On 20 August the Panama Canal postponed two draft reductions it had already scheduled, and cut the number of ships it will lock through each day instead. Most coverage filed the two halves separately — a reprieve on draft, a tightening on transits. They are the same decision. The Canal is spending transit slots to buy back depth, and the bill for that trade does not land on the ships that keep their cargo.
One definition first. The canal has two sets of locks, and this concerns the newer one — the Neopanamax locks, opened in 2016, 51.25 metres wide and normally operated to a draft of 15.24 metres, or 50 feet. The original 1914 chambers are a separate passage with a separate, shallower limit.
What actually happened
Advisory to Shipping A-29-2026, issued on 20 August, moved the reduction to 48.0 feet from 26 August to 2 September, and the reduction to 47.5 feet from 3 September to 1 October. In the same document it cut daily transits: from 3 September, nine slots at the Neopanamax locks and twenty-five at the Panamax locks; from 15 September, the Panamax figure drops again to twenty-three.
The Authority gave its reasons in its own numbers, which is worth crediting because most water managers do not. Accumulated rainfall for the current hydrological year is down by about a third and inflows to the watershed by rather more — the Authority puts the shortfall in inflows in the mid-forties of a percent. Gatun Lake stood a little above 84 feet in mid-August, roughly a foot under its five-year average for the month, with a published expectation of falling toward the high 82s by the middle of October. The Authority also said, in terms, that conditions were changing faster than it had anticipated and that further adjustments were possible. None of that is a body concealing a problem.
Why the two halves are one decision
A draft limit and a transit limit look like the same kind of restriction and are not. They answer different physical problems, and confusing them is where most of the commentary goes wrong.
Maximum authorised draft is a depth question. It is set by how much water stands over the lock sills and along the channel, which is to say by the level of Gatun Lake. When the lake falls, the usable depth falls with it, and the Canal publishes a lower ceiling.
Daily transits are a volume question. Every lockage moves a large quantity of fresh water from the lake toward the sea. Water-saving basins recover part of it and cross-filling recovers more, but the net draw is real and it scales with the number of lockages. Fewer transits means less water leaving the lake, which means the lake sits higher than it otherwise would, which means the depth holds.
So the two levers are connected in one direction: transits are the thing you spend to keep draft. That is what the 20 August advisory did. It did not relieve a constraint. It moved the constraint from the ships that transit to the ships that want to.
Who pays, and in what currency
The two restrictions send their bills to different addresses, and this is the part worth pricing.
A draft cut is paid in tonnes, by whoever was loading deep. It is spread thinly across every laden ship using the larger locks, it never appears as a line item, and it is invisible in freight indices because it shows up as a slightly smaller parcel rather than a higher rate. A centimetre of immersion is worth roughly sixty-five tonnes on a Kamsarmax — the figure a draft survey turns into a bill of lading number — but the real loss for any given ship is only the gap between her own summer draft and the new ceiling. Most ships were not loading to 50 feet. The restriction is felt unevenly and reported quietly.
A transit cut is paid in dollars and days, by whoever does not hold a booking. It concentrates rather than spreads. The market for the remaining slots is an auction, and auctions do not distribute scarcity evenly — they hand the whole cost to the marginal bidder. Through August that bidder has been paying conspicuously: an LPG carrier sailing in ballast took a slot for a reported record of about 4.6 million dollars in mid-August, and an Authority spokeswoman told reporters that median auction results had roughly tripled from around fifty-five thousand dollars in late 2025 and early 2026. On 20 August, thirty-six vessels were waiting without a reservation, at an average of about eight days in each direction.
The order of events matters here, and it is where most of the commentary has the causation backwards. That premium was built by demand. Gas and energy cargoes rerouted by the closure of the Strait of Hormuz, together with Asian buying of US energy exports, had already been bidding into a tight window through the spring and summer — before the Canal touched supply at all. On 3 September the Canal starts taking slots out of that same market. Restricting supply into a queue that is already being bid up by a separate shock does not produce a proportionate price move. It produces a disproportionate one.
There is a third change in the same advisory that almost nobody has reported, and it is the most revealing of the three. From 3 September the auction is split into four vessel categories — LNG and LPG carriers in one group, dry bulk and general cargo in another, containers and vehicle carriers in a third, chemical and crude tankers in a fourth — and a customer already holding a reservation may not take an additional slot unless the bid is uncontested. That is not a water measure. It is a decision about who is allowed to bid against whom. Carving gas carriers into a pen of their own is the Authority conceding, in regulatory language, that one segment had been clearing the market at everyone else's expense.
Those two bills fall on different businesses. The tonne cost lands on cargo interests with long-run contracts of affreightment, and it is absorbed. The slot cost lands on spot and time-critical trades — gas carriers, reefer, boxships chasing a service window — and it is not absorbed, it is bid for. The 20 August advisory moved weight from the first group to the second, and the Panamax locks, where most of the transit cut falls, carry a great deal of the trade that has never bought an auction slot in its life.
Which ships are actually in the draft band
Before going further it is worth being precise about who the draft ceiling touches at all, because it is a narrower group than the coverage implies. The old 1914 chambers take ships up to 32.31 metres wide — the dimension that gave the Panamax class its name — and most dry bulk still transits them under their own, lower limit. Ships too wide for those chambers book the Neopanamax locks and pay more to do it. That is the tonnage these draft advisories govern: laden container ships, gas carriers, and the bulk carriers big enough to justify the larger chamber, Kamsarmax and above. Our guide to bulk carrier sizes sets out how the classes map onto those dimensions.
What this is not
Two readings should be discarded before they take hold, and one of them was ours.
The first is that scheduling several dated steps in a single advisory shows unusual foresight. It does not. It is the Canal's ordinary drafting practice. Advisory A-22-2023 carried three dated levels in one document, and it existed only to postpone a step that A-20-2023 had set three weeks earlier. A batched schedule is a plan, not a commitment, and the Canal has revised its own more than once. Anyone reading the 2026 pairings as a signal of anything is reading a house style.
The second is that the Canal Authority is failing. Nothing in the record supports it. The advisories are numbered, dated, public and issued ahead of effect; the postponement was published fifteen days before the step it deferred; the administrator has said plainly that 2023 and 2024 were the rehearsal. This is a water manager doing arithmetic in the open. The question is not how Panama is managing a shortage. It is whether the freight market has noticed which lever moved.
What the forcing actually implies
The Climate Prediction Center's discussion of 13 August puts an El Niño Advisory in effect and gives better than a 90% chance of a very strong event through the coming Northern Hemisphere autumn and winter, with a 69% chance of an event exceeding anything since 1950 on a three-month relative index during October to December. The relative index matters: it is measured against the tropical-mean sea surface temperature, so the threshold is a higher bar than the same number on a raw anomaly, and quoting it without that qualification overstates it.
The timing is where the trade press has been sloppy, ourselves included in draft. The relationship between El Niño and this particular watershed is well established — Smith and colleagues, working on Panama Canal Watershed discharge, find the ENSO signal strongest in the December-to-March window, and the World Weather Attribution group found four of the five driest years in the modern record fell in El Niño years. But a reservoir integrates a rainfall deficit rather than tracking it. An event peaking in the fourth quarter of 2026 does not do its damage in the fourth quarter of 2026. It does it to the dry season that follows, which is to say the first four months of 2027, entered from a lake that never refilled.
Which reframes what happened on 20 August. Deferring a draft cut by four weeks and paying for it with transit slots is a sensible move if the constraint is a few weeks of shortfall. It is a considerably more expensive one if the lake has to be carried through to April.
What would have to be true for this to be cheap
Treating the current arrangement as a manageable seasonal squeeze requires one of two things to hold.
The first: the rains come back before the lake is drawn down, the deferred steps are cancelled rather than merely postponed, and the transit cuts are restored in the fourth quarter. Possible — the Canal restored draft in 2024 after the previous event, and the watershed can recover quickly when the pattern breaks. It requires the watershed to recover while the driving event is, on the forecaster's own numbers, still intensifying.
The second: the auction absorbs the whole cost without it reaching anyone else. For a single fixture that is correct — a record slot price is one owner's problem. It weakens as the count grows, because a cargo that cannot buy a slot does not evaporate. It goes around, it waits, or it does not go, and each of those consumes tonne-miles the fleet was not carrying last quarter.
Both are testable, which is more than most market assumptions manage. The trouble is that they are usually held together while nobody tracks whether either is coming true.
The close
The thing to watch is not the next draft number. It is which lever the next advisory pulls. If the Canal defers 47.5 feet again and takes more transits out, it has decided that per-ship cargo capacity is the thing worth protecting and that the queue can carry the cost — and the auction, not the freight index, is where that decision will be priced. If instead it lets the draft step land on 1 October and restores slots, it is protecting throughput and the cost goes back into the cargo hold, quietly, spread across everyone.
And watch the auction categories, which take effect on the same day the slots come out. If the four-group split holds, the Canal has permanently changed who competes for scarcity, and the segment that was clearing the market on 14 August will not be able to do it again in the same way. That is a structural change dressed as a drought measure.
These are opposite trades with opposite winners, and the market has no position on any of them. It is watching a depth number that the Canal has just told it is negotiable.
Sources
- Autoridad del Canal de Panamá — advisories to shipping A-22-2026, A-25-2026 and A-29-2026, and published Neopanamax lock dimensions
- NOAA Climate Prediction Center — ENSO Diagnostic Discussion, 13 August 2026
- Smith et al., “Physical controls and ENSO event influence on weathering in the Panama Canal Watershed,” Scientific Reports, 2020
- World Weather Attribution — analysis of low water levels in the Panama Canal, May 2024
- Trade press reporting of August 2026 auction results, waiting times and rerouted energy flows — gCaptain, The Maritime Executive and Lloyd's List — cross-checked against the Fairway ETA Maritime Data Hub
Related on Fairway ETA
- Bulk Carrier Sizes Explained — the classes that sit in the affected draft band
- SCNT and PC/UMS — how the canal charges, as distinct from what it will accept
- Global Chokepoints 2026 — where Panama sits among the world's constrained passages
- Six Marks on the Hull — why a permitted draft and an actual draft are different numbers