Supply Chain Action Points
Read this first — the conclusion, and the moves to make:
- This week, convert 100% of next quarter volume into bookings with a named vessel and a written roll policy, and stop treating a rate agreement as a space agreement.
- Cap any single string at 60% of a month's volume for the next two quarters; at 0.9% idle there is no charter market to substitute into when a loop breaks.
- Report blocked ships, frozen capacity and idle share on one line every month, and treat idle under 1% as a standing instruction to pre-approve air or rail for the top 10% of SKUs by margin.
- Get equipment release confirmed in writing 10 days before CFS cut-off; the binding constraint has moved from slot to box, and the box is decided earlier than the slot.
- By 15 July, compute weeks of cover against a seven-day roll for your top 10 SKUs; anything under two weeks gets buffer stock or a second origin before the next allocation round.
Analysts estimate roughly 50 ships were blocked in the Persian Gulf and about 10% of global container capacity frozen in port congestion by late June, with the idle fleet down to 0.9% of the total. Red Sea rerouting via the Cape and Middle East tensions continue to suppress effective supply.
The Analysis
Three figures came out of the same estimate: roughly 50 ships blocked in the Persian Gulf, about 10% of global container capacity frozen in port congestion, and an idle fleet at 0.9% of the total. Two of them describe damage. One describes the ability to absorb it, and that is the one worth reading twice.
Fifty ships is a fleet count with no denominator. Fifty 8,000 TEU ships and fifty 24,000 TEU ships are different events, and neither tells you what happened to the capacity that was supposed to be moving. Ten percent frozen is a better number because it has a base, but it is a stock: capacity parked, not capacity destroyed. The 0.9% changes how you plan, because idle fleet is the buffer a network spends when something breaks.
Put the two together and the arithmetic is not subtle. The frozen capacity is roughly eleven times the entire spare fleet. That ratio is the story. The ship count is the caption.
A number without a denominator is a headline. Fifty vessels stuck is a fact about ships, not about space, and space is what you book. The estimate that carries information is the ten percent, because it says one tenth of the world box fleet is currently holding cargo it cannot discharge, and the 0.9% says there is almost nothing left over to replace it.
The ten percent is a stock, and stock is only half the drain. Rerouting around the Cape does not park vessels, it consumes them: the same ship spends more days at sea per round voyage, so the same fleet delivers fewer voyages a year. That is a second, quieter loss, and it never appears in a congestion count because those ships are moving. Capacity parked and capacity burned both show up in the same place eventually, which is the space you cannot get.
The loop is the part worth understanding, because it explains why the three numbers held steady rather than improving. Longer voyages mean fewer round trips per vessel per year, which means arrivals bunch instead of spread, which means discharge terminals queue, which means more capacity counts as frozen, which means idle stays at 0.9%, which means the next disruption has nothing to absorb it. Each step feeds the next. A network in that state does not return to normal when the trigger eases; it returns when the queue clears, and the queue clears last.
Here is where the accounting gets interesting. A ship waiting off a terminal is employed, not idle. It has a charter, a crew, a full load and no revenue-producing movement. So the two headline metrics are counted in different buckets and they describe the same ship twice, once as frozen capacity and never as idle. That means the 0.9% understates how tight the market is rather than overstating it. Near-100% utilization is not a measure of efficiency. It is a measure of the fact that the denominator has stopped telling you anything.
Which changes the shape of the cost, not just the size. With a spare fleet to charter, a disruption is a price event: you pay more and you get the box. With the buffer gone, it is a timing event: you pay more and you still may not get the box. Those are different problems and they do not have the same fix. A price event is managed with a budget. A timing event is managed with inventory, and inventory decisions are made weeks before the price is quoted. Most planning models still treat this as the first kind, which is why the surprise is always about timing when everyone expected it to be about money.
Who feels it first is not who shouts loudest. Space goes before price, and it goes in a fixed order. A large buyer with named space and a service commitment is protected on the slot and exposed on everything downstream of it, meaning equipment release, free time at destination, and the inland leg, because boxes end up in the wrong places when rotations stretch. A mid-size importer with a contract but no named allocation gets rolled first, because a contract is a rate agreement and rolling costs the carrier nothing but a phone call. A small spot buyer pays and waits, and cannot plan beyond the next booking. On the export side the shortage arrives earlier than most people expect, because empties get pulled toward the high-value headhaul and secondary origins lose their equipment first.
Transmission runs in a fixed order too: space, then equipment, then schedule reliability, then rate. Rate is last. That ordering is the single most useful thing in this story, because waiting for the rate signal means the failure has already happened in a form money cannot fix. By the time the index moves, the rollovers have been happening for weeks and the stockout is scheduled.
What the coverage hides is a broken metric. When a buffer metric approaches zero it stops measuring the buffer and starts measuring the absence of one, and the number keeps printing as if it still meant something. Utilization at 100% reads like a market at full stretch. It actually reads like a market where the reserve line has been deleted from the report. The derived figure that still works is the ratio, ten percent divided by 0.9%, roughly eleven to one, and the trend of that ratio. Watch the ratio, not either level. If the ratio widens, the system is losing its ability to substitute, and no amount of negotiation at the booking desk changes that.
Run the arithmetic you can actually run without a freight rate. Assume 40 FEU a month, and assume one booking in ten gets rolled by a week, which is my assumption, not a published figure. That is four FEU a month, forty-eight FEU-days of schedule loss. What those days cost depends on your own numbers, a line-down hour, a promotion window, a penalty clause, and I am not going to put a dollar value on cargo I have not seen. What you can compute today is cover: weeks of inventory per SKU against a seven-day roll. If your top ten SKUs sit under two weeks of cover, you are one roll from a stockout, and that is a purchasing problem wearing a freight costume.
The counterfactual is worth one line. If there were a charter market to rent from, this would be an expensive quarter and then it would end, because price would clear it. There is no charter market at 0.9%. So the disruption converts into delay, and delay converts into inventory, and inventory converts into a decision your operations team has to make without a rate to anchor on. That is why the useful question is not how high rates go. It is how many weeks of cover you hold when the next ship does not arrive.
Lock space, not rate, while idle sits under one percent. The rate is renegotiable in a phone call. The sailing is not, and the booking you did not confirm in June is the one you will be explaining in August.
↑ Back to the key points