Stablecoin growth is no longer only a payments, digital-assets or compliance question. New 2026 evidence shows that it can alter the quality, location and intraday behaviour of bank funding. Treasury leaders need a flow-based control model before adoption becomes material.
The finance question is where the money lands
Stablecoins are often discussed through the customer interface: faster settlement, programmable transfers and access to digital markets. A bank balance sheet sees a different picture. When a household or company exchanges a conventional deposit for a stablecoin, the deposit may remain somewhere in the banking system, move to another institution, change from retail to wholesale funding, or leave the system's commercial-bank reserve base. Those outcomes are economically different even when the customer experiences one simple conversion.
That distinction is the core finance issue for 2026. The BIS Annual Economic Report 2026 says wider stablecoin adoption could change bank funding and credit provision, with the result depending on reserve composition, use and regulation. The practical implication is not that every bank should assume a deposit flight. It is that treasury cannot model stablecoins as a single product balance. It must model the complete money path.
This report sets out an operating approach for finance, treasury, asset-liability management and business leaders. It treats stablecoin activity as a set of balance-sheet states and payment flows, connects those states to liquidity metrics and pricing, and turns uncertain adoption into measurable management decisions. It is a control framework, not a prediction about the winning form of digital money.
What the 2026 evidence changes
The BIS's June 2026 assessment uses stylised scenarios to show why reserve design matters. If an issuer holds bank deposits, retail deposits can be replaced by a concentrated wholesale balance. If the issuer holds government bills, banks may also lose high-quality liquid assets. If an issuer can hold central-bank reserves, household conversion can remove both retail deposits and central-bank reserves from commercial banks. Each route can weaken liquidity, but through a different mechanism.
New empirical research adds an intraday dimension. A February 2026 Federal Reserve Bank of New York staff report links on-chain activity with wholesale interbank payments. It finds that banks serving stablecoin issuers face higher payment demand and greater sensitivity to primary-market activity. The authors also find that partner banks hold more reserves and that their loan share contracts relative to peers. This is research evidence, not a universal forecast, but it shows why end-of-day balances alone can miss the operating burden.
Together, the sources point to three separate treasury questions:
What funding changes when customers acquire or redeem stablecoins?
Which institution receives the resulting balances and in what regulatory category?
When do the gross payment flows occur, and how large can they become inside the day?
A bank that answers only the first question is measuring product volume. A bank that answers all three is managing liquidity risk.
Build the stablecoin flow map
Start with economic roles, not product names
The same legal entity can be a customer, reserve bank, settlement bank, distributor or liquidity provider. Treasury should therefore build an activity-level map covering minting, redemption, exchange, safeguarding, reserve investment, market making and payment settlement. For each activity, the map should identify the initiating party, account owner, currency, legal claim, settlement rail, expected timing and balance-sheet destination.
This avoids a common classification error: treating an issuer's large deposit as a replacement for thousands of granular retail balances. Both are deposits, but they differ in concentration, pricing sensitivity, operational volatility and regulatory treatment. The Basel liquidity principles require banks to meet daily obligations and withstand institution-specific and market-wide stress. That principle supports a behavioural view of the funding, not a label-based one.
Define three reserve-path scenarios
The BIS scenarios provide a useful minimum set for management modelling.
Bank-deposit reserve path. Customer deposits move to the issuer and are redeposited with one or more banks. System-wide deposits may be broadly unchanged, but the recipient banks hold more concentrated wholesale funding while other banks lose retail balances. The scenario must distinguish system effects from the bank's own position. A bank can be neutral in aggregate data and still face a material franchise loss or a large issuer concentration.
Government-bill reserve path. The issuer uses incoming deposits to buy short-term government securities. Sellers may redeposit proceeds, but banks that sell bills lose assets that may otherwise support their liquidity buffers. The model should capture both liability migration and any change in high-quality liquid assets, including settlement timing.
Central-bank reserve path. Where the institutional design permits an issuer to hold central-bank money, conversion can reduce commercial-bank deposits and reserves together. The effect is more direct and should be modelled as a distinct structural scenario, not as a sensitivity around the deposit-reserve case.
Banks should add jurisdictional and business-model variations, but these three paths create a common language for the board, treasury, risk and product teams.
Connect the map to the bank's liquidity stack
Measure regulatory and economic liquidity together
The Basel Liquidity Coverage Ratio is designed to ensure enough high-quality liquid assets for a 30-day stress. The Net Stable Funding Ratio addresses the stability of funding over a one-year horizon. Stablecoin scenarios can affect both, but regulatory ratios should not become the entire dashboard.
Treasury should calculate four views for each scenario: regulatory LCR and NSFR; internal survival horizon; intraday minimum liquidity; and the marginal cost of restoring the target buffer. That final measure converts a ratio movement into a business decision. If a new stablecoin service consumes scarce reserve capacity or requires more contingent funding, its economics should include that cost.
The model also needs distributional detail. Segment expected conversions by customer type, channel, currency, geography and time zone. Identify whether funds are likely to return to the same bank, another bank or a securities account. Assign behavioural assumptions to ordinary days, high-volume events and redemption stress. Keep management overlays visible rather than burying them inside a single coefficient.
Add intraday stress to the scenario engine
Daily closing balances can hide gross movements. A bank may finish with an acceptable reserve position after processing a sequence of large mints and redemptions, yet experience severe liquidity demand during peak windows. The New York Fed paper's evidence makes an explicit intraday layer essential for banks that serve issuers or provide settlement.
At minimum, test concentrated minting, concentrated redemption, simultaneous customer outflows, delayed incoming payments, a settlement-rail interruption and the loss of an expected funding source. Measure peak cumulative outflow, lowest projected reserve balance, collateral usage, payment-queue pressure and time to recover the operating buffer. Assign action thresholds before the test is run.
This is also where operations and finance must share one clock. Treasury needs live or near-live visibility into issuer account movements; operations needs pre-agreed escalation points; the business needs cut-off and capacity rules; and management needs authority to restrict activity if liquidity crosses a defined floor.
Price the service for the balance sheet it uses
Stablecoin-related revenue can arrive through transaction fees, reserve services, custody, foreign exchange, distribution or client retention. The cost side is less visible. It may include additional reserves, committed facilities, collateral, operational coverage outside conventional hours, concentration limits, compliance capacity and the opportunity cost of a more liquid asset mix.
A credible product contribution model should therefore charge for four resources: structural funding consumed, intraday liquidity consumed, contingent liquidity reserved and operational resilience required. It should also recognise franchise effects. If customers move low-cost operating deposits into a coin while the bank receives a smaller share back as rate-sensitive issuer funding, nominal deposit totals can conceal a loss of value.
Finance should review profitability by cohort rather than by total volume. Separate customers whose activity creates two-way flows from those that consistently remove deposits. Separate issuer-service revenue from the economics of the bank's own distribution channel. Show downside economics under a higher reserve requirement, a lower redeposit share and a more concentrated redemption pattern.
The decision rule is simple: growth should be approved against risk-adjusted balance-sheet value, not transaction count.
Put control ownership around the flow
Create a stablecoin funding passport
For each material service or partnership, maintain a compact funding passport that records the end-to-end flow, relevant accounts, reserve path, customer segments, contractual rights, liquidity assumptions, data feeds, limits and contingency actions. Name an executive owner and a model owner. Review the passport when the issuer changes reserve policy, a new settlement rail is introduced, or customer eligibility expands.
The passport should also identify evidence dependencies. Treasury must know whether it receives timely reserve-composition information, whether account balances are segregated, what redemption terms apply and how quickly liquidity can move across entities. Where evidence is unavailable, the scenario should become more conservative rather than more precise-looking.
Establish early-warning indicators
A useful dashboard combines slow structural measures with fast flow measures. Track retail-to-stablecoin conversion by segment; stablecoin-related wholesale funding concentration; issuer deposit volatility; gross mints and redemptions; intraday reserve range; payment peaks; LCR and NSFR sensitivities; collateral utilisation; funding restoration cost; and product return after liquidity charges.
Indicators need thresholds linked to decisions. A rising conversion rate might trigger repricing or a deposit-retention action. Higher intraday volatility might trigger additional prefunding. Deteriorating evidence quality might reduce a limit. A concentrated issuer balance might require diversification or a capped exposure. A dashboard without actions is reporting, not control.
Pre-authorise contingency actions
The contingency funding plan should incorporate stablecoin-specific events: a redemption surge, an issuer transferring reserves, loss of a payment rail, delayed collateral mobilisation or a rapid customer shift following market news. For each event, set the order of liquidity sources, responsible decision maker, communication path and recovery target.
Pre-authorisation matters because digital-market activity may continue outside the bank's traditional treasury window. A plan that depends on assembling a committee after the liquidity threshold is breached is not an operational plan.
A 90-day management agenda
Days 1-30: establish exposure. Inventory current issuer, exchange, custody, distribution and settlement relationships. Trace customer conversion and redemption paths. Reconcile product data to account and payment data. Identify the top uncertainties and assign conservative interim assumptions.
Days 31-60: quantify impact. Build the three reserve-path scenarios, segment behavioural assumptions and add intraday stresses. Calculate LCR, NSFR, survival-horizon and profitability effects. Set provisional limits and data-quality requirements.
Days 61-90: embed decisions. Approve the funding passports, pricing charges, dashboard thresholds and contingency actions. Run a joint treasury-operations exercise that includes an out-of-hours redemption surge and delayed inflow. Present the results as choices: capacity, economics, controls and residual uncertainty.
The deliverable is not a perfect forecast of adoption. It is a repeatable system that can absorb new evidence without redesigning governance each time the market changes.
The strategic conclusion
Stablecoin activity can redistribute deposits, change their behavioural quality and generate sizeable intraday payment demand. The effect on an individual bank depends on its role, customers, counterparties and reserve path. That uncertainty is a reason to build better measurement, not to wait for a single industry answer.
Banks that connect digital-product data to funding, liquidity and pricing will be able to distinguish attractive services from balance-sheet-intensive volume. They will also be better placed to manage customer choice without treating every conversion as a crisis. The finance function's job is to make the movement of money legible, costed and governable.
FAQ: Stablecoin bank liquidity risk
Do stablecoins always reduce total bank deposits?
No. If an issuer redeposits reserves with banks, deposits may remain in the system while shifting from granular retail balances to more concentrated wholesale funding. Other reserve choices can produce different outcomes, so the bank must trace the full path.
Why can liquidity weaken if deposits stay in the banking system?
Funding quality and location matter. Concentrated or rate-sensitive wholesale deposits may be less stable than retail deposits, and the receiving bank may differ from the bank that lost the customer balance. Regulatory and internal liquidity measures can therefore deteriorate even if system-wide deposits look unchanged.
What should banks measure beyond LCR and NSFR?
They should measure internal survival horizon, intraday minimum liquidity, peak cumulative outflow, collateral usage, payment-queue pressure, concentration and the marginal cost of restoring management buffers.
Are stablecoin issuer deposits attractive funding?
They can support revenue and balances, but their value depends on concentration, volatility, contractual terms, segregation, pricing and operational demands. Banks should assess them as a distinct funding cohort rather than assume they replace retail deposits one-for-one.
What is the first practical step for a bank with limited exposure?
Create a flow map covering customer acquisition and redemption, reserve destination, settlement timing and account ownership. Even a low-volume inventory will reveal data gaps and clarify which scenarios deserve quantitative testing.
