Stablecoin depeg risk: the contingency plan a payment operator needs before the weekend it happens

Stablecoin depeg risk is the risk that a token you accepted at one dollar does not convert back into one dollar. For a bank, a credit union or a licensed money service business, that gap is not academic. It is a loss on a payment you have already promised to deliver. Most institutions settling cross border hold the token for minutes rather than months, and that short window limits the damage. It does not remove it. This article sets out how to measure the exposure, what should stop settlement, and who gets to make that call at two in the morning on a Saturday.

What does it actually mean when a stablecoin loses its peg?

A payment stablecoin trades at par because the issuer promises to redeem one token for one unit of the reference currency. Redemption is the anchor. Market prices follow it. A depeg begins when holders doubt that promise, or doubt how fast it can be honoured. The token then trades below par while redemption queues lengthen. Two different things can break here. The first is the reserve itself, if the assets behind the token are worth less than the tokens outstanding. The second is access to the reserve, when the assets are sound but temporarily unreachable. A token can be fully backed and still trade below par for a weekend. Keep the two apart in your plan. They call for different responses and they last for different lengths of time.

Scale matters as well. A price a fraction of a cent away from par is ordinary market noise. Every traded instrument does that. A depeg worth managing is a sustained, visible discount that holds across more than one venue. Write your own definition down before the day you need it. An undefined trigger becomes an argument.

Why does stablecoin depeg risk matter if you hold the token for minutes?

Short holding periods reduce exposure. They do not make it zero. Count the moments when the value is yours to lose. You hold the token from the moment a sender's funds convert to the moment your payout partner accepts them. You hold it again in any working balance kept on hand to fund payouts before the matching inflow arrives. You hold it in transit between wallets and between chains. Each moment is brief. Hundreds of them overlap. The figure that matters is your peak aggregate balance at any instant, not the average holding time quoted in a sales deck.

Timing works against you too. Stress in the banking system tends to surface late on a Friday or over a holiday. That is when your correspondent is closed, your liquidity provider is thin, and your escalation path runs through one phone number. A plan that assumes a Tuesday morning is not a plan.

What did the March 2023 USDC episode teach payment operators?

The clearest recent lesson came from a bank failure rather than a blockchain failure. Circle disclosed in March 2023 that 3.3 billion dollars of USDC reserves were held at Silicon Valley Bank, which regulators had closed days earlier. The token traded below par across that weekend and returned to par once the deposits were made whole. The chains never faltered. Blocks were produced, transfers confirmed, balances stayed accurate. The stress came entirely from the banking leg behind the token.

Three practical conclusions follow. First, stablecoin depeg risk is mostly a banking and custody question in unfamiliar clothing, so assess the reserve the way you would assess any counterparty. Second, information moves faster than committees. Prices reacted before most compliance teams had opened a file. Third, the firms that came through calmly were the ones holding small balances under a pre agreed pause rule, not the ones with the best analysis.

A depeg is rarely a technology event. It is a liquidity event that arrives wearing a token, and the firms that handle it well are the ones that fixed their limits and their pause rule long before the weekend it happened.

How do you size your real exposure?

Start with measurement, not policy. Pull a week of settlement data and chart your token balance minute by minute. Look for the peak, not the mean. Then split that balance into three buckets.

The first bucket is customer money in flight against a specific payment instruction. The second is your own float, held to smooth payouts and cover timing gaps. The third is anything stranded: failed payouts, unclaimed amounts, small residues on chains you no longer use actively. The third bucket is where exposure quietly accumulates, because nobody owns it. Sweep it on a schedule and the sweep alone will cut your risk more than most policy work.

Now attach a consequence. If the token traded below par for a full business day, what is the loss on your peak balance, and whose profit and loss absorbs it? If the answer takes longer than a sentence, your limits are not yet real.

What should a stablecoin depeg contingency plan contain?

A good plan is short and executable by whoever is on duty, including someone who has never seen a depeg. It names limits, triggers, actions and owners, and it fits on one page. The steps below are the minimum for an institution that settles cross border in stablecoins.

• Set a maximum balance, by token and in aggregate, with a lower ceiling for weekends and public holidays.
• Define the trigger in numbers: a stated discount to par, sustained for a stated number of minutes, visible on two independent venues.
• Name the one role that can halt new settlement in a token, and a named alternate for nights and weekends.
• Write down the substitution path: which token, chain or correspondent you move to, and how long the switch takes in practice.
• Agree the customer and partner message in advance, so nobody drafts disclosure language under pressure.
• Decide now how payments already in flight are treated, and at what value they are delivered or refunded.
• Log every action with a timestamp, because an examiner will ask how the decision was made, not only what it was.
• Rehearse it once a year as a tabletop exercise and keep the notes with the plan.

Notice what is absent from that list. There is no requirement to predict a depeg, and no modelling exercise. The plan exists to convert a fast moving event into a small number of decisions you already made.

Who decides to pause settlement, and how fast can you move?

Speed here is a governance problem before it is a technical one. Decide now which role can stop the institution accepting a token. In most places that authority sits with the treasurer, with the chief compliance officer consulted and the chief executive informed. Put the delegation in writing and hand it to the operations team, because the person on duty at the weekend is rarely the person who wrote the policy.

Then test the mechanics rather than the intention. Can you disable a token in your payment platform without a code change and without your vendor? Can you reach every payout partner inside an hour, on a Sunday? Does one screen show you the live balance by token and by chain? Each of those has a yes or no answer, and finding out during an incident is expensive.

Payments already in flight are the hard case. A customer whose funds are on chain is owed either delivery or a refund, and your plan should say which, and at what value. Redemption rights against a permitted issuer, which the GENIUS Act requires in the United States, belong to the holder of record and take time to exercise. They are a recovery mechanism, not an intraday tool. Do not design your Saturday around them.

How does token choice change the risk you carry?

Diversification helps, but only if it is real. Holding two tokens issued against similar reserves, custodied at similar banks, does little. Look at what genuinely differs: the composition of the reserve, the quality and frequency of the attestation, the redemption terms, and the jurisdictions involved. S&P Global Ratings publishes a Stablecoin Stability Assessment that grades tokens on the strength of that backing, which is a useful external opinion to set beside your own file.

Then be honest about corridors. A payout partner who accepts only one token removes your choice at exactly the moment you want it. Ask each partner, in writing, which alternates they will take and how long a switch needs. That answer belongs in the contingency plan, not in an email thread somebody has to go and find.

Where StableNet fits

StableNet, built by SpendTheBits, is a cross border B2B payment and settlement platform for banks, credit unions, licensed money service businesses, exchange houses and remittance fintechs. Settlement runs in regulated stablecoins such as USDC and USDT on public blockchains and completes in minutes with on chain auditability, which is what keeps the exposure window short enough to manage. The platform is ISO 20022 native, with pacs.008 customer credit transfers, pacs.009 interbank legs, pacs.002 status reports and pacs.004 returns inside head.001 envelopes, tracked end to end by UETR, so a payment held or returned during a market event carries a message an examiner and a counterparty can both read. Compliance is built in, with KYB and KYC onboarding, KYT, sanctions and PEP screening, FATF Travel Rule data in IVMS101 form, a compliance workbench and a tamper evident audit trail. SpendTheBits is a Bank of Canada registered payment service provider and a named finalist in the Swift Hackathon 2026 Technical Challenge.