There is a simple question we repeatedly ask about infrastructure:

Is it open or closed?

It sounds reasonable.

It is also often the wrong question.

A shipping corridor can remain legally open, physically passable and commercially impaired for years.

Ships still move.

Goods still arrive.

The system continues to function.

But traffic is lower.

Insurance is more expensive.

Transit takes longer.

More inventory has to be financed.

And more of the world's shipping capacity is consumed simply moving the same cargo over longer distances.

Nothing has technically failed.

But the system is no longer providing the same service at the same cost.

That distinction is the starting point of Article 3 in Series II.

And unlike the first two rows of this series, it gives us something genuinely encouraging.

Article 1 examined electricity.

There, substitution is extremely difficult.

A large grid connection can take years. Critical equipment can also take years. There is often no alternative network waiting beside the first one.

Article 2 examined frontier computing.

There, substitution exists in theory but becomes much harder at the leading edge. Advanced packaging, high-bandwidth memory and specialised equipment can create bottlenecks measured in quarters or years.

Shipping is different.

A vessel can change route within days.

Alternative corridors exist.

Some energy flows can use pipelines.

The global logistics system has repeatedly demonstrated that it can adapt.

That makes maritime chokepoints the most substitutable physical layer we have examined so far.

But it also reveals a different type of cost.

When the shortest route becomes too risky, the cargo does not necessarily disappear.

It takes another route.

For Asia–Europe traffic, rerouting around the Cape can add roughly 3,500 nautical miles and 10–14 days.

That sounds like a transportation problem.

It is actually a balance-sheet problem.

The carrier uses more fuel.

The importer waits longer.

Goods spend another two weeks in transit.

That means another two weeks of working capital tied up in inventory that cannot yet be sold.

Then there is the fleet itself.

A ship completing a longer voyage cannot simultaneously complete another voyage.

So a fixed fleet carries less cargo per year.

No ships need to disappear for effective shipping capacity to fall.

The same fleet simply becomes less productive.

This is how a disruption thousands of kilometres away can eventually increase freight rates on routes that never touch the affected corridor.

The event happens in one place.

The cost propagates through capacity.

This creates one of the most interesting paradoxes in resilience.

Failure is easy to see.

A factory stops.

A shipment does not arrive.

A shelf becomes empty.

Successful adaptation looks completely different.

The shipment arrives.

Just later.

The invoice is larger.

The importer carries more inventory.

The retailer's margin gets slightly thinner.

Eventually the final price rises.

By then, the original disruption has disappeared inside freight rates, working capital and margins.

There is no line on the consumer's receipt labelled:

chokepoint disruption.

And there is rarely a line in a board report labelled that either.

The system has successfully prevented the visible failure.

That success makes the cost harder to identify.

Nothing failed, so nobody adds up what it costs to keep everything working.

There is another assumption worth questioning.

We tend to imagine that recovery is simply disruption running backwards.

Threat rises.

Traffic leaves.

Threat falls.

Traffic returns.

But companies are not starting from the same position when they return.

They have already rewritten schedules.

Insurance arrangements have changed.

Contracts have changed.

Customers have adapted to new delivery times.

Networks have been reorganised around longer routes.

Returning to the old route therefore requires another reorganisation.

And the risks facing a carrier are asymmetric.

Staying on the longer route is expensive.

Returning too early and losing a ship can be catastrophic.

Those two outcomes are not equivalent.

So falling incident counts do not automatically produce proportional traffic recovery.

The extraordinary risk premium can gradually become part of ordinary operations.

Risk does not disappear. It normalises.

“Open” and “closed” are useful legal descriptions.

They are poor operational metrics.

A better dashboard would ask:

What share of normal traffic is passing?

At what insurance rate?

With what transit time?

How much fleet capacity is being consumed by the alternative?

These variables tell us whether infrastructure is actually usable.

And this principle extends far beyond shipping.

An electricity network can technically function while a new customer waits years for connection.

A hospital can remain open while access to treatment deteriorates.

A supplier can continue accepting orders while delivery moves from weeks to a year.

A system does not have to stop operating to lose much of its usefulness.

The article's base path, at 50%, is a permanent premium.

Traffic partially recovers, but elevated insurance and routing costs settle into the baseline rather than disappearing.

The 25% path is repeat disruption.

Another significant chokepoint event resets the recovery and pushes companies further toward treating alternative routing as standard network design.

The 15% path is genuine normalisation.

Security improves durably, insurance moves back toward previous levels and traffic returns.

And the final 10% path is the one that matters most for systemic risk:

two chokepoints impaired at the same time.

Why is this different?

Because the alternative to one disruption depends on spare capacity elsewhere.

If two corridors are impaired simultaneously, the reroute from the first can collide with congestion created by the second.

Individual chokepoints are relatively substitutable.

The network of chokepoints is less so.

That is the failure mode simple contingency planning often misses.

This is one of the few substrates in Series II where preparation can change the outcome quickly.

Start by mapping physical flows.

Not just suppliers.

Routes.

Which goods cross the Red Sea?

Which depend on Hormuz?

Which use Malacca, Panama or the Turkish straits?

Several apparently independent suppliers may depend on the same corridor.

If so, supplier diversification has not created route diversification.

Then replace contracted transit assumptions with actual transit times.

A logistics model built around old Asia–Europe journey times can remain systematically wrong even though every individual shipment eventually arrives.

Most importantly, pre-negotiate the alternative.

The ship itself can change course quickly.

Commercial arrangements cannot.

A company that already has alternative routing priced, carriers contracted and customers informed can react in days.

A company beginning those negotiations during the disruption may need weeks.

That difference is the practical value of resilience.

For most households, maritime disruption is more likely to appear through prices and delivery times than through shortages.

That changes the useful response.

The rational expectation is not empty shelves.

It is slightly higher prices for imported goods and longer delivery times that may persist long after the original security situation improves.

For large imported purchases, realistic delivery assumptions matter more than reacting to dramatic headlines.

Insurance deserves more attention.

In the case examined in the article, war-risk pricing moved before the military escalation.

That is important because it is not commentary.

It is behaviour.

Someone exposed to the risk changed the price of carrying it.

Transit counts provide the second useful signal.

And divergence between cargo types provides a third.

If tankers and container ships make different decisions about the same water, the difference tells us something important about the economics of the cargo rather than simply the security of the route.

Series II now has enough rows for the comparison itself to start producing information.

Electricity:

poor substitutability, measured in years.

Frontier computing:

high concentration at critical layers, with substitution measured in quarters or years.

Maritime chokepoints:

real alternatives, with rerouting measured in days or weeks.

And yet the third category generates by far the most visible alarm.

That is revealing.

Public attention does not necessarily track substitution time.

It tracks visibility.

Ships can be filmed.

Routes can be mapped.

Attacks produce images.

A grid connection queue does not.

A qualification bottleneck inside advanced packaging does not.

So the physical constraint that dominates the news can be easier to work around than the quiet infrastructure constraint nobody sees.

That gives us another distinction worth keeping:

Visibility is not severity.

And it reinforces why Series II uses the same instrument for every physical layer:

Concentration.

Criticality.

Substitution time.

Not how dramatic the event looks.

Not how many headlines it produces.

Not whether somebody describes the system as “open.”

The useful question is simpler:

How much of the system can actually be used, at what cost — and how quickly can we replace what cannot?

That is OPEN IS NOT THE SAME AS USABLE.

Next in Series II:

Water — the fourth row.

THRIVE IN CHAOS
Decision Intelligence for an Uncertain World

Analysis → Forecast → Recommendations
Signal → Meaning → Action → Stability

Forecasts are probability-based analytical assessments, not certainties. This material supports independent judgment and does not constitute financial, legal or investment advice.