessels from using PGSA, or whether the letter is the limit of the response.

Vessel volume fourth. Pre-conflict baseline was around 100 transits per day. The 26 PGSA-coordinated vessels on 20 May was a surge. If coordinated transits climb above 60 per day, the operational regime is settling. If they stay below 20, the GCC letter is biting.

Trump fifth, and always. Rubio is running the conventional play. Trump can override it at any time.

Three takeaways

The map matters more than the toll. Tolls, Bitcoin insurance, and the PGSA launch were institutional plumbing. The 20 May map adds the territorial claim that makes the regime structural.

Muscat is the actor to watch this week. The GCC letter narrowed the diplomatic space. Oman’s non-signature placed it publicly outside both the Iranian regime and the regional opposition.

The series is moving fast. The drift thesis from Vol II survived four days before the territorial claim, the regional response, and the institutional ramp-up all landed inside the same news week. Vol IV will be whichever of the four scenarios resolves first. Scenario one is back in the lead.

We are 84 days into the crisis. The most concrete institutional move so far is the 20 May map. The first formal pushback came three days later from the GCC. What happens next depends on whether China and the United States break their silence.

Third in the East of Suez series in Unpacking Complexity. Vol I (The Repricing of Hormuz, April) established the five structural mechanisms by which the strait was being permanently repriced. Vol II (19 May) named the operational structure and identified Oman as the load-bearing actor. Vol III responds to the PGSA map publication on 20 May, the ISW assessment on 22 May, and the GCC IMO letter of 23 May.

Sources and Further Reading

• Vol II of the East of Suez series, “East of Suez, Vol II,” Unpacking Complexity, 19 May 2026

• “Iran asserts jurisdiction over UAE and Oman waters in new Strait of Hormuz map,” Euronews, 22 May 2026 (contains the ISW assessment)

• “Iran Draws New Maritime ‘Oversight Zone’ Across Strait of Hormuz,” gCaptain, 20 May 2026

• “Iran establishes Persian Gulf Strait Authority,” Container News, 20 May 2026

• “Maths behind Hormuz toll: Is paying Iran for transit cheaper than blockade?” Al Jazeera, 21 May 2026 (the Habibi interview)

• “Iran claims it coordinated passage of 26 vessels out of Hormuz in 24 hours,” Al Jazeera, 20 May 2026

• “First coordinated VLCC transits through Hormuz raise cautious hopes of a thaw,” Splash247, 20 May 2026

• Institute for the Study of War, Iran Updates (rolling daily assessments)

• “Israel/US-Iran conflict 2026: Reopening the Strait of Hormuz,” House of Commons Library Research Briefing CBP-10636

Updated 19 May 2026. Since first drafting: Iran launched the Persian Gulf Strait Authority on 18 May as the formal institutional mechanism for managing Hormuz traffic, with confirmed tolls of around $2 million per vessel. A Bitcoin-backed insurance facility for transiting ships was announced the same day. Brent crude is trading at $118-122. The structural thesis below has been validated by events at a speed I did not expect.

Sometime in mid-March, Japan stopped being a postwar maritime ally of the United States. There was no announcement, no joint statement, no diplomatic note, no headline in the Nikkei. A Japanese-flagged tanker simply transited the Strait of Hormuz under Iranian coordination, paid whatever it paid, and continued to Yokohama. The Iranian foreign ministry added Japan to its friendly list. Tokyo declined to comment.

That is what alliance erosion looks like in 2026. Quiet. Operational. Already done.

The friendly list

Since the US-Israel strikes on Iran on 28 February, the Strait of Hormuz has operated under what Iran calls a coordinated safe-passage regime. In plain language: Iran decides who passes and who pays. The friendly list as of early May includes China, Russia, India, Iraq, Pakistan, Malaysia, Thailand, the Philippines, Bangladesh, and Japan. Western-flagged vessels reroute via the Cape of Good Hope, adding around 6,000 nautical miles and three weeks to the trip. War-risk insurance on Hormuz transits sat at roughly twenty times pre-war levels by April.

On 18 May, Iran institutionalised the regime by launching the Persian Gulf Strait Authority, an official body that issues transit permits and reportedly collects around $2 million per vessel. The same day, the Ministry of Economy announced a Bitcoin-backed insurance facility for ships transiting the strait, a deliberate workaround to exclusion from Western marine insurance markets. The two-tier maritime regime now has its own administrative agency and its own crypto-rails.

The friendly list is not a clean shield. The Chinese-flagged JV Innovation was struck while transiting on 8 May, with 17 Chinese crew aboard. US CENTCOM attributed the strike to “Iran-aligned naval forces” without naming a unit. Attribution is contested and Iran is unlikely to have ordered an attack on its single most important diplomatic patron, but the practical effect is the same: even friendly-flag vessels now pay a risk premium, and Beijing has skin in the game it did not previously have.

The legal status of all this is contested. Chatham House published a careful analysis on 16 April pointing out that Iran’s discriminatory pricing offends UNCLOS Article 38 on transit passage through international straits. The argument runs into the same problem every UNCLOS argument runs into, which is that there is no court with jurisdiction to enforce it. China and Russia vetoed the Bahraini UNSC resolution on 7 April. The UK and France announced an international defensive mission on 17 April, conditional on a sustainable ceasefire that does not yet exist. We are operating in a post-UNCLOS Gulf, and the lawyers can sort it out later.

Musandam

Oman.

Hormuz is 21 nautical miles wide at its narrowest point. Traffic moves through two unidirectional lanes under an IMO traffic separation scheme that has been in place since 1979. Outbound, away from the Gulf, runs mostly through Iranian waters. Inbound, returning to the Gulf, runs mostly through Omani waters near the Musandam Peninsula. Half the strait, by operational geography, is not Iranian at all.

Oman’s Transport Minister, Said al-Maawali, stated on the record in April that imposing fees on transit contradicts the international agreements Oman is committed to. Oman ratified UNCLOS in 1989. Iran proposed a joint 10-point plan under which the two countries would jointly regulate strait traffic, including coercive measures against non-cooperating vessels. Oman publicly declined the fee component. On 18 May, Iranian FM spokesman Esmail Baghaei claimed that technical teams from both countries had met to coordinate strait passage. Muscat did not confirm. On 17 May, Oman publicly welcomed Germany’s call for the strait to be reopened under international law. Read together, the diplomatic signals are theatre. Muscat is running the playbook it has used since the Qaboos era. It maintains active channels with every party and avoids formal alignment with any of them.

In practice, Oman could declare its territorial waters open to free transit and strip Iran’s toll regime of half its enforcement geography. It probably will not do so explicitly, because Iran is a neighbour and Sultan Haitham bin Tariq has every reason to maintain the Qaboos-era neutrality. Oman does not need to act explicitly. The mere fact of non-enforcement on the inbound lane creates a workaround that any sufficiently competent shipping line can use. A Greek or Norwegian tanker that wishes to avoid paying Iran can transit inbound through Omani waters and reverse-flag for the outbound leg. Brokers in London are already pricing this.

Sulfur, acid, pistachios

Hormuz is famous for oil. Other cargoes are squeezed harder right now.

The Middle East produces about a third of global elemental sulfur, mostly as a byproduct of oil and gas desulfurization. When the strait stopped moving in February, sulfur stopped with it. Sulfuric acid prices CFR US Gulf went from $155 per tonne in late February to $400 by 6 May. Sulfur FOB US Gulf doubled. On 1 May, China, which produces 40% of global sulfuric acid, banned exports of smelter-byproduct acid to protect domestic fertilizer production during spring planting. Electronic-grade acid was specifically exempted, a deliberate carve-out for the global semiconductor industry.

Sulfuric acid sits inside almost every consumer-facing supply chain, which is why a shortage cascades into so many places at once. It feeds phosphate fertilizer (food prices), copper hydromet extraction (every wire and every electric vehicle), nickel and cobalt processing (every lithium-ion battery), uranium leaching (nuclear fuel), titanium dioxide (paints, pigments, sunscreen), and semiconductor fabrication. Around 20% of global copper supply uses solvent extraction-electrowinning, which is a sulfuric acid bath. Chile imports roughly a million tonnes of acid from China annually. Some Chilean producers reported being down to 30 days of supply by mid-April.

LNG belongs in this list too. Qatar ships around 80 million tonnes a year through Hormuz, most of it to Asia. Japan’s friendly status lets Tokyo’s importers keep buying at coordination rates. Korea and Taiwan, not on the list, are not so lucky. European LNG buyers who switched from Russian pipe gas to Qatari LNG after 2022 are now paying the bill twice.

Then there is the friendly-list cargo trade. Iran produces around 250,000 tonnes of pistachios annually, and roughly 90% of global saffron. These now flow to friendly markets at preferential rates and to Western markets via roundabout routing at premium prices. The Christmas-snack market in Germany and the wedding-saffron market in the United States will be quietly more expensive this year, and the supply chain origin will not appear on any retail label.

What you actually pay

Container traffic is the largest consumer story but not the only one. Asia-Europe shipping was already rerouted via the Cape because of Houthi attacks in the Red Sea since 2023. The Hormuz crisis layers on top. A Western-flagged container ship from Shanghai to Rotterdam that pre-2023 transited Suez at around $1,800 per TEU now moves via Cape at around $4,500, with war-risk premiums adding several hundred dollars per TEU on the Hormuz-adjacent legs. Pass-through to retail is a few dollars of markup on a typical electronics item, more on white goods, and disproportionately more on perishables that cannot afford the time penalty.

The petrol pump is where most households notice first. Brent crude is trading around $118-122 against a pre-war baseline of $74-78, retesting $122 on 15 May after the first US-Iran kinetic sea engagement since 2020 near Larak Island. North American gasoline is roughly a third higher than January. UK and EU pump prices are higher again because fuel duty amplifies the crude pass-through. A typical North American commuter is paying well over $1,500 more for fuel this year. The UK equivalent runs into the low four-figure sterling range.

LNG is the next layer for households. Japan’s friendly status keeps Tokyo utilities buying at coordination rates. Korean and Taiwanese utilities pay full premium and pass it through. European households who switched from Russian pipe to Qatari LNG after 2022 are now paying that premium twice over. Winter 2026-27 heating bills in Germany, the Netherlands, and the UK are likely to land sharply above the post-2022 baseline.

Insurance is the invisible layer. When marine war-risk premia move twentyfold, the reinsurance market reprices behind them. Lloyd’s syndicates rebalance. Reinsurance pricing cascades into commercial property, then into homeowner and auto policies through 2027. Consumer insurance premiums in developed markets are likely to rise meaningfully over the next eighteen months from this trigger alone, separate from anything climate-driven. The increases will arrive in small steps across many renewals.

The bulk-commodity hit is the biggest single line. Brazilian fertilizer importers are seeing CFR sulfur at $1,150 per tonne against pre-war levels of $525. Brazil’s 2026-27 grain harvest is at risk if farmers cannot afford to fertilize at peak planting. That feeds into global soybean and corn prices in the second half of 2026, and into Chinese pork prices, because Chinese pork depends on Brazilian soybean meal. Fertilizer is the proximate cause of food price moves at retail.

Electric vehicles take a quiet hit through the sulfuric acid chain. Battery-grade nickel and cobalt processing in Indonesia and the DRC needs sulfuric acid. The annual cost declines the industry had counted on have stalled. Tesla, Ford, GM, BYD all face the same input pressure. The consumer EV market slows on price even as the geopolitical case for it accelerates.

The second-order monetary effect is the one most likely to surprise. If sustained commodity-driven inflation forces central banks to hold rates higher for longer, every mortgage refinancing in the next two years pays the bill. The household impact runs into hundreds of dollars a month for years. Fertilizer crunches and maritime regimes become interest-rate stories.

Four futures, four price paths

The next 90 days run down one of four scenarios. The cost trajectory in each differs by an order of magnitude.

Scenario one: Oman’s quiet workaround holds.

Likelihood: High.

Iranian toll regime hollows out through Q3 as inbound traffic concentrates in Omani waters. A two-tier regime institutionalises. Container shipping settles at 25-30% above pre-war levels and stays there. Sulfuric acid prices ease as Middle East sulfur partially resumes via friendly-flag shipping but remain elevated 50-80% above pre-war. Brazilian fertilizer crunch produces a 2026-27 grain price spike but not a catastrophe. Global food inflation runs 2-3 percentage points above baseline through 2027. Insurance market shifts permanently toward Shanghai and Mumbai pricing for non-Western cargo. London loses share it does not get back.

Scenario two: Iranian escalation, contained Gulf war.

Likelihood: Medium.

Iran mines the Omani lane or attacks an Omani-flagged vessel. UK-France mission goes operational. US re-escalates Project Freedom. Strait closes effectively for 60 to 180 days. Brent crude in the $140-180 range. Container costs double again from current levels. Sulfuric acid CFR US Gulf reaches $600-800. Food crisis becomes acute in import-dependent countries: Egypt, Bangladesh, sub-Saharan Africa. Consumer price index in developed countries jumps another 3-4 percentage points within a quarter. This is the scenario where households notice in a hurry.

Scenario three: China brokers a face-saving fudge.

Likelihood: Medium.

An Iran-Oman Joint Maritime Authority is created on paper. Tolls become token. Chinese soft loans finance Iranian reconstruction against future oil deliveries. Container rates retreat toward 15% above pre-war by year-end. Sulfuric acid normalises within six months. Western insurance loses share permanently but the price shock dissipates. The political consequences exceed the economic ones. China gets credit for solving a crisis the United States could not solve, and the Atlantic alliance loses prestige it cannot rebuild.

Scenario four: Trump goes erratic.

Likelihood: Low, but with wide variance.

Either he declares victory and walks away, which is cost-neutral and prestige-negative, or he launches surprise strikes when bored, which is cost-disastrous and all bets off. The variance on this scenario is so wide it dominates the expected-value calculation for the whole system, which is why investment banks are charging premium rates on Gulf-related hedging.

On the rankings

These are calls, not measurements. The reasoning behind each:

Scenario one rates High because Gulf crises since 1987 have repeatedly settled into uncomfortable equilibria rather than escalating fully. The 1987-88 Tanker War, the 2019 Saudi facility attacks, and the 2020 Soleimani aftermath all bent back toward stability under operational constraint. Oman’s 50-year diplomatic record reinforces the equilibrium prior. China and Russia want stable flows for themselves. The Pentagon prefers de-escalation. The case against the High rating is that PGSA institutionalisation on 18 May suggests Iran is settling in for a long regime rather than seeking exit. That weakens the equilibrium prior somewhat. The historical default still dominates.

Scenario two rates Medium because kinetic escalation is the base rate for Gulf crises that do not settle. The USS Stockdale-IRGC engagement on 15 May and the Chinese-flagged JV Innovation attack on 8 May show the system is already kinetic at a low level. Post-war Iran has reconstituted under Khamenei’s son with less to lose. The case for an upgrade to Medium-High is defensible if you take current trajectory at face value. I am holding at Medium because Iran does not have a structural incentive to escalate against its own toll revenue.

Scenario three rates Medium because Chinese capacity to broker is demonstrated by the March 2023 Saudi-Iran restoration and Xi’s 21 April call with MBS publicly stating Hormuz should remain open. The constraint is timing and Iranian acceptance. Iran wants formal sovereignty validation, not Chinese-engineered ambiguity. The faster the Trump deadline pressure tightens, the lower this rating runs. Could move in either direction in the next 14 days depending on the Pakistan mediation channel’s next round.

Scenario four rates Low because most Trump behaviour is already captured inside the other three scenarios. The announce-and-pause pattern is the alternating-state version of Scenario one. The Low rating reserves likelihood for the distinctively unanchored moves: walking away and declaring victory, or surprise strikes when bored. The variance attached is wide enough that the option-value of hedging against it remains material even with a Low base rating.

The ratings most likely to need revision in the next 30 days are Scenarios two and three, in opposite directions, depending on whether Chinese mediation produces substantive movement before Trump’s “days, not weeks” deadline expires.

Britain, 1968

The United States is not actually driving outcomes in this crisis, China and Russia drive one side through the UNSC veto, while the UK and France build the legal-institutional coalition. Oman controls operational reality through lane geography. Japan signals alliance erosion through quiet acceptance of Iranian-coordinated passage. The United States is reactive, oscillating between Trump’s instincts and the alliance system, alternately announcing blockades and pausing them within 48 hours.

This pattern looks like Britain’s “East of Suez” withdrawal of 1968-71. Harold Wilson’s government pulled back from Gulf commitments after the 1967 sterling devaluation forced a fiscal reckoning. The structure dissolved in stages. Aden in 1967. The Trucial States independent by 1971. The Royal Navy’s Gulf squadron drew down through the 1970s. By the time anyone noticed, the transition was complete. There was no single withdrawal document and no Parliamentary debate framed in the terms history has since used for it.

The same thing is happening to the United States right now. The maritime hegemon function has been parceled out among other actors without a formal transition: enforcement to the UK-France mission, operational geography to Oman, alternative flows to China, and tacit assent to the new regime by Japan. Washington has more ships than anyone else in the Gulf. The rules of the game are being set by others. The Pacific allies have read the trend correctly. Tokyo’s silent defection is the leading indicator.

If you take Japan’s behaviour seriously, the postwar maritime order has already ended in practice.

Five things to watch

One. Whether the UK-France mission language uses “Omani lane support” rather than “freedom of navigation” generally. That phrasing change signals tacit policy alignment.

Two. Any Iranian-flag harassment of an Omani-lane vessel. First incident is the bellwether.

Three. Whether Tokyo publicly comments on its passage arrangement. Continued silence means the arrangement is holding.

Four. Whether the Pakistan mediation channel produces a substantive Round 2 in Islamabad.

Five. Whether ADNOC’s Sultan al-Jaber escalates his public diplomacy. UAE is the most exposed Gulf state and the most angry. They are the canary.

Three takeaways

One. The postwar maritime order is unraveling. Tokyo’s quiet acceptance of Iranian-coordinated transit is more telling than anything coming out of Washington this quarter. When the most-aligned Pacific ally hedges operationally, the underlying alliance structure has already shifted.

Two. Most household cost will appear in bills people do not immediately connect to a Gulf crisis. Home insurance renewals, winter heating, the slight EV markup, the mortgage refinancing that comes due in 2027. Petrol gets the headlines. The other lines compound.

Three. Industrial chemistry is a faster signal than political news. Sulfuric acid prices and war-risk premia moved before the ceasefires were announced and before the political pieces ran. For early reads on what is happening in the maritime order, watch chemicals and shipping rates, not press conferences.

We are not watching a Gulf crisis. We are watching the consequences of a maritime hegemony that has already ended, and the supermarket aisle is where most of us will read the bill.

Pushback welcome from readers in marine insurance and reinsurance, commodity trading on the sulfur and copper side, Asian shipping line operations, Saudi or UAE strategic policy, and especially anyone with eyes on Japanese foreign or trade ministry positioning.

Sources and Further Reading

• “The Strait of Hormuz, shipping, and law,” Marc Weller, Chatham House, 16 April 2026 (the load-bearing legal analysis on UNCLOS Article 38 and Iran’s transit fees)

• “US-Iran ceasefire and nuclear talks in 2026,” House of Commons Library Research Briefing CBP-10637

• “Israel/US-Iran conflict 2026: Reopening the Strait of Hormuz,” House of Commons Library Research Briefing CBP-10636

• “Oman’s Legal Position Challenges Iran’s Bid to Impose Hormuz Fees,” Alhurra, April 2026 (the Omani lane geography argument and the Al-Mulla analysis)

• “Who controls the Strait of Hormuz? Iran’s toll plan could reshape global maritime order,” Sanjeet Ruhal, TRT World, May 2026

• “Trump Iran deadline: Hormuz Authority talks stall,” Israel Hayom, 18 May 2026 (on the PGSA launch and Iran-Oman coordination claim)

• “China to Ban Sulfuric Acid Exports as War Hits Supply,” Bloomberg, 10 April 2026

• “Global sulfur prices surge on Middle East disruptions, China export ban,” Marina Silveira Lima, S&P Global Platts, 7 May 2026

• “China’s sulfuric acid restrictions set to squeeze miners,” S&P Global Energy CERA, 16 April 2026

• “China Halts Sulphuric Acid Exports as War Squeezes Global Supply,” Exiger, April 2026

• “How China’s Sulphuric Acid Ban Will Hit Metals & Fertiliser,” Supply Chain Digital, April 2026

• “Strait of Hormuz Closure 2026: What It Means for Your Supply Chain and Shipping Routes,” Carra Globe (updated May 2026, vessel and mariner stranded figures)

• 2026 Strait of Hormuz crisis (Wikipedia, regularly updated, useful for incident timeline)

In 1845, roughly a million people in Ireland depended on a single potato cultivar called the Lumper. It was a mediocre potato by every measure that should have mattered. Bland, watery, the agricultural equivalent of taupe. But it grew on bad land, and tenant farmers had only bad land, so everyone planted Lumpers. When Phytophthora infestans arrived from the Americas, it did not need to be sophisticated, the path was already laid. One genotype, one farm, one country, gone in four years. A million dead, two million emigrated, and the Atlantic crossing changed politics on three continents.

The famine is usually told as a story about British policy, which it also is. Underneath that policy failure sits a structural one. The monoculture made the famine inevitable. Phytophthora supplied only the timing.

This is not a piece about potatoes.

What an LLM actually does

Mechanically, a large language model samples from a high-dimensional space of plausible next tokens given training data. Different models trained on overlapping corpora with similar architectures will converge on similar samples. That convergence is the design. The output is statistically average, with average defined relative to whatever the model ate. At the moment, what the model ate is mostly Reddit, Wikipedia, and the collected works of every blogger who has ever used the word “stoked”.

You can see the visible markers if you look: the phrase “delve into” appears in every academic abstract now, “Let’s unpack” has become the verbal tic of the educated class, the em dash has colonised LinkedIn, most consulting decks has the same three-bullet rhythm. Every product launch announcement reads like every other product launch announcement, because they all started life as the same prompt t\:o the same model. These are surface tells. The thing underneath is harder to see, which is that the structure of how arguments get made is converging too.

Why it feels like augmentation

This is the tricky part. From inside the compression, you feel sharper, you finish the email faster, you draft the memo cleaner, your code compiles on the first try more often. Each interaction is a small win. The losses sit at the population level, invisible to anyone counting wins one user at a time. The cost is creativity and originality, the eccentric paragraph, the weird metaphor, the thing a particular human writes because their particular brain produces it, all of it gets sanded toward the median.

What looks like a swarm of LLM-augmented humans is closer to a field of Lumpers.

The pattern is not new

If this sounds dramatic, consider that it has happened before, and recently. In the run-up to 2008, almost every major bank was using a variation of the same risk model, the Gaussian copula formula popularised by David X. Li in a 2000 paper. The model was elegant and tractable. It was also wrong in a particular way that became invisible because everyone used it. When US housing prices started correlating across regions, every bank’s models failed at the same time. Debt and fraud helped, but the crash was structural. Everyone was running the same equation, and the equation could not see what was coming because it had been written before the thing existed. The formula had a fan base inside quant desks for almost a decade.

Or consider the Cavendish banana, the one in every supermarket. It is the second monoculture banana in living memory. The first, the Gros Michel, was wiped out commercially by Panama disease in the 1950s. The industry replanted with Cavendish, a less flavourful variety with one virtue: it was immune to the strain that killed Gros Michel. A new strain, Tropical Race 4, is now methodically eating Cavendish plantations from Asia outward. Banana people have been issuing warnings for twenty years. Banana people, as it happens, do not control the marketing budget.

The pattern repeats. Optimise for efficiency at the cost of variance. The system runs without trouble until it meets something the design did not anticipate. Then it fails everywhere at once.

Where this leads

Project the curve forward. As LLM outputs flood the open internet, they become training data for the next generation of models. Researchers call this “model collapse,” and the current evidence suggests it is real and not easily fixed. The signal-to-noise ratio of the latent space worsens with each iteration and the homogenisation compounds. In ten years, half the internet will be a hall of mirrors made of mildly helpful blog posts trained on mildly helpful blog posts.

In policy terms, you get a world where ministerial briefs across G20 capitals start sounding like one another. Where boardroom strategy decks across Fortune 500 firms reach the same three-option framework. Where journalists in Nairobi and Helsinki and São Paulo file pieces with the same structural rhythm because they all reached for the same draft assistant on deadline.

None of these people are colluding. They are just planting Lumpers.

The deeper risk is that systemic errors propagate without anyone catching them. When everyone reasons through the same model, the model’s blind spots become the system’s blind spots. The 2008 analogue, applied to every domain of human judgement, takes roughly a decade to arrive on current trends.

I would not bet against it. I will also not be writing a policy paper recommending we do something about it, because that policy paper will also be written by an LLM and will say the same thing as everyone else’s policy paper.

The hedge

The technology is here. The people refusing it will be poorer, not wiser. The work is to keep heterogeneity alive on purpose.

For individuals, treat LLM output as a first draft to argue with. The moment you accept it as a finished thought is the moment the compression wins. Read writers from before the training cutoff. They will sound strange because they are. Keep a few people in your life whose minds work in ways that surprise you, including some who irritate you. (Especially those). If you write for a living, write at least one piece a month without any assistance, including the parts of your own brain that have absorbed the AI cadence by osmosis. Crop rotation, basically.

For organisations, the problem is harder. If every analyst is using the same tool to summarise the same reports, the firm’s collective intelligence is now bounded by the tool.

The hedge is structural: pay for dissent.

Hire some people who are bad at AI on purpose. Keep at least one team running on paper and instinct, even if their output is slower, because they will catch the thing the model missed. The model could not see it. France has been doing a version of this for wine since 1935. The AOC system makes cultivar diversity legally mandatory and the food is better for it.

For nations, this is an unsolved problem and probably the most important one nobody is working on. A country whose policy class all uses the same handful of American foundation models has outsourced a layer of its sovereign cognition without noticing.

The cabin in west Cork

In December 1846, a magistrate from Cork called Nicholas Cummins wrote a letter to the Duke of Wellington describing what he had seen in the cabins of west Cork. In one hovel he found six figures, skeletal, huddled under a horse-cloth, who he took for corpses until one of them moaned. The Times printed his letter on Christmas Eve. The British public read it and went back to their lives. Christmas is, after all, Christmas.

The thing about monocultures is that they look fine for a long time, until they do not, and the warning signs are structural rather than dramatic. Convergence in vocabulary. Convergence in framing. The vague feeling, reading three newspapers from three different countries, that you have read this article already.

Plant something weird this year.

Reading list

For people who want to keep planting weird things.

The Great Hunger by Cecil Woodham-Smith (1962). The definitive narrative history of the Irish famine. Worth reading for the British class assumptions that made certain outcomes invisible to decision-makers in real time and embarrassing in retrospect. The template recurs in every subsequent monoculture failure, including the ones currently in progress.

Recipe for Disaster: The Formula That Killed Wall Street by Felix Salmon, Wired, February 2009. The best short essay on cognitive monoculture in finance, written before the language existed for it. Twelve pages and you understand 2008 better than most people who lived through it.

The Botany of Desire by Michael Pollan (2001). The chapter on the potato is the relevant one. Connects agricultural monoculture to consumer preference, which is the demand-side mechanism people usually leave out when they blame the supply chain.

Banana: The Fate of the Fruit That Changed the World by Dan Koeppel (2008). The Cavendish and Gros Michel story in full, including the geopolitics. The phrase “banana republic” is not a metaphor, it was a business model. Funnier than a book about fruit logistics has any right to be.

The Curse of Recursion: Training on Generated Data Makes Models Forget by Shumailov et al, Nature, July 2024. The model collapse paper. Read the actual paper, not the headlines about the paper.

The Master and His Emissary by Iain McGilchrist (2009). A different kind of monoculture argument, this one about hemispheric dominance in Western thought. Overreaches occasionally. Worth the price of admission for the chapters that work.

The Library of Babel by Jorge Luis Borges (1941). Six pages. About what happens when you have access to all the information and no way to filter it. Predates LLMs by eighty years and remains the best thing written on the subject.

Canada’s GDP per capita fell two percent over the five years to 2024, the worst sustained decline since the Great Depression. The 2025 per-capita number was essentially flat after two consecutive years of decline. The OECD’s long-run outlook projects Canada will rank last among all thirty-eight member countries in real per-capita GDP growth through 2060.

The country’s response is to build its way out. Ottawa has announced a Major Projects Office to accelerate sanctioning of nationally significant infrastructure, an expanded mandate for the Canada Infrastructure Bank, and a regulatory framework reorganised under the One Canadian Economy initiative to move projects faster. The Spring Economic Update 2026 leads with a “second-fastest G7 growth” headline and an infrastructure pipeline measured in hundreds of billions of dollars.

Speed has not been the problem. The procurement that produced the failed PPPs of the last twenty-five years is still in place, along with the demand modelling that supported them, the capital matching that funnelled short-horizon private equity into long-ramp assets, and the missing phase-gate accountability that other mature project industries treat as basic. Adding capital to an unchanged structure produces the structure’s failures at scale.

The demographic foundation

Every merchant risk infrastructure project is underwritten on demand assumptions. The assumptions depend on population growth, working-age cohort size, household income trajectory, and willingness to pay user fees. The reality of the country those assumptions are supposed to track has shifted in ways most of the public-facing forecasting has not accounted for.

Canada’s total fertility rate sits at approximately 1.26, well below the 2.1 replacement level and at a historical low. The 65-and-over share of the population now exceeds the under-15 share and the gap is widening. The federal government has revised immigration targets sharply lower for 2025-27 from earlier trajectories, removing one of the principal supports under aggregate GDP growth and exposing the per-capita reality more directly. Productivity growth has been weak for over a decade. Household financial stress is elevated, with shelter costs consuming more than half of median household income in major urban markets and youth unemployment running above the post-2016 average.

These conditions matter for infrastructure because the demand cases that supported Canadian toll roads sanctioned in the 1990s do not match the country of 2026, and the demand cases being constructed in 2026 will not match the country of 2040. A forecast that assumes 2008-era population growth, 2008-era working-age employment, and 2008-era willingness to pay user fees, denominated in 2026 dollars, is a marketing document with footnotes.

The infrastructure community knows this. The procurement frameworks have not yet been required to.

What the Canadian PPP record actually shows

The 407 ETR is the standing example of how to give away the value of a public asset for a one-time fiscal payment. The Mike Harris government of Ontario privatised the toll road in 1999 for approximately $3.1 billion under a 99-year concession. The concession holders have generated multiples of that figure in dividends over the following twenty-five years, with the asset itself valued at substantially more than the original sale price. The Auditor General of Ontario subsequently found that the concession structure limited the province’s ability to regulate tolls, which now sit among the highest in North America. The structure transferred long-run upside to private holders for a payment that has been retrospectively understood as a significant discount to fair value. The mechanism that produced this was a procurement structure that prioritised closing the deal over pricing the asset.

The Champlain Bridge replacement was originally structured as a PPP and was progressively pulled back into direct federal procurement after concerns about lifecycle cost, tolling, and risk allocation. The completed bridge opened in 2019 under federal ownership with no toll. The decision to abandon the PPP structure mid-process was the correct one. It would have been a less costly correction to make before the procurement started rather than during it.

The Ottawa Confederation Line LRT was procured as a PPP through a consortium led by SNC-Lavalin. The system opened in 2019 and experienced sustained operational and reliability failures, including a major derailment in September 2021. The public inquiry led by Justice William Hourigan concluded that procurement and governance failures had compromised both delivery and operational integrity. The City of Ottawa has continued to absorb operating and capital costs the original PPP was structured to insulate it from. The asset is now in active service with a maintenance and reliability profile the original commercial case did not anticipate.

The Réseau express métropolitain in Montreal, structured by CDPQ Infra as a hybrid public-private project with Quebec’s pension manager as the principal sponsor, has been delivered substantially over its original budget and behind its original schedule. The hybrid model produced both Quebec pension exposure and a project subject to ongoing criticism over routing, station design, and integration with the existing transit network. The structure shielded the political principals from direct accountability for the overruns by placing the financial risk inside a pension institution whose oversight model is not designed for project delivery accountability.

The Confederation Bridge has performed competently for nearly thirty years on an availability-payment structure that did not depend on demand forecasts. The Anthony Henday Drive components of Edmonton’s ring road were delivered through PPPs that the Alberta auditors have generally treated as successful. The Vancouver Canada Line opened ahead of schedule in 2009 and has operated reliably since. BC Hydro’s John Hart Generating Station replacement was delivered on schedule and within budget.

PPPs in Canada have worked when the underlying asset is technically straightforward, the demand profile is predictable, the structure is availability-payment rather than demand-risk, and the public counterparty has retained the regulatory levers it needs to govern operation. PPPs have failed when demand risk has been transferred to capital that cannot bear it, when the asset is operationally complex, when the political incentive favoured signing over getting it right, or when the regulatory framework has not kept pace with the structure of the deal.

The acceleration agenda does not address any of the variables in the second list.

Why more money makes it worse

Capital chasing badly structured projects produces more failures, not fewer.

The Canada Infrastructure Bank, established in 2017 with a $35 billion mandate, has had a deployment record variously described as cautious, slow, or strategic depending on the political vantage of the commentator. Whatever the framing, the CIB’s evolution has demonstrated that a capital institution operating inside a flawed structural framework is forced to participate in the framework’s failures. The CIB cannot, on its own, fix the procurement practices of provincial transit agencies, the demand modelling practices of the major consultancies, or the capital matching practices of the sponsor community. It can refuse to participate, which it has occasionally done, or it can participate on its own terms and accept the resulting risk profile.

The federal Major Projects Office is intended to accelerate the sanctioning and delivery of nationally significant infrastructure. The acceleration model assumes the binding constraint is regulatory friction. The actual binding constraint in most failed Canadian infrastructure projects of the last two decades has been structural rather than regulatory. Demand has been allocated to capital that could not bear it, capital horizons have not matched asset ramp profiles, operational accountability has been displaced from sponsor to operator, and procurement has rewarded financial close over delivered performance. None of these are frictions acceleration would fix. Acceleration delivers more of each, faster.

A federal infrastructure push that delivers $200 billion of capital through the existing project pipeline, the existing procurement frameworks, the existing demand modelling capacity, and the existing capital matching practices, will produce a higher absolute volume of failed projects at the existing failure rate. The taxpayer will pay for the failures the same way the taxpayer has paid for Ottawa LRT, the equity holders of Cross City Tunnel in Sydney, the lenders of the Indiana Toll Road, and the public balance sheets that absorbed Carillion’s UK PFI portfolio after its collapse. The cost will be socialised because the assets will be essential. The political incentive will favour quiet restructurings over public failures.

On a successful PPP the upside accrues to private holders. The downside is absorbed by the public when the asset is essential, which it usually is. The risk transfer that justified the structure on day one turns out to have been imaginary by the time it matters. Adding capital makes the bill higher.

What structural reform looks like

The discipline that would fix this is regulatory, unglamorous, and unpopular with everyone whose fees crystallise at financial close.

Consider what a merchant risk transport project would look like under gates that worked. Take a tolled bridge, an inland port, or a regional rail link. The procurement sequence might run as follows.

The development phase would close on a forecast that had reconciled, in writing, to Statistics Canada’s regional population projection, the Parliamentary Budget Officer’s productivity baseline, and the relevant provincial economic outlook. Variances between the consultancy forecast and the agreed federal-provincial baseline would be a procurement disclosure item. A forecast that depended on demand the demographic numbers could not support would not progress to underwriting.

The underwriting phase would close on a capital structure where lead equity matched the asset’s ramp profile. A project with a seven-to-ten-year ramp would not take short-horizon private equity as its lead equity, and the CIB and Major Projects Office sanction criteria would say so explicitly. Where demand risk could not be absorbed by the equity stack, the public contingent liability would be priced and disclosed in the procurement documents on day one, capped, with the workout playbook agreed before the asset opened.

The construction phase would close with an operations director named, in the room, and accountable for the case being handed over. Critical development personnel would carry compensation that did not vest until the asset reached steady state. The receiving team would have a contestation right at the gate, with real consequences for the handing-off team if the deliverable did not meet the acceptance criteria.

The operations phase would open with twelve months of customer pre-validation behind it: letters of intent from anchor users with volumes and indicative pricing, a documented pipeline of prospective customers, and independent industry surveys with sample sizes large enough to mean anything. The first board meeting of the operating company would receive a paper comparing actual ramp against the conservative case underwritten for debt service, with the marketing case shown separately. A covenant holiday would sit over the ramp years.

None of this is exotic. The aerospace primes have done versions of it for decades, the oil and gas majors run major projects under Front-End Loading frameworks, and the better regulated utilities use stage gates as standard. Canadian transport procurement does not.

The reforms do not produce ribbon-cuttings, and ministers who require them rather than sanctioning projects on quarterly timelines tend not to last in office. Each one, if adopted, would prevent more value destruction than the entire Major Projects Office is likely to add through acceleration.

The point about who pays

Canada has access to the capital, the political mandate to spend it, and the institutional capacity to deploy it. What is missing, in the part of the system that determines whether the spend produces assets that operate as promised, is the structural discipline.

Adding capital to the system without adding structure produces what the system has been producing. A pipeline of essential assets the public will eventually have to absorb, at the same time as the per-capita income that should have supported them is declining, and at the same time as the demographic base that should have driven their demand is shrinking.

The frame, of aggregate growth and infrastructure volume rather than per-capita performance and structural reform, is the same metric-capture move that produces the failed merchant risk projects this argument started with. Choose the number that supports the political case, defer the structural reckoning, and push the consequence forward to people who will not be in the room when it arrives.

The country deserves a debate about structure rather than a press release about capital.

Reading list

Ludovic Phalippou, “An Inconvenient Fact: Private Equity Returns and the Billionaire Factory” (Oxford Saïd Business School, 2020). The cleanest demolition of the PE outperformance claim. Fees as the actual product.

Eileen Appelbaum and Rosemary Batt, Private Equity at Work: When Wall Street Manages Main Street (Russell Sage Foundation, 2014). The standard sector-by-sector account of what happens to companies inside a PE hold period. Strong on employment and supplier effects.

Brendan Ballou, Plunder: Private Equity’s Plan to Pillage America (PublicAffairs, 2023). Written by a former DOJ counsel. Recent, accessible, anchored on the consumer harm transmission.

Eduardo Engel, Ronald Fischer and Alexander Galetovic, The Economics of Public-Private Partnerships: A Basic Guide (Cambridge University Press, 2014). The standard academic treatment of PPP economics. Strongest on risk allocation theory.

Aidan Vining and Anthony Boardman, various articles on Canadian PPP performance (Journal of Comparative Policy Analysis, Canadian Public Administration, 2008 onward). The Canadian empirical literature on what worked and what didn’t, by the two scholars who have spent the most time on it.

Justice William Hourigan, Report of the Ottawa Light Rail Transit Public Inquiry (Government of Ontario, 2022). The case study. Demonstrates concretely how procurement structure compromises operational delivery.

Auditor General of Ontario, reports on Highway 407 ETR (2003, 2009, and subsequent). The original sin of Canadian PPP privatisation, documented contemporaneously by the relevant officer.

National Audit Office (UK), Investigation into the government’s handling of the collapse of Carillion (HC 1002, 2018). The clearest official account of what happens when essential infrastructure delivery passes through a private balance sheet that fails.

Lord Justice Sheen, MV Herald of Free Enterprise: Report of Court No. 8074 (Department of Transport, UK, 1987). The canonical project-structure-as-safety-failure document. Required reading for anyone thinking seriously about commercial pressure in regulated transport.

James Reason, Managing the Risks of Organisational Accidents (Ashgate, 1997). The framework most modern safety thinking still relies on. The Swiss cheese model is Reason’s.

Charles Lammam et al., Standard of Living in Canada: Trends and Drivers (Fraser Institute, 2025). The per-capita GDP work that has set the public debate frame in Canada.

OECD, Economic Survey of Canada 2025. The international comparative perspective on Canada’s productivity and per-capita performance. The long-run projection that has Canada last in the OECD through 2060 comes from the same workstream.ream.

Imagine your GP at the pub on a Friday, three pints in, having a quiet moan about the modern patient. She would tell you that half her caseload is now people in their fifties who feel mostly fine but have a long list of small things wrong at the same time. A bit of high blood pressure. A dodgy gut. Knees gone. Sleep wrecked. Old injuries that never quite healed. None of it killing them. None of it fixed by treating any single problem on its own.

Doctors call this multimorbidity. It barely existed a century ago and now accounts for most of what GPs actually see. As Voltaire put it, the art of medicine consists in amusing the patient while nature cures the disease. Three hundred years on we have better drugs and busier patients, but the principle holds. Bodies have opinions about being fixed. So do planets.

The patient I want to talk about today is the planet, and the news from the clinic is mixed.

You are not actually you

Bad news first. You are not what you think you are. You are you plus about thirty-nine trillion gut bacteria, plus the mitochondria in every cell that still carry their own DNA from before they joined the team two billion years ago, plus a population of mites living on your face right now (sorry), plus a virome that outnumbers your human cells by ten to one. Biologists call the composite creature a holobiont. The word never caught on at parties.

The planet works the same way. Biosphere, plus eight billion people, plus livestock that outweighs all wild mammals combined by ten to one, plus infrastructure, plus finance, plus an information layer that did not exist as a global system before about 1995 and now runs at AI speeds less than five years old. Three layers, three different clocks. Most of the trouble lives in the gap between them.

Bad arteries

The classic check at any new patient appointment is the cardiovascular workup. Arteries narrow, blood backs up, tissue downstream dies. The horror of a heart attack is that one tiny blockage in one tiny vessel can take you out.

The planet has bad arteries too. About a fifth of the world’s seaborne oil squeezes through the Strait of Hormuz between Iran and the Arabian Peninsula. Roughly the same share of Chinese trade goes through Malacca, between Malaysia and Indonesia. Add Suez, Panama, the Bab-el-Mandeb. Each is a narrow passage where one event can stop a system thousands of times its size.

I remember a moment from air operations over Iraq. I was watching the IFF on the air defence radar, the screen that should be busy with returns: civilian traffic, allied movements, our own aircraft. Overnight it went clean. The Combined Air Operations Centre had sanitised the airspace ahead of strikes. No aircraft, no warning. The thing you usually saw was simply not there. That is what a system stopping looks like from inside, when you are the one watching the screen.

When the Houthis started shooting at ships in the Red Sea, global shipping rerouted around the Cape of Good Hope. The body does the same thing around a blocked artery, paying metabolic cost for the new route. The cost in this case showed up as higher freight rates, higher insurance premiums, more carbon per container, and a Suez Canal Authority that lost most of its revenue overnight. Egypt is now feeling that as fiscal pain. Domino set up.

The diabetes of finance

Stress hormones are useful for a sprint. Useless if you cannot switch them off. Hold cortisol on for years and the body stops listening to it. Same with insulin: that is how type 2 diabetes works. Your pancreas is shouting and the receptors are wearing earplugs.

Central banks are the planet’s stress hormone system. They do not build anything. They signal. From 2008 to 2022 the major ones flooded the global system with signals: rates near zero for a decade, balance sheets the size of small economies, quantitative easing (a polite name for buying bonds with money that did not previously exist).

Around 2009 a serving central banker explained it to me over a coffee. The shadow banking system was warehousing bad debt to put off the reckoning. Central banks were suppressing rates artificially to keep the patient breathing. He was not complaining; he was describing the trade. I came out of that conversation knowing the system was already broken in a structural sense, and the people running it knew.

The 2022 rate hikes did less than the old models predicted. The patient had developed receptor resistance. Doses up, effects down, eventually nothing works. Monetary diabetes. And the doctors managing it are increasingly being told what to prescribe by the family. About one in ten central banks now faces direct political pressure in any given year, almost always to ease policy regardless of what the diagnosis says. The legal independence usually still stands. The independence in practice is bleeding out.

William Osler, the godfather of modern medicine, said medicine is a science of uncertainty and an art of probability. Spare a thought for the doctor in this clinic.

Antibiotics, and Canada’s empty cupboard

Your gut hosts somewhere between five hundred and a thousand bacterial species. They digest fibre your cells cannot, make vitamins your cells cannot, and quietly run a lot of your immune system. A heavy course of antibiotics can wipe out gut diversity for months. In that window you become vulnerable to a bug called C. difficile that thrives when its competitors are gone. Severe cases are treated with a faecal transplant. Yes, it is what it sounds like. Try not to dwell.

Two centuries of industrial modernisation worked like a civilisational antibiotic course. Out went local craft, regional cuisines, minority languages, traditional agriculture, indigenous governance, mutual aid societies. Many were oppressive and not much missed. The casualty was diversity itself, because diversity is what absorbs shocks when the dominant structure breaks.

Consider what Canada looked like during the pandemic. The country was importing finished refined product into the east while shipping crude west, with no operational pipeline running coast to coast. Energy East had been cancelled. Terminal Norcan, where I ran the operating company at the time, sat as the marginal barrel into Quebec and Atlantic Canada, dependent on volatile import flows in a year when global shipping itself was breaking. Imagine the alternative. Pipeline integration east to west. A refinery system that could route around problems. Norcan as one piece of a redundant national network rather than the relief valve. We would have spent the pandemic less exposed to global volatility, with thicker buffers and more pricing optionality. Instead we got the opposite. Single corridors, single suppliers, single points of failure.

That is dysbiosis at the national scale. Four crops feed most of humanity. A handful of languages dominate global commerce. About ten companies handle most of the information layer. A drought in one wheat region now moves bread prices across forty countries. The system is efficient and brittle, and the brittleness only shows up when something hits it.

The immune system, in patches

Your immune system is not one thing. It is layers. Skin and mucous membranes as physical barrier, innate responses that punch first and ask questions never, adaptive responses with memory for things they have seen before. Failure rarely takes everything down at once. Usually one compartment goes and the rest carry on. People with autoimmune flares often look completely fine for years between episodes. People with bad allergies have entirely functional immune systems that have simply decided pollen is a war crime.

The global system has an institutional immune system: UN, WHO, WTO, the Bretton Woods crowd, regional alliances, treaty networks, central bank coordination, scientific academies, national regulators, municipal public health.

Some compartments still work brilliantly. Vaccine development answered Covid in about a year, the fastest such response in history. Global trade keeps functioning. Child mortality keeps falling. Extreme poverty by World Bank measures is still declining, slower than the 2000s but still going the right way.

Other compartments are gone. Climate coordination, thirty-three years on, has produced emissions stabilisation only because China’s economy slowed. Nuclear non-proliferation below the P5 has quietly collapsed, with North Korea, India, Pakistan, Israel and now plausibly Iran sitting outside the regime. Pandemic preparedness is worse than it was in 2019 because attention moved on. AI governance does not really exist yet. Migration governance is in open breakdown across several regions.

The patient has selective immune failure. Some compartments responding well, others giving up entirely. That description is more useful than “institutional decline” because it tells you which bits to stop relying on.

Adolescent

I said earlier the patient was fifty. I was being kind.

Industrial civilisation is about two hundred and fifty years old, three human lifespans, embryonic in evolutionary terms and barely adolescent in cultural ones. The symptoms fit. Identity formation weaponised into nationalism. Risk-taking without consequence modelling, visible in everything from unchecked carbon release to unbridled AI deployment to expanding nuclear arsenals. Narcissism amplified by mirror surfaces, otherwise known as social media. The bubble-and-crash cycle of modern finance as adolescent impulse control under hormonal surge.

The human prefrontal cortex, the bit responsible for long-term planning and impulse suppression, does not finish wiring up until around age twenty-five. The civilisational equivalent is the institutions that can plan across decades: independent central banks, independent courts, scientific academies, planning commissions. Current political populism is going at these institutions with a screwdriver across most democracies.

We are unwiring the adolescent brain during the window when it most needs to wire up. Most adolescents survive their twenties. Many spend the next forty years paying for choices made while the wiring was still under construction.

Prognosis

The patient lives. Multimorbid patients persist for decades on good management. They do not run marathons. The post-1945 order will not perform the way it did at thirty, and expecting it to is the kind of denial GPs encounter every day in their fifty-somethings.

The acute risk worth watching is the mismatch between the system’s clock speeds. Capital markets move in milliseconds. News in hours. Politics in years. Climate in decades. When the fast layers structurally drown out the slow ones, the slow signals cannot get through. Most multimorbid patients die of the thing that moved fastest while everyone was watching the thing that moved loudest.

The first rule of medicine, attributed to Hippocrates and possibly invented later by people quoting him, is to do no harm. Worth remembering when the next genius reaches for a cure that destabilises everything else.

What you do with this on Monday

A clinician’s questions, when you read next week’s news. Is this symptom or cause? Acute or chronic? One system failing or several interacting? What would a sensible doctor track next, and what would she discount as noise?

In multimorbidity, what gets watched matters more than what gets prescribed.

Have a good weekend.

The strait is being repriced. That is the part markets are missing.

Oil dropped ten percent on Friday. The S&P hit a new high. Heating oil futures fell thirteen. A friend texted me that peace was breaking out.

What happened on Friday was not peace. Iran’s Foreign Minister declared the Strait of Hormuz “completely open for commercial vessels” for the remainder of the current ceasefire, via a coordinated route already announced by Iran’s Ports and Maritime Organisation. On the same day, Iran’s Deputy Foreign Minister told a different press pool that Tehran rejected any temporary ceasefire and wanted a comprehensive regional settlement. Two members of the same cabinet, different positions, same afternoon. That is a regime hedging publicly so it can walk back either statement depending on what the other side does next.

Trump posted his thanks to Tehran in all caps and kept the US blockade of Iranian ports in place. Markets treated the combination as a peace signal and sold off war-risk premium. That is the error this piece is about.

This piece is not about whether a US-Iran deal happens in the next 60 days. It might. I put the probability at around 40% for a narrow transactional deal, meaning ambiguous Hormuz access, partial asset release, a face-saving nuclear formula, and Lebanon parked on a separate track. Both sides need the reprieve. Both principals have incentives to sign something.

The assumption that just died

For thirty-five years, from the end of the Iran-Iraq tanker war in 1988 until 28 February 2026, there was a working assumption underneath every Gulf-origin energy contract, every refinery offtake agreement, every insurance premium, every SPR release calculation, every LNG long-haul route, every petrochemical plant siting decision in East Asia. The assumption was that the Strait of Hormuz was a transit commons, guaranteed by the US Fifth Fleet, unpriced in the base case of most corporate and sovereign models, with tail risk handled by war-risk insurance riders that occasionally spiked and then returned to baseline.

I want to be careful here. I am not suggesting the Strait will stay closed, or that the current crisis will not be resolved, or that Gulf oil will not flow again at scale. It probably will. What I am saying is that the category of assumption has changed. Before February, Hormuz access was a near-constant in everyone’s models. After February, it is a variable. Variables require hedging, optionality, inventory, alternative routing, and capital allocated against the possibility of disruption. The shift from constant to variable is a balance-sheet event even if the variable’s realised value returns to something close to the old constant.

This distinction is what markets are missing. A ten-percent drop in crude on Friday prices a return to normal. It does not price the cost of having learned that normal is optional.

What the 1988 ceasefire actually taught us

When Iran and Iraq agreed to ceasefire in August 1988 after eight years of tanker attacks, mine-laying, and the US Navy’s Operation Praying Mantis, markets treated the resolution as a full reset. Insurance premia normalised within about 18 months. SPR discussions faded. LNG long-haul routing returned to pre-war assumptions. By 1990, most of the lessons of the tanker war had been forgotten outside naval staff colleges and a narrow circle of war-risk underwriters at Lloyd’s.

Then 1990 happened. Then 2003. Then the 2011-12 Strait closure threats over nuclear sanctions. Then 2019 and the Fujairah tanker attacks. Then the Soleimani killing. Each episode was treated, in the immediate aftermath, as a discrete event that resolved rather than a data point in a series. Each produced a price spike followed by a return to baseline. Each was followed by someone, somewhere, writing a piece saying this time it was different, and the chokepoint risk was now permanently priced in.

It never was. Because between episodes, the US Fifth Fleet went back to being the silent guarantor, and the market returned to treating Hormuz access as a constant.

What has changed in 2026 is not that there is a crisis. Crises are cyclical. What has changed is the mechanism by which the crisis has been produced and the mechanism by which it is being resolved. Both break the 1988-2026 model in ways that do not reset even when the ships start sailing.

The five mechanisms of permanent repricing

The repricing operates through five distinct mechanisms, each of which persists through any reasonable near-term political resolution.

One. War-risk insurance volatility itself is now priced.

Before February, war-risk premia for Hormuz transit sat around 0.125% of hull value. In the three weeks before the 28 February strikes, as Iran signalled, they moved to between 0.2% and 0.4%, an increase of roughly a quarter-million dollars per very large tanker transit. Post-strikes, coverage became unavailable for most tonnage for extended periods. Lloyd’s syndicates withdrew. P&I clubs imposed restrictions.

The part that does not reset is this. Insurance is priced off the distribution of expected losses, not the central tendency. A market that has now observed one full closure event with mining, 21+ attacks on merchant ships, and a US blockade cannot write the same distribution curve it was writing in January. The realised variance is too high. Reinsurance treaties will be renegotiated at the next renewal cycle. Annual premia for Gulf-origin vessel deployment will carry a structural increment that persists even when transit premia return to near-baseline.

I spoke with a reinsurance contact at Munich Re equivalent tier a few weeks ago and the working assumption internally was a 15-25 bps permanent increment on Gulf-origin marine exposure going into 2027 renewals, regardless of political outcomes. That is a small number per hull, but multiplied across the 100-140 pre-war daily transits and across full-year volumes, it is a multi-billion dollar annual transfer from energy economies to reinsurance markets. It does not go away when the mines are cleared.

Two. Routing optionality is now a required capital expenditure.

Before February, the bypass infrastructure around Hormuz (Oman’s Duqm, Salalah, and Sohar ports on the Arabian Sea, the UAE’s Fujairah, the Saudi East-West pipeline to Yanbu, the Abu Dhabi Crude Oil Pipeline to Fujairah) was treated as legacy infrastructure with modest ongoing utilisation. The underlying assumption was that bypass capacity was nice-to-have rather than need-to-have because Hormuz was reliable.

That assumption should also be reviewed. Every Gulf energy producer, every Asian importer, and every oil major with Gulf exposure is now running capital planning scenarios that assume Hormuz disruption risk has moved from a tail event to a base-case variable. Duqm’s expansion plans, which were chugging along at sensible pace, will now accelerate. Fujairah’s storage capacity build-out is already being bid forward. The ADCOP pipeline, the East-West pipeline, and Iraqi outlet options through Turkey (when that political relationship permits) are all now strategic assets rather than redundant ones.

The capital cost of this repricing is enormous and the beneficiaries are predictable. Any port, pipeline, or storage asset outside the Strait but within the Gulf region sees a permanent structural uplift in valuation. Any asset inside the Strait faces a structural discount. This is already visible in infrastructure-fund bid-ask dynamics for regional assets, though it will take another year for the repricing to fully capitalise.

Three. Strategic petroleum reserves are being restructured, quietly.

The US, China, India, and Japan all entered 2026 with SPR policies calibrated to the pre-2026 Hormuz assumption. Each of them has, in the six weeks since the crisis began, begun quiet policy reviews of both stockpile size and geographic distribution. China, which gets roughly a third of its oil through Hormuz and had about a billion barrels in reserve at the crisis start, is the most exposed and is already signalling expanded stockpile targets. India has begun discussions on doubling its strategic reserve. Japan has initiated a review. The IEA itself said this week that resuming flows through Hormuz is the single most important variable in easing pressure on the global economy, which is also an admission that the pre-crisis SPR framework was inadequate.

SPR expansion at sovereign scale is not a one-year process. It is a 3-7 year commitment involving siting decisions, storage construction, filling logistics, and diplomatic coordination. Once begun, it does not unwind on a ceasefire announcement. The decisions being made in ministerial offices this quarter will shape oil demand dynamics into the 2030s. Expanded SPR buying is also, mechanically, a durable bid under crude prices that offsets some of the demand destruction from elevated prices. Markets pricing a return to $75 Brent are not modelling this bid.

Four. The US Navy is no longer an unconditional guarantor.

For forty years, the working assumption was that the Fifth Fleet was a commons-provider. Freedom of navigation in the Strait was a public good the US provided because it was in US interest and because it was also in the interest of every Gulf producer, every Asian consumer, every European LNG importer, and every major shipping flag.

The US blockade that took effect on 13 April, even in its selective form (blocking only vessels entering or leaving Iranian ports, permitting transit to non-Iranian ports), breaks that model. The Fifth Fleet has now been used as a coercive economic instrument, not a commons-provider. Every capital that imports Gulf energy now has to model the US Navy as a variable whose deployment depends on US political interests rather than as a constant.

For allies, this is uncomfortable but manageable. For non-allies and partial-allies, it is a structural change. Beijing in particular now has an incentive to accelerate its own blue-water capability, its Gulf port access through Gwadar and Djibouti, and its bilateral security relationships with Gulf states. The Gulf states themselves now have to hedge their security posture in ways that did not previously make sense, which is why you are seeing accelerated GCC-China defence dialogue and why the Saudi-Iran rapprochement of 2023 looks in retrospect like prescient positioning rather than opportunism.

This is not reversible. You cannot unuse a capability once it has been used. The norm that governed the Gulf for 35 years has been broken, and nobody will forget that it can be broken.

Five. The toll route precedent is now out in the world.

Iran’s Larak Island toll route, where vessels pay the IRGC directly for safe passage, handled 50 of 72 Strait transits between 6 April and 13 April. Those are Lloyd’s List figures and they are probably conservative.

What Iran has demonstrated is that a middle-power state with geographic advantage, sub-state paramilitary assets (in this case the IRGC Navy’s small-boat fleet), and mining capability can monetise chokepoint access through direct toll extraction without triggering a full-scale war from the affected parties. This is a meaningful innovation in coercive geoeconomics. It is the maritime equivalent of what the Houthis have been doing in the Red Sea, but executed by a state actor with more capability and more leverage.

Every chokepoint state is watching. Turkey controls the Bosphorus and Dardanelles. Egypt controls Suez. Malaysia, Indonesia, and Singapore control Malacca. Yemen’s parties, including those aligned with Iran, control Bab-el-Mandeb. Even Panama, with a different geography and political context, has the canal. The precedent that a state can extract sovereign rent from international navigation, coercively, and survive the consequences, changes the incentive structure in all of these places. Not immediately. But the decade ahead of us will look different because of this than the decade behind us did.

Bloomberg noted this week that the Hormuz crisis has revived anxiety over Malacca specifically, and that is the correct instinct. The analogy will be extended to every bottleneck that matters.

The adjacent industries few are pricing

Three adjacent industries have been hit harder than crude and will take longer to recover, and each of them has consequences that propagate into food prices, manufacturing costs, and geopolitics in ways that will outlast any ceasefire.

Fertilizer

The Persian Gulf accounts for roughly 46% of global seaborne urea transit and around 30% of ammonia transit. Qatar alone supplies a significant share of global LNG, which is the feedstock for nitrogen fertilizer production in India, Bangladesh, Pakistan, and Egypt. When Qatari LNG stops flowing, fertilizer plants across South Asia stop running, even though they are thousands of miles from the Gulf.

FOB Middle East granular urea went from $436-494/ton before the war to $604-710/ton by 19 March. Southeast Asia granular urea moved from $490-498 to $750/ton over the same period. Urea at the Port of New Orleans is up 25%+. Sulphuric acid prices are rising because a third of the world’s sulphur is a byproduct of Gulf oil and gas refining.

The asymmetry with oil is what matters. Oil can be rerouted around Hormuz through Saudi pipelines to Yanbu and via Fujairah. Ammonia cannot. There is no ammonia pipeline bypassing the strait. There are no strategic fertilizer reserves in any G7 country. There are no military escort programmes for fertilizer cargoes. A ship captain braving drone attacks prefers to carry crude, because the per-voyage economics favour oil over bulk nitrogen.

And the timing is worse than the oil disruption. The Northern Hemisphere planting season is now. Farmers who cannot secure urea in April and May do not make up the shortfall in June. They plant less, switch crops, or leave fields fallow. The consequence shows up in supermarket prices in six to twelve months, not weeks. NDSU’s agricultural economics group is forecasting elevated fertilizer prices through 2027 even under a hypothetical immediate full reopening, because the planting cycle has already been disrupted and the global inventory position cannot be rebuilt in a single season.

The World Food Programme estimates that acute food insecurity could rise by 45 million people if the Iran war runs through mid-2026 with oil above $100. That number is sensitive to how long fertilizer remains disrupted, not how long oil does.

The Russian beneficiary nobody wants to name

Russia accounts for 23% of global ammonia exports, 14% of urea exports, and with Belarus 40% of potash exports. Russian export infrastructure is entirely independent of Hormuz. The Kremlin does not need a ceasefire, a military escort, or a diplomatic breakthrough to ramp deliveries. It needs orders, and it is getting them.

Importers in Nigeria and Ghana are already pre-purchasing Russian fertilizer for Q3 2026. Once these trade relationships are established, they do not unwind on a ceasefire. They calcify into dependency. Moscow has played this hand before, and it knows how to convert commodity leverage into political leverage. The 2026 fertilizer crisis is doing for Russia in African and South Asian fertilizer markets what the 2022 gas crisis did for it in selective European energy relationships: creating durable reliance in countries whose neutrality matters to Western strategic interests.

This is a structural geopolitical transfer that neither a US-Iran deal nor a Hormuz reopening will reverse. By the time trade flows could revert, the commercial relationships, credit lines, and shipping contracts will have been re-papered around Russian supply. That is a Kremlin gain the West cannot un-price by declaring peace.

Petrochemicals and LNG to Europe

The Middle East accounts for roughly a quarter of global petrochemical feedstock exports, concentrated in Saudi Arabia, the UAE, Qatar, and Iran itself. Naphtha, ethylene, propylene, and methylene derivatives flow to plastics and specialty chemicals manufacturing across Asia and Europe. Qatar’s LNG shutdown alone removes roughly 20% of European LNG imports at a moment when European gas storage is entering the summer refill season underweight.

The consequences propagate into European manufacturing cost structures in a way that the post-2022 gas crisis already previewed. German chemicals exposure is the obvious case; BASF, Covestro, Evonik, and Lanxess all carry Hormuz-proximate feedstock sensitivity through the Qatari LNG channel. But the less-obvious case is UK, French, and Italian plastics and packaging manufacturers, who do not buy from the Gulf directly but whose input costs move with global polymer pricing.

European industrial competitiveness versus US and Chinese rivals takes another structural step down from a Hormuz disruption of this magnitude, and the step down is sticky for the same reason the German chemicals sector has not recovered from 2022. Capacity gets written down. Production gets relocated. Jobs move.

Shipping rates beyond tankers

One industry-adjacent effect worth flagging because it hits the shipping-fund allocators reading this. War-risk insurance withdrawal and rerouting costs are now pricing into container, dry bulk, and LNG shipping rates, not just tankers. Red Sea rerouting around the Cape of Good Hope is adding 10-14 days to Asia-Europe container voyages. Add a Hormuz disruption on top, and Asian exporters lose two major routes simultaneously.

This is the cost structure that drives H2 2026 inflation prints in Europe and which the ECB will be trying to read through for monetary policy purposes. The structural re-pricing in marine logistics is not reversing at the pace markets are assuming.

What the traders got right and wrong on Friday

The traders got one thing right. A face-saving pause serves both Washington and Tehran right now, and the probability of a narrow deal is non-trivial. Selling off some of the war premium was defensible.

What they got wrong was the magnitude of the rally, and the magnitude is measurable.

The mispricing in oil

The base-rate argument here requires a comparable event, and there is not one.

The easy move is to point at 1988, 2007, 2012, 2019, and 2020 and derive a 37% historical probability of durable resolution from ceasefire announcements in comparable Gulf episodes.

None of those episodes involved: the assassination of Iran’s supreme leader, direct US and Israeli strikes on Iranian territory, a formal US naval blockade, actual mining of the strait, 21+ confirmed attacks on merchant shipping, an IRGC toll regime in commercial operation, the Fifth Fleet deployed as a coercive economic instrument rather than a commons-provider, and a regime in the middle of a post-Khamenei succession crisis. The closest historical analogue is 1987-88, and the current situation is arguably more combustible than that one because Iran now has drones, precision anti-ship missiles, and an explicit track record of sub-state coercion that has succeeded in extracting sovereign rent.

The honest base rate is therefore materially lower than 37%, and a probability estimate has to be built from first principles rather than historical analogy.

Durable (90+ day) resolution requires all of the following to hold simultaneously. Iran must relinquish the Larak Island toll revenue, which is currently generating $100m+ per week into IRGC hard-currency accounts, the most lucrative revenue stream the regime has had in years. Iran must also abandon its strongest remaining deterrent card after the loss of its nuclear programme, proxy network, and leadership. Israel must halt operations in Lebanon, which is Iran’s stated red line and which Netanyahu has explicitly refused to cross. The US must offer credible post-JCPOA commitment mechanisms to a regime that watched Washington walk away from the last deal. A succession inside Tehran must produce a winner who can sell this concession without being overthrown by the hardliners in the IRGC.

The joint probability of those conditions, if you assume 60% for each independently (which is generous), is approximately 8%. If you assume 70% for each, which is very generous, the joint probability is approximately 17%. A defensible range is 10-20% for genuinely durable resolution, meaning 80-90% probability of some form of breakdown within the ceasefire window or shortly after.

Fair-value Brent under each scenario: $80-85 (durable resolution, higher than before because the structural floor from the five mechanisms holds), $115-125 (collapse/escalation with full re-closure), $92-95 (muddled middle, which is actually the modal outcome at roughly 50% probability).

Probability-weighted fair value at a 15/50/35 split (durable / muddled / collapse): (0.15 × $82.50) + (0.50 × $93.50) + (0.35 × $120) = $12.38 + $46.75 + $42 = approximately $101/bbl

Observed Brent after the announcement: approximately $87/bbl.

Mispricing: approximately $14/bbl, or 14% too low. Markets are pricing 55-60% probability of durable resolution when the defensible probability is 10-20%. That is a categorical gap between what the tape says and what the underlying situation supports.

The mispricing in equities

S&P hit a new all-time high on the announcement. The mispricing here is harder to isolate because equities aggregate many factors. But the Hormuz channel alone, through oil cost sensitivity (roughly 0.3% S&P earnings impact per $10/bbl mispriced) and the VIX collapse from ~22 to ~17.50 (a ~20% volatility reset that prices away tail risk which has not actually been resolved), suggests equities are 4-6% above fair value on this news alone. This is meaningful when compounded with pre-existing valuation concerns about AI capex and concentrated mega-cap exposure.

The asymmetric payoff

At a 15/50/35 probability split on durable/muddled/collapse, the payoff structure from current levels is:

• Long oil: expected return approximately +16% from $87 toward $101 fair value, with tail upside to $120+ under collapse

• Long equities at all-time highs: expected return approximately -4 to -6% with tail downside of 10-15% under collapse

The asymmetry is stark. Markets are positioned the opposite way. When cross-asset markets move in internally consistent directions on thin information, that is the signature of positioning-driven flow (CTAs, vol-targeting, risk-parity, retail FOMO) rather than fundamentals. The Friday move priced the optimistic scenario at something close to certainty, which is never how real uncertainty resolves.

Follow the money

Who is on the other side of this trade? Worth asking, because the answer tells you how the Friday rally actually got manufactured and who pays for it.

The institutional book

The professionals did not miss this. The professionals positioned for it before it happened, and most of them started in late February.

One commodity fund manager with three decades of experience across the top macro and multi-strat shops launched a commodity-focused vehicle in late 2025 that had been largely absent from oil as of February 2026, on the view that the market was oversupplied and the price reflected it. Within days of the Iran conflict starting, that fund established long positions in the mid-part of the Brent and WTI curves (late 2026 into mid-2027). When the Hormuz closure came, those positions were increased. The fund returned 17.1% in Q4 2025 and was up 13.4% year-to-date through March 2026. In a recent interview, the manager articulated his central argument in one sentence: “Once the Straits reopen, it will take 50 to 100 days to get back to 80% of previous oil flow. The shores will need to be secured. There will be convoys. That tail adds further to the equilibrium price.”

That is exactly the mispricing calculation I ran above, done by someone managing real risk with real capital, not writing a Substack.

The broader numbers confirm the pattern. Hedgeweek’s Q1 2026 allocator survey shows the proportion of allocators reporting commodity hedge fund exposure at 33%, but forward intent (where they plan to allocate next) at 59%. That is one of the sharpest single-cycle moves in the dataset. Energy sector returns were +34% in Q1 2026 alone. Energy hedge fund AUM is growing faster than any other single sub-strategy. Large commodity trading houses with flex capacity and storage optionality do not publish their positioning, but the historical pattern in shocks of this magnitude is clear: physical merchants make supernormal profits during disruptions, usually several multiples of normal-year earnings. The 2026 numbers will print in twelve months and they will follow the same pattern.

The retail book

On the other side of those institutional longs is a rally-chasing retail bid that got triggered Friday morning when Iran’s announcement hit the tape.

CTAs, which are systematic trend-following strategies, are estimated to be buying approximately $40bn of S&P 500 exposure this month, mechanically, because price thresholds crossed. That is $40bn of flow that is not based on a view. It is based on a rule. When the rule reverses, that $40bn becomes a seller at whatever the market is offering.

Retail equity flows into broad-market ETFs and mega-cap tech hit multi-month highs in the 48 hours after the Iran announcement. Robinhood activity spiked. “Peace dividend” was trending on financial Twitter by Friday afternoon. The narrative spread faster than the facts, which is a signature of a positioning-driven move rather than a fundamentals-driven one.

The pattern is symmetric on the oil side. Retail short-oil ETF inflows (XLE puts, USO puts, oil short ETFs) hit fresh highs on Friday. The crowd is selling oil at $87 that the professionals bought at $95-100 and intend to sell back at $100-110 in the collapse scenario or $82-87 in the muddled scenario (flat to their cost basis, plus carry). The professionals’ downside is bounded. The retail upside is negligible.

The flows into Russia

The quietest trade in the whole complex is the one happening in Russian commodity markets. Russian urea and ammonia orderbooks are filling fast with Nigerian, Ghanaian, Brazilian, and South Asian buyers locking in Q3 and Q4 2026 deliveries at prices well above pre-war baseline. Russian oil at the Urals-Brent discount has narrowed materially as Asian buyers have revived relationships that were dormant through the sanctions regime.

This is now a commercial reality visible in freight bookings at Novorossiysk and Taman, in letters of credit being written in Mumbai and Lagos, and in the tonnage being chartered for the Russia-to-Global-South corridor. Russian state and para-state entities are booking the gains. Western allocators who are long Russian sovereign exposure through the small number of legal channels available (Indian local-currency bonds that recycle Russian trade flows, specific Chinese equity holdings, commodity derivatives on Russian-produced metals) are participating. Everyone else is watching capacity they used to own go permanently to Moscow.

The scorecard

If this piece’s thesis is roughly right, and Brent gravitates to $95-105 over the next 60-90 days while equities give back 3-5% of the post-Friday rally, the ledger of winners and losers will read approximately:

Winners. Commodity hedge funds positioned long oil in early March (Fulcrum, the multi-strat macro funds, specialist energy funds, trading houses). Physical merchants with Gulf flex capacity. Russian and Russian-proxy commodity exporters. Arabian Sea port operators and bypass infrastructure owners. Reinsurers on 2027 renewal cycles. SPR-refilling sovereigns buying at discounts to the eventual equilibrium price. Selected large-cap integrated majors (XOM, CVX, Shell, BP) that print cash at $95+ Brent. Fertilizer producers outside Hormuz exposure.

Losers. Retail investors who bought the peace-dividend rally at all-time highs. Emerging-market importers of nitrogen fertilizer, especially in Africa and South Asia. European industrial chemicals manufacturers carrying Qatari LNG feedstock exposure. Airlines and refiners short jet fuel into a structurally elevated crack spread environment. Passenger transport operators with unhedged fuel exposure. Treasury ministries in the Global South that have to choose between fertilizer subsidies and currency defence.

The transfer. Tens of billions of dollars over the next two years, flowing from retail and unhedged consumers to commodity professionals and vertically integrated physical merchants. This is roughly what happened in 2022, at similar scale, and for the same reason. Dumb money runs into the narrative the professionals are selling into. The professionals are selling into it because their models told them the narrative was wrong.

The falsification test

The conditions under which this thesis is wrong:

If by the end of 2027, war-risk premia for Hormuz transit have returned to their pre-2026 baseline and stayed there for two full renewal cycles, the insurance mechanism is disproved. If SPR policy in the US, China, India, and Japan is substantively unchanged from 2025 levels and geographic distribution, the stockpile mechanism is disproved. If infrastructure-fund transactions for Arabian Sea ports show no durable premium versus comparable Gulf-interior assets, the routing mechanism is disproved. If the US Fifth Fleet visibly returns to pure commons-provider posture with no further use as a coercive economic instrument, the navy mechanism is disproved. If no other chokepoint state begins to experiment with toll extraction or its functional equivalent by 2028, the precedent mechanism is disproved.

If all five of those falsifications hold, I am wrong. If any one of them fails to falsify, the thesis stands in part. If three or more fail to falsify, the thesis is durable.

I personally won’t be betting against all five.

What to do about it

Practical implications.

If you run energy-exposed operations (transportation, manufacturing, petrochemicals, logistics), your fuel hedging policy needs to be rewritten, not tweaked. Assume a permanently elevated volatility premium on Gulf-origin product and hedge accordingly. A 2025 hedging book is not fit for a 2026 world.

If you allocate capital to infrastructure, revisit your Gulf-region asset valuations with the chokepoint bypass premium explicit in your model. Duqm, Salalah, Sohar, Fujairah, and the pipeline network are structurally more valuable than they looked six months ago. Intra-Gulf assets without bypass optionality are structurally less valuable.

If you sit on boards with sovereign counterparty risk in the region, the composition of that risk has changed. The US security guarantee is now contingent in ways it was not before, which means the governance of Gulf-linked investments has to price US political volatility alongside regional political volatility.

If you run supply chains that assume timely delivery of Gulf-origin product, add six to ten days of working inventory in perpetuity. The tail risk on transit disruption is now fat enough to justify the working capital cost.

If you invest in shipping, the flag of convenience question has been reframed. Flag selection now has a risk implication that depends on which nations Iran grants toll exemptions to, which changes on political timescales. Your counterparty diligence on charterers needs a geopolitical overlay that it did not need in 2023.

If you run or allocate to agriculture, food production, or grain trading, fertilizer cost pass-through over the next two planting cycles is the variable to model. The structural repricing in nitrogen and phosphate markets will outlast any Hormuz reopening, and Russian market share gains in African and South Asian fertilizer trade are durable rather than reversible.

If you are in insurance, you already know all of this, and you are repricing now.

If you are a policymaker, I am not going to tell you what to do. But I will suggest that the assumption your predecessors operated under is gone, and you have an opportunity to shape what replaces it. That opportunity will not last.

The rodeo

Versions of this argument get made and dismissed after every Gulf crisis. 1988. 2003. 2011. 2019. 2020. Each time, the commentators who said the lesson was permanent were wrong in the medium term even when they turned out to be right in the longer one.

This cycle will likely not follow that pattern. The structural repricing through insurance, routing, SPR policy, US Navy posture, and the toll precedent has independent momentum. The Russian capture of fertilizer market share is already calcifying. The petrochemical and LNG channels are already flowing consequences into European industrial costs. And the institutional capital has already moved. The mispricing is not a prediction. It is a pattern that is already visible in the positioning reports, the commodity fund flows, and the forward curves.

The Friday rally was not a peace dividend. It was a wealth transfer from retail flows and positioning-driven buying into the hands of professional money that had already positioned for the opposite outcome. The professionals will sell back to the retail bid over the next two quarters at higher oil prices and lower equity prices than Friday’s close.

Whether this piece’s specific numbers are right matters less than whether the direction is right. On the direction, I would take the under on peace at current odds every time.

Sources and references

On the Hormuz crisis and the 17 April announcement

• NBC News, “Live updates: Iran declares Strait of Hormuz ‘completely open’; Trump says U.S. blockade ‘will remain in full force’ until peace deal”, 17 April 2026

• Bloomberg, “Iran Says Hormuz Strait Now Completely Open For Commercial Ships”, 17 April 2026

• Al Jazeera, “Iran war live: Strait of Hormuz open ‘for remaining period of ceasefire’”, 17 April 2026

• Wikipedia, “2026 Strait of Hormuz crisis” (for crisis timeline, toll route details, and transit data)

• Wikipedia, “Islamabad Talks” (for 11-12 April negotiations background)

• CNBC, “A few tankers and ships are going through the Strait of Hormuz”, 15 April 2026

• USNI News, “Strait of Hormuz Traffic Down As U.S. Blockade Appears to Deter Some Ships”, 14 April 2026

• Al Jazeera, “No ships ‘make it past US blockade’ in Hormuz strait in first day: Pentagon”, 14 April 2026

• Chatham House, “The Strait of Hormuz, shipping, and law”, April 2026 (for the legal framework of the blockade and UNCLOS analysis)

• Council on Foreign Relations, “U.S.-Iran Peace Talks Hit an Impasse. What Comes Next?”, James M. Lindsay, 13 April 2026

• NPR, “No Deal: U.S.-Iran peace talks in Islamabad fall through”, 12 April 2026

On fertilizer and adjacent industries

• Carnegie Endowment for International Peace, “Fertilizer isn’t getting through the Strait of Hormuz, which could lead to a global food crisis”, March 2026

• Carnegie Endowment for International Peace, “Beyond Oil: Hormuz Closure Puts Russia in the Lead in the Fertilizer Market”, March 2026

• CNBC, “Fertilizer prices surge amid Iran war, sparking food security warnings”, 25 March 2026

• Farmdoc Daily (University of Illinois), “Strait of Hormuz Closure and Fertilizer Supply Risks for U.S. Agriculture”, Arita, Chakravorty, Kim, Lwin, Steinbach, March 2026

• Brownfield Ag News, “Fertilizer prices to stay elevated through 2027 even if Strait of Hormuz reopens”, interview with Shawn Arita, NDSU Ag Risk Policy Center, April 2026

• Wisconsin Farmer / ICIS, “Strait of Hormuz shutdown chokes global oil and fertilizer supplies”, March 2026

• The American Prospect, “Iran War Sends Fertilizer Prices Sky-High”, April 2026

• Farm Policy News (University of Illinois), “Strait of Hormuz Blockade Could Further Hit Fertilizer Prices”, April 2026

On market positioning and flows

• Hedgeweek, “Exclusive: The ex-Rokos, Brevan veteran quietly building a commodities fund”, April 2026 (source for the institutional positioning narrative and the 50-100 day recovery argument)

• Hedgeweek Q1 2026 Allocator Survey (for forward-intent data on commodity hedge fund allocation)

• Intellectia, “Hedge Funds Pile Into Bullish Bets as Iran Ceasefire”, April 2026 (for CTA flow data and the tech-to-energy rotation analysis)

• Goldman Sachs prime brokerage commentary on CTA positioning, April 2026 (cited in Intellectia)

• Middlefield, “April 2026 Market Commentary” (for Canadian energy infrastructure positioning and rate-market implications)

• McKinsey, “Navigating a new era in commodity trading”, March 2026 (for trading house flex capacity and historical shock-profit patterns)

• J.P. Morgan Global Research, “Oil Price Forecast for 2026”, Natasha Kaneva, February 2026 (for pre-crisis demand-supply baseline)

• Trading Economics and Investing.com Brent and WTI price data, 14-17 April 2026

• Fortune, daily oil price tracking, 15-17 April 2026

On insurance and war-risk pricing

• Lloyd’s List Intelligence, Hormuz transit data and toll route analysis, April 2026

• Kpler maritime intelligence data on Hormuz transits, April 2026

Methodological note

The probability estimates in this piece (15/50/35 for durable resolution / muddled middle / collapse) are built from first principles rather than historical base rates, for the reasons given in the “Mispricing in oil” section. The fair-value Brent calculation and the asymmetric-payoff analysis are the author’s own work. The joint-probability derivation of the 10-20% durable-resolution estimate assumes approximate independence across the five conditions, which is a simplifying assumption; the true joint probability is plausibly lower if the conditions are positively correlated (all tend to hold or fail together) and higher if negatively correlated. Readers with different priors should rerun the calculation with their own inputs.

Reinsurance-market commentary reflects conversations with participants in the Gulf marine insurance market and is attributed in general terms rather than to specific institutions.

Equity-market sensitivity numbers (0.3% S&P earnings impact per $10/bbl oil move, VIX sensitivity to geopolitical tail risk) are approximate back-of-envelope estimates drawn from historical crisis episodes. A more rigorous analysis would use sector-level earnings elasticity models.

All views are the author’s own and do not reflect the positions of any institution, board, or investor with which the author is affiliated.

Someone sent me one of those infographics last week. “Countries with the Best Healthcare,” ranked by something called the Numbeo Health Care Index. Ecuador at number six. Thailand at eight. Canada thirty-fourth. The United States fortieth.

It’s nonsense, of course. Numbeo is a crowd-sourced survey. The “index” reflects how satisfied internet users feel about their local healthcare. It’s TripAdvisor for hospitals. But the fact that it circulates as though it measures something real tells you how badly we want a simple number for a complicated system.

So I tried to build a better one. I pulled data from the Lancet’s Healthcare Access and Quality Index (which measures deaths from 32 conditions that shouldn’t kill you if care is available), the Commonwealth Fund’s Mirror Mirror 2024 report (which surveyed patients and clinicians across ten countries on 70 performance measures), OECD Health at a Glance 2025, and the WHO. Eight pillars. Twenty-five countries.

When you plot composite outcome against per capita spend, something ugly emerges. There’s an efficiency frontier, a curve through the countries delivering the most health per dollar: Taiwan, South Korea, Japan, Finland. Most European systems cluster in a band above it, spending more for slightly better outcomes.

Then there’s the United States. $14,885 per capita according to OECD figures drawing on the 2024 CMS National Health Expenditure Accounts. A composite score below Canada’s, below the UK’s, below Spain’s. It sits alone, far below the frontier, like a state-of-the-art airport terminal where the baggage system doesn’t connect to the gates.

I’ve spent most of my career in infrastructure. Ports, energy terminals, transportation, sovereign investment. I’m used to looking at systems where assets exist but throughput fails, where money flows in but value doesn’t flow out, where the problem isn’t the equipment but the architecture connecting the equipment to the customer.

That’s what I see when I look at healthcare. The clinical capability is there. The transmission lines connecting it to patients are damaged, and almost nobody in the system gets paid to fix them.

The Windfarm With No Power Lines

The Canadian Institute for Health Information surveyed 467 imaging sites for its 2022-23 Medical Imaging Inventory. Canada has roughly 11 MRI units per million people, 28th out of 33 OECD countries. Japan has 57. The US has about 35. Those Canadian machines run an average of 15.3 hours per day, 97 hours per week. Only 17% of sites run around the clock. Nearly 14% of the equipment is over fifteen years old.

The machines sit idle overnight. Patients sit waiting for weeks.

In energy infrastructure, we’d call this a capacity factor problem. You’ve built the windfarm but you haven’t built the power lines. The electrons exist. Nobody can get to them.

The standard response: we don’t have enough staff to run the machines longer hours.

That answer assumes a staffing model from 1985. A study out of Northwestern Medicine, published in June 2025 covering nearly 2,000 clinical cases, found their in-house generative AI system boosted radiologist report completion efficiency by an average of 15.5%, with some hitting 40%. Unpublished follow-on work reported gains up to 80% for CT interpretation. Separately, AI-enabled acquisition tools from multiple manufacturers have cut scan times by up to 75% while improving image quality. Workforce projections estimate a US shortage of 42,000 radiologists by 2033.

The answer to that shortage is not 42,000 new training places. It’s redesigning the workflow so a radiologist spends time on judgment, not mechanics.

But who decides whether to redesign? The professional bodies who control scope of practice. The hospital systems whose staffing agreements are built around existing workflow. The EMR vendors who’d need to integrate tools they didn’t build. Some radiologists see AI as assistance. Others see replacement.

Nobody in that chain has an unambiguous incentive to change, even though the patient waiting three months would benefit immediately.

Similar story everywhere you look in healthcare. The technology exists. The evidence exists. The incentive doesn’t.

A Toll Booth Every Hundred Metres

In an energy terminal terminal, product arrived by ship, sat in tanks, left by truck and pipeline. At every stage: what’s the throughput, where are the bottlenecks, what’s the cost per unit delivered?

Healthcare never asks this at system level. Not because people haven’t thought of it, but because the answer is embarrassing.

According to the CMS National Health Expenditure Accounts, cross-referenced with OECD Health Statistics, US administrative overhead consumes approximately 15.5% of total healthcare spend. The OECD reports Canada at 8.1%, Germany at 6.8%. Taiwan’s National Health Insurance Administration publishes 2.4%. The OECD average sits around 3%.

The arithmetic: total US health expenditure in 2024 was $4.9 trillion. The excess admin above the 3% OECD benchmark is roughly 12.5 percentage points, which yields $600 to $750 billion depending on where you draw the line. That’s more than the US defence budget, and none of it touches a patient.

If you described a port where 15% of throughput cost went to documentation between the ship and the truck, you’d fire the operations manager.

The 15.5% persists because every dollar of that overhead is revenue to someone. The billing company, the prior-auth vendo, the health IT firm selling workflow software to manage the workflow that the other firm’s software created. A $600 billion administrative layer has spawned a $600 billion administrative industry, and that industry lobbies, donates, and hires.

The waste has a constituency. That’s why it likely persists.

The Brain Doesn’t Know It Has Two Doctors

Every hospital runs an electronic medical record system. Many of these systems don’t talk to each other, including, embarssingly inter-provincial healthcare in Canada. A patient moving between providers may as well arrive from another planet. The information exists. It sits behind proprietary walls, incompatible formats, and vendor contracts designed to prevent interoperability, because interoperability reduces switching costs, and reduced switching costs reduce vendor power.

In energy, this would be like every wind turbine manufacturer running on a different voltage. You’d never build a grid.

The silos go deeper than software.

I know a former head of one of Canada’s leading medical faculties who observed, only slightly joking, that psychiatrists and neurologists should be the same profession. They’re treating the same organ. The clinical logic is obvious. The schools would never agree, because the departments are separate fiefdoms with separate budgets, separate tenure tracks, separate journals, and separate identities built over a century.

So: a patient with depression and early-stage Parkinson’s sees two specialists who may never communicate, whose records live in different systems, whose treatment plans develop independently, whose interventions may actually conflict. The brain doesn’t know it has two doctors. The billing system does.

The cost shows up as duplicated tests because records don’t transfer, delayed diagnoses because information sits in someone else’s silo, treatments that conflict because the left hand doesn’t know what the right hand prescribed. And hovering over all of it, an administrative layer whose primary function is managing the complexity the silos created.

The technology to fix this has existed for years. Open APIs, standardised formats, patient-owned records. The incentive does not, because the people profiting from closed systems are the people selling closed systems.

Distributed Energy, Distributed Care

The old electricity model was centralised: large generating stations, long transmission lines, step-down transformers. Expensive, fragile, poor at serving remote areas.

The new model is distributed: solar on rooftops, batteries in communities, microgrids. Generation close to demand. The grid as coordination layer, not a delivery bottleneck.

Healthcare is stuck in the old model. Large hospitals are the power stations. The “transmission losses” are immense: missed appointments, chronic conditions deteriorating between visits, conditions escalating to emergencies because the upstream system never caught them.

OECD Health at a Glance tracks avoidable hospital admissions, defined as hospitalisations for asthma, COPD, heart failure, and diabetes that should be manageable in primary care, standardised for age and sex. The US sits at approximately 680 per 100,000. Germany: 570. The frontier countries, Italy, Switzerland, Portugal, range from 180 to 250. That gap has nothing to do with clinical capability. Those are patients whose upstream infrastructure failed them.

The distributed model moves primary care, chronic disease management, mental health, and routine diagnostics to the community and the home. The hospital concentrates on acute care, complex surgery, intensive care. You don’t use a nuclear power station to charge your phone. You shouldn’t use a tertiary hospital to manage diabetes.

Except distributed models reduce hospital revenue. A patient managed in the community doesn’t generate an inpatient admission. Fee-for-service, per-admission funding, procedure-based billing, the entire payment architecture penalises the shift.

The hospital administrator who halves avoidable admissions through community prevention has just slashed their own budget. Patients are healthier. The institution is weaker. The board is unhappy.

Energy utilities faced exactly this when distributed solar began cutting into kilowatt-hour sales. The regulatory solution: restructure payments so utilities earn from grid management and reliability, not from volume sold. Healthcare hasn’t made the equivalent move.

Follow the Money

Wendell Potter, a former Cigna executive turned industry critic, publishes annual analyses of SEC filings for the seven largest for-profit US insurers. His 2025 figures: collective revenue of $1.7 trillion, up roughly 300% from 2015. Operating profits of $54 billion. Combined CEO compensation of $146 million, drawn from proxy statements.

The ACA requires insurers to spend at least 80% of premiums on medical care, the medical loss ratio. Congress thought this was a cost control. The insurers figured out that 20% of a bigger premium is a bigger number. Raise the family premium 6% and your margin grows without improving a single outcome. The regulation designed to limit the industry became the incentive to expand it.

Economists William Lazonick and Öner Tulum, examining SEC filings for the fourteen largest publicly traded pharmaceutical companies from 2012 to 2021, found $747 billion spent on stock buybacks and dividends versus $660 billion on R&D. A separate US House Oversight Committee investigation covering 2016-2020 reached similar conclusions: $578 billion to shareholders, $522 billion to R&D. More capital leaving through shareholder returns than entering through research.

A 2023 study from the Institute for New Economic Thinking, led by Ekaterina Cleary, tracked NIH-funded publications linked to all 356 drugs approved from 2010 to 2019. They found that NIH funding, totalling roughly $230 billion over the decade (about 40% of NIH’s total budget), contributed to published research associated with every single approved drug. The NIH annual budget in fiscal year 2024 was approximately $48 billion.

The public funds the science. The private sector commercialises it. The surplus goes to Wall Street.

Then several major companies abandoned neuroscience R&D. Too risky. Returns too uncertain. They walked away from the brain because it didn’t fit the quarterly earnings cycle.

The OECD’s Mental Health Benchmark, published in 2021 and updated in Health at a Glance 2025, estimates that approximately two-thirds of people needing mental health care in OECD countries don’t receive it. The US Health Resources and Services Administration reports, as of September 2024, that circa 62% of designated Mental Health Professional Shortage Areas are rural.

Two-thirds of the people who need help can’t get it. The clinicians exist. The delivery infrastructure to connect them to patients at scale does not, and the industry that might have built it took the money and left.

What I’d Build

Every element here exists somewhere. No country has assembled all of them.

1. Universal coverage through competing non-profit sickness funds at under 5% administration. Prevention spending ring-fenced at 8% of total health budget.

   1. That 8% deserves explaining. The OECD reports member countries currently average about 3%. A 2017 systematic review by Masters and colleagues at the University of Liverpool, published in the Journal of Epidemiology and Community Health, examined 52 studies of public health interventions in high-income countries. They found a median return on investment of 14.3 to 1. The range was wide: negative returns on some influenza vaccination programmes through to 221:1 for lead paint control. National-level interventions showed even higher median returns of 27:1. The authors acknowledged likely publication bias toward positive results, and methodologies varied, so the 14:1 figure is a median of heterogeneous data, not a universal constant. But even halved, 7:1 is an extraordinary return. At 3% allocation, we’re chronically underinvesting in the highest-yielding capital in any healthcare system.

2. Mental health ring-fenced at 10%, delivered digital-first.

   1. That threshold comes from the Lancet Commission on Global Mental Health (2018), which recommended it as a minimum for high-income countries. Most OECD nations fall short. France leads at approximately 13%. Germany and the Netherlands hover around 10%. Most others sit between 5% and 8%.

3. Mandatory data interoperability. Open APIs, standardised formats, patient-owned records.

4. Pharmaceutical governance that strips tax credits from companies returning more in buybacks than they invest in R&D.

Estimated cost: $5,500 to $7,000 per capita. That range is based on the per capita spend of the highest-performing OECD systems: Japan at $5,250, Finland at $5,540, the Netherlands at $7,570, adjusted for the expanded prevention and mental health coverage I’ve described. The US currently spends $14,885 and delivers worse outcomes than all of them.

What Stops It

Nothing here requires technological breakthrough. It requires something harder: realigning incentives so the people who run the system benefit from the system working, rather than from the system being complicated.

Between insurer operating profits, excess administration above OECD benchmarks, and pharmaceutical shareholder returns exceeding R&D investment, the value being extracted from the current architecture runs somewhere between $750 billion and $1 trillion annually, depending on where you draw the boundaries. The healthcare and pharmaceutical sectors collectively spend over $700 million per year on lobbying, according to OpenSecrets data. The medical schools will not merge departments. The EMR vendors will not open their interfaces. The insurers will not reduce margins. The hospitals will not redirect revenue from beds to prevention.

None of this requires villainy. It requires only the ordinary operation of institutional self-interest in the absence of countervailing force.

I’ve worked in industries where entrenched operators insisted the system couldn’t change, right up until it did. Maritime shipping said containerisation was impossible. Energy said distributed generation would crash the grid. Ports said automation would never handle bulk cargo. In every case, the technology existed for years before the institutions accepted it. The lag between what’s technically possible and what’s institutionally permitted is where the waste accumulates and the patients wait.

Healthcare is in that lag right now. The clinical capability is extraordinary. The delivery infrastructure is decades behind. The incentives point the wrong way.

The system will change when the cost of maintaining it exceeds the political cost of reforming it. I suspect we’re closer to that threshold than the incumbents believe. But I’ve been wrong about tipping points before, and anyone who tells you they know when institutional inertia breaks is selling something.

What I do know is infrastructure. And this infrastructure is overdue for replacement.

Not repair. Replacement.

I’ve written before about the gap between expert scopes, the space where the biggest risks accumulate because nobody’s trained to look there. Healthcare might be the most expensive example I’ve encountered. The clinical expertise is extraordinary. A neurosurgeon in Boston, a cardiologist in Zurich, an oncologist in Tokyo, they are among the most skilled professionals alive. But the system connecting that expertise to the people who need it was designed by committees, distorted by lobbyists, and hasn’t been seriously re-engineered since the mid-twentieth century. An energy engineer would spot the capacity factor problem in an afternoon. A shipping logistics manager would see the routing failures in a week. A healthcare administrator, standing inside it, often can’t. That’s what scope blindness costs when the infrastructure is your health.

If you’re new here, Calamity Jane is the story of how I accidentally spent twenty-five years collecting crises across five industries. This piece is what happens when I point the same lens at a sector I don’t work in.

Sometimes the outsider’s view is the only honest one.

Sources

1. Numbeo, Health Care Index by Country 2026

2. GBD 2019 Healthcare Access and Quality Collaborators, “Assessing performance of the Healthcare Access and Quality Index,” The Lancet Global Health, Vol 10(12), December 2022

3. Blumenthal D, Gumas ED, Shah A, Gunja MZ, Williams RD, “Mirror, Mirror 2024: A Portrait of the Failing U.S. Health System,” Commonwealth Fund, September 2024

4. OECD, Health at a Glance 2025: OECD Indicators, OECD Publishing, Paris, November 2025

5. Canadian Institute for Health Information, Canadian Medical Imaging Inventory 2022-2023, CIHI, Ottawa (survey of 467 sites, 92.7% participation rate from publicly funded facilities)

6. Northwestern Medicine / McCormick School of Engineering, “New AI Transforms Radiology with Speed, Accuracy Never Seen Before,” June 2025 (study covering 1,957 clinical cases; 15.5% average efficiency gain, up to 40% individual gains; unpublished follow-on reporting up to 80% for CT)

7. CMS, National Health Expenditure Accounts 2024 (total national health expenditure $4.9 trillion; administrative overhead calculated as percentage of total)

8. OECD Health Statistics 2025 (administrative cost percentages by country; Taiwan figure from National Health Insurance Administration annual report)

9. Potter W, “2025: Big Insurance’s $1.7 Trillion Year,” Health Care un-covered, February 2026 (analysis of SEC 10-K filings and proxy statements for UnitedHealth, CVS/Aetna, Cigna, Elevance, Humana, Centene, Molina)

10. Lazonick W & Tulum Ö, analysis of SEC filings for 14 largest publicly traded pharmaceutical companies, 2012-2021 ($747B buybacks/dividends vs $660B R&D)

11. US House Committee on Oversight and Reform, Drug Pricing Investigation, July 2021 ($578B to shareholders vs $522B to R&D, 2016-2020, 14 companies)

12. Cleary EG et al., “Contribution of NIH funding to new drug approvals 2010-2019,” Institute for New Economic Thinking Working Paper, 2023 (tracked NIH-funded publications linked to 356 FDA-approved drugs; $230B in associated NIH funding; 99.4% of approved drugs had linked NIH-funded research)

13. NIH fiscal year 2024 budget: approximately $48 billion (congressional appropriation)

14. OECD, A New Benchmark for Mental Health Systems, OECD Health Policy Studies, June 2021 (two-thirds treatment gap estimate)

15. HRSA, Designated Health Professional Shortage Areas Statistics, September 2024 (61.85% of Mental Health HPSAs in rural areas)

16. Masters R, Anwar E, Collins B, Cookson R, Capewell S, “Return on investment of public health interventions: a systematic review,” Journal of Epidemiology and Community Health, 71(8):827-834, 2017 (52 studies in high-income countries; median ROI 14.3:1; range -21.3 to 221; national-level median 27.2:1; publication bias acknowledged)

17. Lancet Commission on Global Mental Health and Sustainable Development, The Lancet, October 2018 (10% minimum mental health budget recommendation)

18. OpenSecrets, Health Sector Lobbying aggregate data (combined healthcare and pharmaceutical lobbying expenditure)

19. Peterson-KFF Health System Tracker, “How does health spending in the U.S. compare to other countries?”, March 2026

Seventeen crises. Five sectors. Four countries. None of it on purpose.

A dear friend of mine used to call me Calamity Jane. For the uninitiated, the original Calamity Jane was Martha Jane Canary, a frontierswoman in the 1800s American West who rode with Wild Bill Hickok, survived smallpox, frontier wars, and barroom brawls, and earned her nickname because disaster followed her everywhere she went. She was never the cause. She was just always there when it happened. At the time I thought this was unfair. Looking back, I owe her an apology. Not for anything I did, but for the sheer actuarial risk of standing next to me for any length of time.

The Early Years: Learning to Sink

I joined the Royal Navy in 1996, which in hindsight was the last quiet year in global security for the rest of my natural life.

In September 2001, I was in the Navy Exchange in Naples, buying a Leatherman and a pair of Ray-Bans. The first tower had already been hit. I watched the second one live, on a screen mounted above the checkout, holding a multitool I hadn’t paid for yet. The Ray-Bans stayed on the counter. We redeployed immediately.

I spent the next several years in the Arabian Gulf doing the kind of work that sounds interesting at dinner parties and tedious when you’re actually doing it. Iraq, maritime interdiction, the full menu. At one point I was serving as a liaison officer aboard the French aircraft carrier Charles de Gaulle off the coast of Pakistan, which is the sort of sentence that sounds made up but wasn’t. The Gulf was not a relaxing posting during this period.

Then there was HMS Nottingham.

HMS Nottingham was a Type 42 destroyer. I use the past tense because she hit a rock off Lord Howe Island in 2002 and damn nearly sank. I was aboard. I was asleep at the time, which is either the luckiest or unluckiest detail depending on how you look at it. The Navy’s formal finding was navigational error. My informal finding is that the universe was not quite done with me yet, and needed me alive for the next twenty years of disasters.

Energy Trading: Perfect Timing

I left the Navy in 2007 and joined Compagnie Financière Tradition in London as an energy broker. I started in 2008.

If you don’t remember what happened in 2008, congratulations on your coma. Lehman Brothers collapsed in September. The entire global financial system spent the next eighteen months on life support. I had been a civilian for less than a year.

The client whose business had been the reason for hiring me pulled out of the market. So within months of starting a new career, my reason for being there had evaporated. I had to reinvent myself on the fly, find new clients in emerging market commodities, and build a new book from scratch.

It got worse. I then had front-row seats, in order, to: the European sovereign debt crisis (Greece discovering that lying about your national accounts has consequences), the Arab Spring (several oil-producing nations catching fire simultaneously), Libya’s civil war and NATO intervention (1.6 million barrels a day of oil production vanishing from the market virtually overnight, which if you are an energy broker is what we call “a busy week”), the Fukushima disaster (one of the world’s largest energy consumers shutting down its entire nuclear fleet overnight), and Russia’s annexation of Crimea (which, if you were in European energy, felt like the first shoe dropping on something much worse).

This was supposed to be my quiet career transition into the private sector.

Quebec: A Masterclass in Political Timing

I was then headhunted to join Investissement Québec, the province’s sovereign investment arm. A deputy minister under Premier Jean Charest personally recruited me. This seemed like a wonderful opportunity. I relocated my family to Montreal, which, if you are an Anglo who speaks fluent French but learned it in France, is a specific kind of fun. Nothing makes a Québécois trust you less than sounding Parisian.

By the time I had settled in, Charest was gone. Pauline Marois and the Parti Québécois were in power. Different party, different language politics, different investment philosophy, different everything. Then Marois got booted. Philippe Couillard came in. New premier, new CEO at IQ, new priorities.

Then François Legault got elected. Fourth premier during my time. At that point I had served under more Quebec leaders than most career provincial civil servants.

I should also mention that during my time at Ressources Québec, the global iron ore price collapsed. I was directly involved in the Cliffs Natural Resources restructuring, which was exactly as enjoyable as it sounds. Quebec had bet heavily on its Plan Nord mining strategy. The commodity gods had other ideas.

After IQ I moved to the Port of Montreal, where I was responsible for one of the largest infrastructure projects in the country: the new container terminal expansion. A multi-billion dollar bet on the future of Canadian trade.

The longshoremen went on strike. Then they went on strike again. Supply chains seized up. Cargo got diverted. Political pressure mounted from every direction. And then, mid-project, a global pandemic arrived. I was now trying to advance a major port infrastructure expansion while the workforce was intermittently walking off the job and the rest of the world was learning what “social distancing” meant.

I left the Port and walked straight into Terminal Norcan. Calm it was not.

Oil Goes Negative: The Terminal Years

In September 2020, mid-pandemic, I became President of Terminal Norcan, Canada’s largest independent petroleum products importer.

My interview with one of the board members took place in his private aircraft hangar. We wore masks and sat ten metres apart. This was the hiring process for running a critical fuel import facility. I accepted the job, showed up, and did not physically meet most of my staff for months. Try running an industrial operation when you’re not allowed to be in the same room as the people doing the work.

I set up a second control room so we could keep operating if one team got infected. I managed a terminal expansion project while steel prices were spiking to levels that made the original budget a work of fiction. And I was signing oil supply contracts in a market that was moving so fast the price could shift materially between the morning and the afternoon.

You may recall that earlier that year, in April 2020, oil prices had briefly gone negative. People were being paid to take oil off other people’s hands. The market I walked into was still reeling from that. When I tell you the supply chain was “disrupted,” I mean that the fundamental logic of buying and selling physical commodities had temporarily stopped making sense.

Then Russia invaded Ukraine in February 2022. European energy markets went berserk. Fuel prices did things that fuel prices are not supposed to do. I was still at the terminal, still importing petroleum products, now doing it during what was essentially a global energy war.

I stayed until late 2023. By that point I felt I had contributed enough.

I won’t pretend all of this was fun. There were nights at Norcan, mid-pandemic, mid-war, when I sat alone after everyone had gone home wondering if I’d got the calls right. I usually had. But you never know that at the time.

The Current Situation

I am now a CEO in the marine transportation sector. I will say nothing further about my current role except that the Strait of Hormuz has recently become relevant to my fuel costs, and leave you to draw your own conclusions about the pattern.

The Pattern

I have been doing the maths. In thirty years of professional life, across the Royal Navy, energy trading, sovereign investment, port infrastructure, petroleum importing, and marine transportation, I have personally been present for:

• The September 11 attacks (watched live, immediately redeployed)

• The Iraq War (in theatre)

• A Royal Navy destroyer sinking (aboard, asleep)

• The 2008 Global Financial Crisis (first year in finance)

• The European sovereign debt crisis

• The Arab Spring

• The Libyan civil war and NATO intervention

• The Fukushima nuclear disaster

• The annexation of Crimea

• Four changes of provincial government in a single posting

• The global iron ore price collapse (holding the provincial investment position during the Cliffs Natural Resources restructuring)

• Multiple port strikes

• A global pandemic (hired into a critical infrastructure role via socially distanced hangar interview, could not meet own staff)

• Oil prices going negative

• Steel prices spiking mid-expansion project

• Russia invading Ukraine

• The Iran/Hormuz fuel crisis (still going)

That is seventeen. Seventeen major crises, across five sectors, in four countries. I was not a bystander for any of them. I was in the room every single time.

I did not plan any of this. Nobody walks into a career saying “I’d like to be present for every major geopolitical and economic disruption of my generation, please”. It just kept happening. And because it kept happening, I kept having to deal with it.

She called me Calamity Jane. I’ve stopped arguing.

What Crises Actually Teach You

None of this makes me special. Thousands of people lived through the same events. But if you spend thirty years in rooms where the plan has just stopped working, you do start to notice a few things.

The first is that most plans fail on the same point: the assumption that conditions will hold. They won’t. The budget was built on last year’s fuel price. The project timeline assumed no strikes. The investment thesis assumed political continuity. Every crisis I’ve been part of started the same way, with a spreadsheet that turned out to be fiction.

The second is that the people who get through it aren’t the ones with the best plan. They’re the ones who can drop the plan fastest. The officer who’s still referring to the briefing document while the ship is listing is not the officer you want in charge. You need the one who looks up, sees what’s actually happening, and starts working from there.

The third is that most of the damage in a crisis isn’t caused by the crisis itself. It’s caused by the delay between the crisis arriving and people accepting that it has arrived. I have watched rooms full of intelligent people spend hours debating whether the situation is really as bad as it looks, while it gets worse. The single most valuable thing you can do when everything goes wrong is say “this has happened, what do we do now” instead of “surely this can’t be happening.”

That’s it. That’s what seventeen crises taught me. Not bravery, not brilliance. Just a short delay between the fire alarm going off and actually heading for the exit.

Though I would still quite like a quiet year. Just one. A library job, maybe.

Knowing my luck, it would probably catch fire.