By Alexander Paetzold, COO, Trade Technologies & MD, Trade Technologies Germany GmbH, and Co-written by William Evans, Trade Advisory Board Member

Ask a treasury team when a guarantee expires and most will read out the date on the instrument. That date is real, but on its own it is a poor guide to what the company can still decide and what it is still exposed to. Some instruments carry a review date months earlier that decides whether they roll on, and some carry a claim window past expiry. Our main article on managing guarantees and SBLCs end to end makes the case that operational control is the real risk that surface’s here. Dates are where that plays out: a date is where a decision either gets made or gets made for you.

Which dates actually govern a guarantee?

Not every instrument has the same timeline. A standby letter of credit tied to a specific export sale, with defined amounts and delivery dates, is usually short. Two years or less is common, and there is often no automatic extension at all, because the tenor is already clear from the underlying transaction. The machinery below does not apply.

Longer project work is different. A seven-year plant refurbishment might be supported by a seven-year bank guarantee or standby letter of credit carrying an annual automatic extension inside it. Here three dates matter rather than just one:

• The non-extension notice date, the cutoff before each anniversary by which the issuing bank must give notice if the guarantee or standby letter of credit is not to extend into another period. The applicant triggers this through its bank, and for different reasons, the bank can also initiate it.
• The expiry date, including the final expiry date beyond which no further automatic extension is effective.
• The claim period, typically the number of days, during which a demand can still be presented after expiry.

A guarantee with no fixed expiry date works differently. It ends when the original instrument is returned or the beneficiary, or its bank, confirms release from the obligations under it. That goes smoothly when both parties agree the underlying business was properly completed and the cover is no longer needed. It stalls when they disagree about that business, or when the beneficiary no longer exists after an insolvency or a merger, leaving nobody to return the document or grant the release. URDG 758 adds a three-year backstop for demand guarantees where no expiry is stated (Article 25), but the practical point is simpler: a defined expiry date, plus any claim period, is what makes an instrument sound, because each party knows its legal position on either side of it.

Why do automatic extensions exist at all?

It is tempting to read an automatic extension as a trap. It is not. It is a deliberate allocation of risk serving both the applicant and its bank.

For the applicant, it buys flexibility. A project may finish early, or its scope may change, and an annual decision point is more useful than a single commitment for the full tenor.

For the issuing bank, the value is credit management. Without the extension structure, a seven-year bank guarantee or standby letter of credit is seven years of risk on the bank's books, priced accordingly. With an annual extension, the bank can give notice ahead of the cutoff that it will not continue supporting the instrument, giving the applicant time to move it elsewhere. That is one reason banks are willing to support long project exposure at all.

Under ISP98, this works because Rule 2.06(a) makes an automatic amendment effective without further notification or consent when the standby letter of credit says so expressly, and demand guarantees achieve the same effect through their wording. The wording carries the weight. Sound practice, reflected in the ISP98 model form for extension, pairs the automatic extension with a stated final expiry date beyond which no further extension is effective. Some beneficiaries on major projects will not accept an automatic extension at all, because they want the security in place for the whole project.

What happens when a review date passes without a decision?

For standbys that have auto extensions, the instrument extends, which is often the correct outcome. Banks generally notify applicants of upcoming automatic extensions, so this is rarely a surprise. None of this is a problem, unless the applicant did not want to extend the standby and missed the decision window without using it.

The practical reality is sharper than the legal position. On paper, an applicant that no longer wants the cover can decline the extension. In practice, once the project has run for months or years, withdrawing the security mid-flight is close to impossible without causing serious problems for the project itself. The beneficiary is relying on it. Pulling it is a commercial event, not an administrative one.

So the deadline that matters sits earlier than the notice date. If a project is winding down or its scope has changed, that conversation belongs with the beneficiary well before the cutoff, so any change can be agreed in the normal course of business between the parties and their banks. Cases where an instrument extended and genuinely should not have are rare, and are usually resolved by amendment.

Can a demand still arrive after the expiry date?

In most markets, yes, to cover the postal run. Some guarantees explicitly include a claim period, sometimes called an extended usage period, that runs past the stated expiry and lets a beneficiary present a demand during it. The length varies by country, bank, and wording, from around a month to a year.

There is also the extend or pay demand. Under URDG 758 Article 23, a beneficiary can present a complying demand asking, as an alternative, for the expiry to be extended, and the guarantor may suspend payment for up to 30 calendar days.

So an instrument at or just past its expiry date is not automatically closed. Treating expiry as the moment a line drops off the exposure register is not accurate.  Banks add the additional notice periods to the final expiry dates in their systems, but applicants may not always understand this short but important additional final life of the standby.

Where does the amendment risk actually sit?

Not in the instrument changing behind your back. Amendments to amount, expiry, or wording happen by agreement between the parties, and the final expiry date is the guidepost everything else sits inside. If both banks and both companies agreed to a change, nobody has been ambushed.

The risk sits in your record of it. An amendment agreed in one place and recorded in another, or recorded late, means the version in your system stops matching the version in force. That is the pattern described in common operational failures: the record drifts just a bit from reality, and nothing forces the two back together until a demand, an audit, or a credit review does it for you.

What does this mean as the portfolio grows?

One instrument with three dates is easy. Several hundred, spread across banks, currencies, and countries, is not, at least not on a spreadsheet built by someone who has since changed roles. The critical dates sit years out, and a spreadsheet cannot reconcile itself against the banks' records.

What a corporate needs is a current view of the whole portfolio by expiry date, country, currency, and counterparty exposure, with review dates visible alongside expiries. Banks track their own exposure for credit purposes, but no single bank sees the whole portfolio, so the consolidated picture is the corporate's to hold.

The aim is not to stop instruments from extending. Most extensions are correct and should happen. The aim is that every extension, lapse, or amendment is something the company chose, early enough for the choice to be real, rather than something it found out about later.