On 25 August the Panama Canal Authority rewrote one line of its booking rules. Slots that are cancelled or become available, it said, would now be allocated in a set priority order: LNG first, LPG second, car carriers third, full container vessels last.
Containers had held that first claim. They now hold the last. For the first time, gas carriers sit at the front of the queue at the most contested waterway in the Western Hemisphere.
It should have been the best news the US Gulf LPG trade had received in three years. Instead, it landed on a fleet that had already stopped queuing. In April, 43 per cent of US Gulf–to-Asia LPG voyages went around the Cape of Good Hope — the highest share since October 2016 — while the count of gas carriers waiting at the canal barely moved.
Priority in a queue you have already left is worth nothing.
The short version
- Panama has stopped being a route and become an allocation system. Neopanamax slots are down to nine a day, and only one of the nine reaches the spot auction.
- Booking demand has exceeded available slots every month since March — utilisation of 113.7%, 105.7%, 113.9%, 105.9% and 112.3%.
- ACP has moved gas to the front in both mechanisms: LNG and LPG are auction Group 1, and now first in line for cancelled slots.
- It arrived late. The Cape took ~43% of US Gulf–Asia LPG voyages in April, a record since 2016.
- The auction cannot cap itself: a slot is worth the sea days it saves. Nineteen days at recent freight is $2.9m — so $2m bids were still the cheaper option.
- And the routing alone removes ~18% of effective fleet capacity from the lane.
A Drought You Cannot See in the Lake
The instinct is to check the water level, and the water level looks fine.Verified
Gatun Lake stood at 84.06 feet on 8 September. Against the fifteen-year mean for that date of 84.53 feet, that is a deficit of less than half a foot. It is four and a half feet above the 79.57 feet recorded on the same day in 2023, at the depth of the drought that shut the canal down to twenty-two transits a day.
The problem is not the level. It is the direction.
ACP's own daily record, which runs unbroken back to 1965, shows 2026 opening almost four feet above normal — 88.00 feet in March against an 84.02-foot fifteen-year average — and falling through the average during the months when the lake is supposed to refill. Between 1 July and 8 September, Gatun lost 0.87 feet. That is the weakest rainy-season performance in the twelve-year record, worse than 2015's minus 0.76 and worse than 2023's minus 0.09.
Advisory A-29 gives the mechanism: cumulative rainfall 34 per cent below the historical average and watershed inflows 44 per cent below, across the hydrological year to August. The lake is high because it started high, not because it is being replenished — and ACP is planning for a potentially severe 2026–27 El Niño hitting the dry season that begins in January.Verified
Cutting Transits, Not Draft
Faced with that, ACP had two levers. It has used one of them and deliberately held the other.Interpretation
Draft limits how much cargo each ship carries. On 4 September, advisory A-33 postponed the reduction to 47.5 feet indefinitely, leaving 48 feet in force; the authority's own published projection holds Neopanamax draft at 49 feet into November. Slots limit how many ships pass at all. Those were cut: nine daily Neopanamax slots from 4 September, and Panamax down to 25 and then 23 from the 15th.
Draft saves water per transit. Slots save transits. Cutting the second while deferring the first is a decision to bank lake volume for the dry season rather than manage today's depth.
Once a waterway rations transits rather than cargo, the allocation rules stop being administration and become the market.
And the system being rationed was already oversubscribed. Aggregating the booking tables ACP publishes in its monthly operations advisories — a series nobody appears to compile — Neopanamax reservations have exceeded available slots in every month since March.
Gas Moves to the Front
Two advisories, eleven days apart, changed where gas carriers stand when capacity is short.Verified
A-29, on 20 August, divided auction access into four groups. LNG and LPG carriers are Group 1. Dry bulk and general cargo are Group 2, container ships and vehicle carriers Group 3, and chemical, crude and product tankers Group 4.
A-30, on 25 August, replaced the sentence in Notice to Shipping N-07 that gave full container vessels priority on carried-over and cancelled slots. In its place: slots that become available for any reason are allocated through a special competition in the order LNG, LPG, car carrier and RoRo, then full containers.
Together they put gas first in both mechanisms by which a scarce slot can be obtained. ACP has published no reasoning; the effect is a structural preference for the two cargo types most likely to abandon the route.Interpretation
The Fleet Had Already Gone
Here is the part that makes the promotion hollow.Verified
In April 2026, roughly 43 per cent of US Gulf-to-Asia LPG voyages routed via the Cape of Good Hope — the highest share since October 2016, on Argus assessments. And the canal showed no sign of gas carriers waiting: Vortexa's count of VLGCs and VLECs at the North Panama waypoint stayed moderate, while the four-week average of oil tankers waiting climbed from 25 in the week to 1 March to 45 by mid-April, a record in its ten-year dataset.
Gas carriers were not queuing. They had stopped coming.
The reason was cost. Auction prices for unbooked Neopanamax slots ran $150,000 to $350,000 earlier in 2026, reached nearly $1.7 million by mid-April and approached $2 million in mid-May. Waiting time for unbooked southbound transits went from about two days to eleven on 17 April, and eighteen days by August.
And freight is only half the cost of going the long way
The nineteen days do not only consume vessel time. They move the cargo into a different pricing month, and that is frequently the larger number.Interpretation
LPG delivered into Asia prices against a monthly index. Add nineteen days and a cargo can slip a pricing month, or miss its delivery window altogether. Whether that helps or hurts depends entirely on the shape of the forward curve.
Through late August the curve was backwardated. Argus assessed the US–Asia paper arbitrage at $385.00 a tonne for September on 28 August, and at $361.50 for October on 1 September — a month-on-month decline of $23.50. On a 44,000-tonne cargo, arriving one pricing month later was worth $1.03 million less.
Add that to the time cost and the detour at 28 August freight cost roughly $3.97 million, not $2.93 million. A $2 million slot was not $0.9 million cheaper than the Cape — it was closer to $2.0 million cheaper. The pricing month roughly doubled the gap.
That reverses under contango. If the forward month is worth more than the prompt, arriving later is not a penalty, and past roughly $21 a tonne of contango the Cape becomes the cheaper option outright even against a $2 million slot. The routing decision therefore embeds a view on the curve as much as on freight — which is why two charterers can look at the same slot price on the same day and reach opposite conclusions.Interpretation
Which explains why the auction has no natural ceiling. A slot is worth what the detour costs, so the rational maximum bid rises with both freight and backwardation. At $80,000 a day the time component alone is $1.5 million; at $160,000 a day it is $3.0 million. Every dollar of freight lifts the ceiling by nineteen — and every ship that gives up and sails the Cape tightens the fleet and raises freight again.Interpretation
Eighteen Per Cent of the Fleet, Without a Single Ship Leaving
The number that matters to anyone pricing US Gulf LPG is not the slot cost. It is what the detour does to effective vessel supply.Interpretation
A round trip via Panama is about 52 days; via the Cape, 90. That is roughly 7.0 voyages a year against 4.1. Each diverted vessel loses about 42 per cent of its effective capacity — not because anything happened to the ship, but because it spends longer at sea per cargo delivered.
Apply that to April's 43 per cent Cape share and roughly 18 per cent of effective fleet capacity disappears from the US Gulf–Asia lane through routing alone. No vessel is scrapped, no yard is late, no cargo is cancelled. The tonnage simply does less work.
Which raises the obvious question: why is nobody short of ships?
Because the same disruption that lengthened the voyages also freed the fleet to sail them. Veson Nautical forecasts Middle East LPG exports falling 48 per cent in 2026, and Vortexa counted a record 150 VLGCs ballasting toward the US Gulf by mid-March. Tonnage that would have been shuttling Ras Tanura to Chiba has been repositioned into the Atlantic basin.
So the lane lost about 18 per cent of effective capacity to routing and gained displaced Gulf tonnage at the same time. If that inflow lifted Atlantic availability by roughly a fifth, the two broadly cancel — which reconciles an alarming-looking capacity calculation with a market where charterers report no difficulty finding a ship.Interpretation
It also carries a warning. The offset is borrowed, not structural. Whenever Hormuz normalises, the displaced tonnage goes home to short-haul Gulf voyages — and if the canal is still rationing transits at that point, the routing loss stops being cushioned.Interpretation
The honest objection
Three arguments cut against this. The Cape share is an April reading taken at the height of the Hormuz disruption, and routing decisions respond to several variables at once — canal cost is one input, not the only one. The 19-day differential assumes a constant 15 knots and ignores bunker economics, which favour the canal further and so understate the slot's value. And ACP's promotion of gas may yet work: a shipper holding a long-term allocation gets certainty the spot market cannot buy, and those allocations are the three slots in nine that never reach auction. It is also fair to note that the fleet-capacity loss has so far been masked by tonnage displaced from the Gulf, so the routing shift has not yet been tested against a normal Middle East trade. The claim here is narrower than "the promotion failed" — it is that priority was granted at the point in the cycle when the economics had already moved the fleet.
What Follows
1 · Price the slot against the detour, not against the toll — and include the pricing month. A Neopanamax slot is worth nineteen days of time charter equivalent plus whatever the forward curve says about arriving a month later. At 28 August values that was close to $4 million, not $3 million, which is why bids near $2 million cleared comfortably. Anyone valuing the slot on freight alone is understating it by roughly a third in a backwardated market — and overstating it in a contango one.
2 · The long-term allocation is now the valuable instrument. Three of nine daily slots sit in LoTSA and never reach the spot market. In a rationed system with gas in Group 1, a multi-year allocation is not an administrative convenience — it is the difference between a 52-day and a 90-day round trip, repeated.
3 · Watch the lake's direction, not its level. Gatun looks comfortable and is falling in the wrong season. The series that matters is the July-to-December change, and ACP publishes it daily and free. If the wet season closes without a meaningful refill, the 2027 dry season begins from a lower base than 2024 did.
4 · For Pacific-facing exporters, this is the quiet argument for proximity. None of it applies to a cargo that never approaches the canal. A West Coast barrel reaching northeast Asia in ten to eleven days carries no slot cost, no auction exposure, no pricing-month slippage and no allocation risk — and it is indifferent to whether Gatun refills. That optionality has never been priced into a Canadian netback, because until this year there was no observable market price for canal access. There is now, and on 28 August it was worth roughly $2 million a cargo.Interpretation
What would change our mind
- A strong end to the wet season. If Gatun recovers materially between now and December, slots return, the auction deflates and the Cape share normalises. ACP publishes the lake level daily — this is checkable within weeks.
- Auction prices decoupling from freight. If clearing prices stall while BLPG3 rises, the slot is being priced on something other than time saved and our model is wrong.
- Gas carriers returning to the queue. If VLGC waiting counts at the North Panama waypoint climb, the promotion is working and the fleet is coming back.
- The Cape share falling with slots still at nine. That would mean routing is being driven by cargo economics rather than canal access, and the causal chain we describe is weaker than we think.
The Queue They Already Left
For three years the canal has been treated as a weather story. It is not. The lake is within half a foot of its fifteen-year average, and the draft cut that was to bite on 1 October has been postponed indefinitely.
What has changed is that the canal now rations transits rather than cargo, and once a waterway does that, its booking rules become the market. Nine slots a day, one of them reachable without a long-term contract, and a published order of precedence deciding who gets the rest.
In August, gas moved to the front of that order for the first time. It is the best position LPG has ever held at Panama, and it arrived to find the berth empty — because the fleet had already done the arithmetic, and at nineteen days and four million dollars a cargo it had stopped asking.
Insight Stream — LNG/LPG & Waterborne Trade Analysis
LNG contract economics, LPG benchmarks and Pacific trade flows, VLGC freight, China PDH and NGL export infrastructure. Independent market intelligence, built only on sources you can check.

