Settling in Stablecoins: How High Risk Merchants Run a Treasury Without a Cooperative Bank

The appeal of stablecoin settlement in high risk payments is straightforward. Card revenue clears into USDT or USDC within hours instead of waiting for a correspondent banking chain that can freeze a payout for reasons no one will explain. Merchants who have lost a month of cash flow to a compliance review at an intermediary bank do not need the pitch.

What the pitch usually leaves out is that settlement in stablecoins moves work onto the merchant. A bank account performed custody, conversion, and record keeping as a bundled service. A wallet performs none of them. Merchants who treat a settlement wallet as a bank account discover the difference at the worst moment, either during an audit or during a payout run.

Custody Comes First

The first decision is who holds the keys. A single hot wallet on a founder's laptop is not a treasury. The workable arrangements are a qualified custodian, or a multi signature wallet with signing authority distributed across named people and a documented recovery procedure.

The test to apply is simple and uncomfortable. If the person who set up the wallet is unreachable for two weeks, can the business still pay suppliers and staff? If the answer is no, the custody design is a single point of failure regardless of how much the balance is earning.

Conversion Cadence Beats Conversion Timing

Merchants settling in stablecoins hold a fiat liability set. Payroll, tax, rent, and supplier invoices are denominated in euro, pound, or dollar. Holding operating cash in any asset other than the currency of those obligations is a position, even when the asset is pegged.

The discipline that works is a written conversion policy: a fixed share of each settlement converted to fiat on a fixed schedule, sized against known obligations for the coming quarter, executed without reference to market view. Discretionary conversion turns a payments function into a trading desk, and the merchants who have lost money on stablecoin settlement almost always lost it at the conversion step rather than on the peg.

Off Ramp Redundancy Is the Real Continuity Plan

A stablecoin balance is only as useful as the route from stablecoin to bank account. That route runs through exchanges, over the counter desks, and payment institutions, each of which can suspend a corporate account for the same reasons a bank can.

Serious operators maintain at least two live off ramp relationships in different jurisdictions, keep both funded and tested with small monthly transfers, and document beneficial ownership and source of funds for each before they are needed under pressure. An off ramp that has never been used is an assumption, not a capability.

Network and Counterparty Choices Are Risk Choices

Network selection sets both cost and reach. Tron carries a large share of USDT settlement volume at low fees but narrows the set of institutional counterparties willing to receive it. Ethereum offers the widest institutional acceptance at higher cost. Layer two networks and Solana sit between the two. Most merchants end up supporting more than one, and the deciding factor is usually which network their off ramps and their suppliers accept without questions.

Issuer choice matters for the same reason. USDC and USDT differ in attestation practice, redemption access, and regulatory positioning across markets. Splitting balances across both issuers is the ordinary answer to concentration risk.

Records Regulators and Auditors Will Ask For

A stablecoin treasury still has to produce a defensible audit trail. That means every settlement traced from the card transaction batch to the on chain transfer to the fiat conversion, wallet addresses documented and attributed to the business, screening applied to counterparties, and a stated accounting treatment for the asset agreed with the auditor before the first year end rather than during it.

Merchants who build this record from the beginning describe stablecoin settlement as an operational upgrade. Merchants who reconstruct it two years later describe it as a mistake. The technology is the same in both cases. The bookkeeping is what differs.