All articles

Building the business case for a document pre-check

5 min read

A document pre-check compares your letter-of-credit documents with the credit and the rules before the bank does. Whether it is worth paying for is a finance question, and it deserves a finance answer: your own refusal records, your own costs, and a payback line anyone can recompute. This article sets out how to build that case so it survives a sceptical reader.

The case has to answer one question: is the expected cost of the refusals a pre-check would prevent larger than what the pre-check costs, meaning the subscription plus the internal time spent running it? Everything below serves that comparison. Nothing in it depends on a particular vendor or plan.

Inputs to gather

  • Presentations per year: from your letter-of-credit register or the bank's payment advices.
  • Refused presentations per year: from the refusal notices the bank sent.
  • Bank fees per refusal: from the bank's fee statements or debit advices, including courier and re-presentation charges.
  • Internal hours per refusal: ask the people who do the rework, and use a fully loaded hourly cost.
  • Days of delay: the gap between the first presentation and payment on refused sets, compared with clean ones.
  • Average credit value and a cost-of-capital rate: the second from treasury, for example the rate on the credit line the delay keeps drawn.

Estimating your own refusal rate

  1. 1Collect the refusal notices for the last 12 months. Each refusal arrives as one notice listing every discrepancy, so each notice is one countable record.
  2. 2Count all presentations in the same period.
  3. 3Cross-check the notices against the discrepancy fee lines on your fee statements. A fee without a notice on file may be a refusal that was dealt with by phone and never logged.
  4. 4Count each refused presentation once, however many discrepancies it carried. Your refusal rate is refused presentations divided by presentations.
  5. 5Note how each refusal was resolved: corrected and presented again in time, or paid on the applicant's waiver. The waiver cases are the ones that carried deadline risk.
UCP 600 Art. 16(c)
When a nominated bank acting on its nomination, a confirming bank, if any, or the issuing bank decides to refuse to honour or negotiate, it must give a single notice to that effect to the presenter.

If your own records are thin, a published benchmark can serve as a sanity check, not as a substitute: ICC Academy estimates that 65-75% of documentary-credit presentations are refused on first presentation (ICC Academy, "11 Questions that will help you master documentary credits", updated 8 August 2024). Many of those refusals are resolved, but each still costs rework, fees and delay. A reviewer will trust your own count over any industry figure, so lead with it.

Payback logic

Annual refusal cost = presentations per year x refusal rate x cost per refusal. On DocAccord's editable default model that is 48 presentations x 50% = 24 refusals x EUR 440.14 = EUR 10,563 a year (the cost per refusal is built up line by line in the companion article on the cost of a discrepancy). A pre-check pays back if the share of that cost it prevents exceeds its annual cost, so the break-even share is the annual cost of the check divided by the annual refusal cost. For every EUR 1,000 of annual cost, including internal time, the check has to prevent 1,000 / 10,563 = 9.5% of that cost, about 2.3 of the 24 refusals. Do not take the prevented share from a vendor: run the pre-check on sets your bank refused last year and compare its findings with the refusal notices. The overlap is your own measured share.

What a pre-check does not change

UCP 600 Art. 14(a)
A nominated bank acting on its nomination, a confirming bank, if any, and the issuing bank must examine a presentation to determine, on the basis of the documents alone, whether or not the documents appear on their face to constitute a complying presentation.

The bank still decides. A pre-check can make a refusal less likely; it cannot make payment certain, and a business case that implies otherwise will not survive review. It does not change the issuing bank's credit risk or the country risk, and it cannot cure what has already happened: a shipment made after the latest shipment date stays late. What it changes is when a discrepancy is found, before the documents leave instead of in the bank's notice, and so whether there is still time to correct it. A finding should cite the rule it rests on so that a person can verify it and decide what to change.

The case on one page

  • Current state: presentations, refusals and refusal rate over 12 months, from your own records.
  • Cost per refusal: bank fee, internal hours and delay, each with its source.
  • Annual refusal cost and the break-even share for the check under review.
  • Measured prevented share, from a trial on last year's refused sets.
  • Limits: the bank still decides under Art. 14(a), payment is not guaranteed, and deadline risk is stated but not priced.

Build the case from your own refusal notices and fee statements, compute a break-even share against the full annual cost of the check, and measure the prevented share on your own refused sets. Say plainly that the bank still decides.

All figures here are model assumptions or published estimates, not market data about your business. This article is not legal or financial advice.

This article sets out the rules as we understand them and is not legal advice. What governs is the wording of the specific credit, the applicable ICC rules as in force, and the determination of the examining bank or customs authority in the individual case.

Check it instead of hoping

The demo runs on realistic sample documents. No account, no credit card.