Why the PDA Is Already Wrong When the Ship Arrives

A proforma disbursement account is a price for something that has not happened yet. You take the vessel, the berth, the services the call will need, and the rates you agreed months ago, and you hand the owner a number. Everything in that number rests on one assumption: that the call you are pricing is the call that will arrive.

That assumption used to be reasonable. It is not any more, and the industry has the figures to say so.

The date the whole estimate hangs on

Liner schedule reliability now runs at roughly 60 to 65 per cent, meaning about two thirds of container vessels arrive when they said they would. Before the pandemic that figure sat between 70 and 80 per cent. When a ship is late, it is late by about five to five and a half days, where the old normal was three to four, according to reporting by The Loadstar on the schedule reliability data.

Congestion is doing the same thing from the other direction. Container capacity stuck waiting has reached 4.31 million TEU, past the previous peak set during the 2022 crunch, per The Maritime Executive. Persistent delays are effectively taking around 1.7 million TEU of capacity out of the market, as Splash 24/7 put it, a fleet the size of a major carrier simply parked.

Put those together and the picture for an agent is simple. The single input your estimate depends on most is the one you can trust least.

The average delay is the least useful number you have

Industry averages get quoted a lot, and they hide the thing that matters. Port operational updates for the week of 12 to 18 August 2026, published by Kuehne+Nagel, put average vessel waiting time at about 1.47 days in Rotterdam, 2.0 in Barcelona and 2.11 in La Spezia, while Buenaventura sat at zero.

Same week, same industry, and the spread between the calmest port and the worst is not a rounding difference. A buffer built for one of those numbers is far too much for a quiet port and nowhere near enough for a congested one. If your estimate carries a single contingency figure across your whole portfolio, it is wrong in both directions at once, and only one of those directions ever gets noticed.

One date moves and everything behind it moves too

An arrival date is not a single fact. It is the peg that a dozen other commitments hang from, and they do not all move together when it slips.

The berth window shifts, and with it the pilot and the tugs. Launch trips get rebooked. Crew flights, arranged around a sign-on date, either get changed at a fee or turn into hotel nights. Anything charged per day, a gangway watchman, a launch on standby, quietly runs longer than the estimate assumed. The provisions ordered for one date arrive for another.

None of those line items is dramatic on its own. Together they are the difference between the proforma and the final, and they accumulate in a place nobody is looking until the invoice goes out and the owner asks a question.

What actually separates a PDA from an FDA

The textbook answer is that a proforma disbursement account is the estimate sent before the call and the final disbursement account is the reckoning sent after it. True, and not the useful part.

The useful part is that the gap between them is information. A PDA that came in ten per cent under is telling you something about that port, that berth, that season, or that assumption. It is the closest thing an agency has to a feedback loop on its own estimating.

Most agencies throw that information away, not out of carelessness but because of where the two documents live. The PDA is a file built for the owner. The FDA is assembled later from invoices, emails and someone's memory of what actually happened. By the time anyone could compare them, they are two unrelated documents that happen to concern the same ship, and comparing them by hand is a job nobody has time for on a Friday.

So the gap gets absorbed. The variance is explained once to the owner, and the lesson it carried is not available the next time that vessel calls at that port.

What you can actually do about it

Not forecast better. Nobody is going to out-predict a 60 per cent reliability rate, and any vendor promising you a more accurate estimate is selling you a better guess about a number that is not knowable.

What is available is traceability. If the proforma and the final are built from the same port call rather than assembled separately, then the difference between them is a fact you can look at instead of a discrepancy you have to reconstruct. The estimate stops being a document you sent and becomes a position you can compare against.

Three things follow from that, and none of them require predicting anything:

Where PortFlow sits, and where it does not

PortFlow builds the proforma from the port call and issues the final against that same call, so the two are not parallel documents that have to be reconciled by hand later. That is the part we will stand behind, and it is the part this whole argument turns on.

What PortFlow does not do is variance reconciliation line by line against your suppliers' invoices. That is a real and separate piece of work, and we would rather tell you it is not here than let the paragraph above imply it. If line-level supplier reconciliation is what you are shopping for, this is not yet the tool for it.

For the wider picture of how the call moves from arrival notice to invoice, we wrote about that in our guide to streamlining port operations.

One agency measured its own quotation preparation dropping from about 35 minutes to under 10. That is one agency and one task. It is not an average and it is not a promise about yours.