Stablecoins for Treasury Teams: A Practical 2026 Guide

The honest answer most treasurers want is whether stablecoins are now a serious tool or still a crypto sideshow. In 2026 the answer has shifted. Stablecoins for treasury are best understood as a fast, programmable settlement rail — genuinely useful for cross-border payment, weekend settlement and reaching markets that correspondent banking serves poorly — and a poor instrument for almost everything else. The technology is no longer the question. Reserve quality, regulatory status and your own controls are.

What changed is regulatory. The United States now has a federal framework for payment stablecoins under the GENIUS Act, signed into law in 2025. The European Union has applied the Markets in Crypto-Assets Regulation (MiCA) in full since 2024, with its specific rules for stablecoin-like tokens already in force. For the first time a corporate or bank treasurer can point to a named statute when asked, in an audit committee, why holding or transacting in a particular token is defensible. That does not make every stablecoin safe. It makes the safe ones identifiable.

What are stablecoins for treasury, and what are they not?

A stablecoin is a digital token, recorded on a public blockchain, designed to hold a constant value against a reference currency — almost always the US dollar, increasingly the euro. The credible ones do this by holding reserves: cash and short-dated government securities held by the issuer, redeemable one-for-one. Tether's USDT and Circle's USDC dominate dollar volumes; PayPal's PYUSD and a growing set of euro tokens sit behind them. These are the only category a treasury should consider.

The distinction that matters operationally is between fiat-backed and algorithmic stablecoins. Fiat-backed tokens are claims on a pool of real reserves. Algorithmic stablecoins try to hold their peg through trading incentives and a sister token, with no hard asset backing. The collapse of TerraUSD in 2022 — which, on the public record, wiped out a very large amount of nominal value over the course of a few days — is the reference event here. No serious treasury function should touch an algorithmic stablecoin under any policy. The category does not belong in this conversation.

What a stablecoin is not is a deposit. There is no deposit insurance, no central-bank backstop and, depending on the issuer and jurisdiction, no guaranteed legal right to par redemption on demand. A well-regulated stablecoin is closer to a claim on a narrow money-market vehicle than to a bank balance. Treating it as cash on the balance sheet — rather than as a short-dated, counterparty-exposed instrument — is the single most common conceptual error, and it is the one auditors will probe first. It also matters how the token sits inside your broader cash hierarchy: a payment stablecoin used in flight to settle a supplier invoice is a transactional balance, not a place to warehouse operating cash. The treasury policy that keeps the two ideas separate is the one that survives a board challenge.

Where stablecoins genuinely earn their place in treasury operations

Cross-border payment is the strongest case. Correspondent banking routes a payment through a chain of intermediary banks, each taking a cut and adding a day. For a corporate paying suppliers across multiple emerging markets, the cumulative friction is real money and unpredictable timing. A stablecoin transfer settles directly between two wallets, on-chain, in seconds to minutes, at a cost measured in cents rather than as a percentage of the principal. For high-frequency, lower-value cross-border flows, the economics are difficult to argue against.

Around-the-clock settlement is the second case, and it is underrated. Conventional payment rails keep banking hours and respect weekends and holidays; a euro payment instructed on Friday evening may not settle until Monday. Blockchains do not close. A treasury managing liquidity across time zones, or settling with a counterparty mid-weekend, can move value when it needs to rather than when the rails reopen. This is the same structural advantage driving institutional interest in stablecoin settlement more broadly. The benefit compounds for firms running thin liquidity buffers: the ability to move dollars at 9pm on a Sunday can mean the difference between a smooth month-end and an overdraft drawn at a punitive intraday rate.

The third is reaching markets that traditional banking serves badly. In jurisdictions with thin correspondent relationships, capital controls or unreliable local rails, paying a supplier or contractor in dollars can be slow, expensive or practically impossible through banks. A dollar stablecoin, with a local off-ramp to convert to fiat, is sometimes the only fast option. The same property that makes this useful — permissionless transfer — is also the one compliance teams must watch most closely, which is why the off-ramp and its KYC discipline matter as much as the on-chain transfer itself.

There is a fourth case worth naming, because it is where serious institutional volume is actually concentrating: intra-group treasury and on-chain settlement of tokenised assets. A multinational moving funds between its own entities, or a desk settling a tokenised money-market fund or repo trade, gains atomic settlement — the cash leg and the asset leg change hands in the same transaction, eliminating the settlement-risk window that conventional plumbing leaves open. This is a narrower, more controlled use than open-market payments, and it is the one where the strongest names in the market — including bank-issued deposit tokens that sit adjacent to stablecoins — are putting their weight. A treasurer evaluating stablecoins should watch this lane closely, because it is where the rail proves itself under institutional scrutiny rather than retail hype.

The risks a treasurer must price: depeg, reserves and counterparty

Begin with the depeg, because it is the risk that turns an operational saving into a loss overnight. A stablecoin is only worth its peg as long as the market believes redemption at par will work. When that belief cracks, the token trades below face value, sometimes sharply, and the discount can persist for hours or days. USDC briefly lost its peg in March 2023 when a portion of Circle's reserves sat in a bank that failed; the token recovered after the deposits were guaranteed, but the episode showed that even the better-run tokens carry tail risk tied to where the reserves are held. A treasury holding stablecoins over a settlement window is exposed to that gap, however briefly.

Reserve quality is therefore the first thing to underwrite. The questions are concrete: what backs the token, who holds those assets, how often is the composition disclosed and audited, and how fast can you redeem for fiat. Tokens backed by cash and short-dated US Treasuries held in segregated, bankruptcy-remote structures, with monthly third-party attestation, sit at the safe end. Tokens with opaque reserves, exposure to commercial paper of uncertain quality, or attestations rather than full audits sit at the other. Under the GENIUS Act, US payment stablecoin issuers face explicit reserve and disclosure requirements; under MiCA, EU issuers face their own. Use those regimes as a filter, not a guarantee — a statute sets a floor, and a treasury still has to do its own diligence above it.

Counterparty risk runs through everything. You are exposed to the issuer (will it honour redemption?), the blockchain (is the network reliable and final?), the wallet or custodian holding your keys (operational and security risk), and any exchange or off-ramp you use to convert. Each is a distinct failure point. A treasury that holds keys directly bears self-custody risk; one that uses a qualified custodian transfers some of that risk but takes on the custodian's solvency and operational standing instead. Neither is free. The discipline is the same as for any short-dated counterparty exposure — concentration limits, name diligence, and a clear redemption path tested before you need it.

One risk gets too little airtime: the irreversibility of an on-chain transfer. A mistyped wallet address or a payment to the wrong chain is, in most cases, unrecoverable — there is no correspondent bank to call and no chargeback mechanism. For a treasury used to the recall and investigation processes of the conventional system, this is a genuine change in operational posture. It argues for address allow-lists, mandatory test transactions before any large transfer, and a four-eyes approval step on payment release. The saving from the rail is real, but it is contingent on a control environment that assumes mistakes cannot be undone.

Accounting and tax: the part that catches treasury teams out

You can build a sound operational case and still trip over how the instrument is booked. Under US GAAP, the Financial Accounting Standards Board issued ASU 2023-08, effective for fiscal years beginning after 15 December 2024, requiring certain crypto assets to be measured at fair value with changes running through net income. That removed the older, punitive impairment-only model for in-scope assets — a genuine improvement — but whether a given fiat-backed stablecoin falls inside that standard or is treated as a financial instrument or receivable depends on its specific terms and your auditor's view. This is not a question to answer informally.

Under IFRS, there is no single dedicated standard, and classification typically turns on the instrument's contractual rights: a stablecoin redeemable for cash may be a financial asset, while others fall under intangible-asset or inventory-style treatment depending on facts and purpose. The practical consequence is that two treasuries holding the same token can carry it differently, and a treasurer should get written classification advice before the first transaction, not after the auditors arrive.

Tax adds another layer. In several jurisdictions, disposing of a stablecoin — including using it to pay a supplier — can be a taxable event, with any difference between cost and value at the moment of transfer potentially crystallising a gain or loss. Even where the peg holds and the economic gain is negligible, the reporting obligation may not be. For high-volume flows this becomes an accounting burden in its own right. The lesson is consistent: model the after-tax, after-accounting cost of a stablecoin payment programme before you assume the on-chain saving is the saving you keep. The systems implication follows directly — your ERP and treasury management system need to capture a cost basis, transfer timestamp and fiat-equivalent value for every movement, or the year-end reconciliation becomes a manual archaeology exercise across block explorers.

The yield temptation, and why a treasury should resist it

Every conversation about stablecoins eventually arrives at the same question from a CFO: if we are holding these tokens anyway, why not earn something on them? The market offers plenty of routes — lending the tokens, depositing them into a protocol, or buying a "yield-bearing" variant that passes through interest on the underlying reserves. The honest counsel is to treat all of these as a different business from treasury operations, and one most corporates have no mandate to enter.

The reason is that yield reintroduces precisely the risks the payment use case was meant to avoid. Lending a stablecoin into a DeFi protocol exposes the treasury to smart-contract risk, to the solvency of borrowers it cannot see, and to liquidity that can evaporate in a stress event — the conditions under which a counterparty most needs its cash are exactly the conditions under which a lending pool freezes. Yield-bearing tokens that distribute reserve interest are a regulatory grey zone by design: the GENIUS Act restricts issuers from paying interest on a compliant payment stablecoin, which pushes any yield offer into a structure that sits outside the clean regulatory perimeter a treasurer was relying on. The moment a token pays you to hold it, you have left the world of settlement plumbing and entered the world of investment products, with all the suitability, classification and risk-committee questions that entails.

There is a defensible middle path for treasuries that genuinely want a return on idle dollars, and it is not a stablecoin at all: a tokenised money-market fund, issued by a regulated asset manager, which is transparent about its holdings and redeems on a known schedule. That is an investment decision made through an investment process, with the right governance attached — not a payment balance quietly repurposed for yield. Keeping the two strictly separate is what lets a treasurer answer the audit-committee question honestly. The programmable-money infrastructure is also what makes B2B embedded finance faster to build, but embedding payments into a product is a deliberate strategic choice, not something a treasury backs into by chasing basis points on its settlement float.

The 2026 regulatory state of play: GENIUS Act and MiCA

Two statutes now define the perimeter, and a treasurer needs to understand both. The US framework is the GENIUS Act, enacted in 2025, which establishes a federal regime for payment stablecoins. It sets out who may issue them, requires full backing by high-quality liquid reserves, mandates disclosure, and draws a firm line: a compliant payment stablecoin is a means of payment, not an investment product, and issuers are restricted from paying interest or yield to holders on the token itself. That last point is the regulatory expression of the stance treasurers should already hold internally — a payment stablecoin is plumbing, not a place to park money for return.

The EU's approach under MiCA is more granular, and the distinction every treasurer operating in Europe must understand is between an e-money token (EMT) and an asset-referenced token (ART). An EMT references a single official currency — a euro or dollar stablecoin is an EMT — and is regulated close to electronic money, with issuers needing the appropriate authorisation and holders generally entitled to redemption at par. An ART references a basket of currencies, commodities or other assets and faces a heavier regime. For practical treasury use the EMT is the relevant category; single-currency, par-redeemable, issued by an authorised entity. If a token you are offered is structured as an ART, that is a signal to slow down and ask why.

The combined effect of these regimes is that the question "is this stablecoin defensible?" now has a documentable answer in the two largest economic blocs. That is the precondition that makes conservative treasury use possible. It also feeds the wider compliance picture — sanctions screening, travel-rule data, transaction monitoring — that any treasury moving value on public chains must address, and which overlaps heavily with the broader RegTech stack a compliance function already runs. One practical caveat: jurisdictional coverage outside the US and EU remains patchy, so a treasury operating in Asia, Latin America or Africa cannot assume an equivalent statute exists, and should treat the absence of one as a reason to hold less and convert faster.

A decision framework: should your treasury hold or transact in stablecoins?

Separate the two decisions, because they carry different risk. Transacting — receiving a stablecoin and converting it to fiat quickly, or sending one and settling immediately — minimises the time you hold the token and therefore your depeg and counterparty exposure. Holding a stablecoin balance over days or weeks is a different proposition: you are now warehousing a counterparty-exposed instrument for convenience or, worse, for yield. The defensible default for 2026 is to transact, not to hold, and never to hold for yield.

Before any programme goes live, four controls should be non-negotiable. First, an approved-token list, restricted to fully-reserved, regulated, par-redeemable tokens from named issuers — typically a major USD EMT and, where relevant, a euro equivalent — with everything else excluded by policy. Second, custody decided deliberately: a qualified custodian for most corporates, with documented key management and segregation, rather than ad hoc self-custody on a treasury laptop. Third, exposure and concentration limits per issuer and per chain, sized as you would size any short-dated counterparty line, plus a tested redemption path so you know you can exit. Fourth, accounting and tax classification agreed in writing with your auditor before the first transaction.

The same diligence instinct that treasuries apply to a new banking relationship should apply to the technology stack behind a stablecoin programme. The decision to settle on-chain sits alongside the wider modernisation of payments infrastructure — the same forces pushing banks toward cloud-native core banking are what make real-time, programmable settlement plausible for a corporate treasury at all. A treasurer piloting stablecoins is not adopting an exotic asset; they are wiring a new rail into an existing payments operation, and it should be governed with the same rigour as any other payment channel — supplier onboarding checks, payment approval workflows, reconciliation, and a kill switch.

The teams that get this right in 2026 will not be the ones that moved fastest into stablecoins, but the ones that wrote the policy before the pilot — the approved-token list, the concentration limits, the convert-by-default posture — and then ran a small live programme through one corridor where the correspondent-banking alternative is genuinely painful. Treat the first six months as a controlled experiment with a hard exposure ceiling, instrument every transfer, and let the reconciliation data tell you whether the saving survives the accounting and tax drag. Scale only the corridors that clear that test. The treasuries that come out ahead will be the ones that proved the rail on a contained set of flows and resisted the temptation to let a settlement tool quietly become a balance-sheet position.