Moving size without moving your own market.

A visible order book shows the best bid and offer and some depth behind it. What it does not show is how quickly that depth thins, how much of it is quoted by market makers who will pull it the moment a large order appears, and how much of it is the same liquidity displayed on several venues at once.

The result is that an order which looks like a fraction of daily volume can move price several percent while it executes. Slippage of that magnitude on a large conversion is a direct, avoidable loss, and it is compounded when the market can see the order working and trades ahead of it.

An OTC desk solves a different problem to an exchange. Instead of exposing an order to a public book, the desk quotes a single price for the entire size, either on a risk basis where the desk takes the position onto its own book, or on an agency basis where the desk works the order across venues and hands back an achieved average. Both remove the public footprint; they allocate the execution risk differently.

For treasury operations — converting revenue, funding payroll, rebalancing between assets, taking a treasury position into stablecoins, or unwinding one — this is usually not an optimisation but a requirement. The alternative is either accepting material slippage or splitting the order across weeks and accepting market risk instead.

This guide covers how OTC pricing actually works, the difference between principal and agency execution, how settlement and counterparty risk are managed, the compliance file a desk will ask for, how stablecoin treasury and fiat off-ramping fit together, what a corporate crypto treasury policy should contain, and the banking architecture that makes it all operable.

A risk quote is a firm two-way price in your size, valid for a short window — often seconds. The desk is pricing the cost of hedging or warehousing your position: the depth available across the venues it can reach, the volatility of the asset, the size relative to normal turnover, the time of day and the liquidity conditions, and its own inventory. The spread you are quoted embeds all of that plus the desk's margin.

An agency or worked order is priced differently: the desk executes across venues over an agreed period against a benchmark — arrival price, an average over a window, or a volume-weighted benchmark — and charges a commission. You keep the market risk during the working period and you gain from favourable moves.

Which is better depends on the trade. For a conversion that must happen now at a known price, a risk quote transfers exactly the right risk. For a large position where the objective is a good average and there is time, an agency execution against a benchmark usually costs less in total.

Comparing desks on headline spread alone is a mistake. What matters is the all-in achieved price, which means quoting the same size at the same moment to multiple desks and comparing, and — for worked orders — measuring implementation shortfall against arrival price afterwards. Desks expect informed clients to do this.

The other variables worth negotiating: settlement window, whether quotes are firm or indicative, minimum sizes, which assets and pairs the desk covers with real depth rather than nominal coverage, and whether the desk will quote in the fiat currency you actually need rather than one you then have to convert again.

In OTC trading, execution risk is small and settlement risk is where losses actually occur. The mechanism that mitigates it is how the two legs of the trade move relative to each other.

The strongest arrangement is atomic or near-atomic settlement, where neither side has an uncollateralised exposure to the other: settlement through a tri-party custodian, an escrow arrangement, or a venue that holds both legs. The next best is a same-day delivery-versus-payment convention with a defined sequence and confirmations at each step. Weaker arrangements — send first, receive later, on a handshake — are where counterparty failures land.

Where a desk asks you to deliver first, the questions are who the legal counterparty is, whether it is regulated and where, what its balance sheet looks like, whether client assets are segregated, what the settlement window is contractually, and what recourse exists. A desk that cannot answer those clearly is not a desk to send assets to first.

Custody is the other half. Treasury assets should sit with a qualified custodian or in a properly governed multi-signature or MPC arrangement, with defined signing policies, role separation between initiation and approval, transaction whitelisting, and tested recovery procedures. Treasury on an exchange account controlled by one person is the single most common structural failure in crypto companies, and it is also what banks and auditors flag first.

The documentation matters as much as the mechanics: a signed master agreement or terms with the desk, agreed settlement instructions verified out of band, and whitelisted addresses confirmed by a second channel. Address-substitution fraud in the confirmation step is a real and recurring loss event.

Institutional OTC desks are regulated or bank-dependent and run onboarding to a standard close to a bank's. Expect corporate documents and ownership chain, beneficial-owner identification, board or authorised-signatory resolutions, expected trading profile by asset and size, and source-of-funds or source-of-wealth evidence.

For crypto delivered into the desk, expect on-chain provenance review. Funds that have touched mixers, sanctioned addresses, high-risk venues or gambling-related flows will be flagged, and the desk may decline the assets rather than the client. This is worth pre-checking on your own side before delivery, because a rejected deposit can leave assets stranded mid-settlement.

Where the counterparty is a token project, the diligence extends to the token: contract, allocation and vesting, the entity that holds the treasury, and whether the tokens being sold are subject to lock-ups or represent a disclosable insider sale. Desks decline trades that would put them on the wrong side of a market-abuse question.

For fiat legs, the desk's banking partners impose their own requirements — corridors they will not serve, currencies they cannot handle, jurisdictions they will not receive from. This is why the fiat side, not the crypto side, is usually the binding constraint on a treasury operation.

Preparing this file once, properly, shortens onboarding from weeks to days and lets you run relationships with two or three desks rather than one, which is what produces competitive pricing and continuity when one desk is unable to quote.

Holding operating reserves in stablecoins solves genuine problems: settlement in minutes across borders, weekend and holiday operation, and access to counterparties who do not bank conventionally. For a business collecting revenue in digital assets, an intermediate stablecoin layer is often the practical way to hold value while fiat conversion is arranged.

It also introduces exposures that a corporate treasury policy has to name. Issuer credit and reserve composition — what backs the token, where it is held, who audits it, and how redemption works at scale under stress. Regulatory exposure — several major jurisdictions now regulate stablecoin issuance and the permitted use of specific tokens for regulated activities. Chain and bridge exposure — the same nominal token on different networks is not the same risk, and bridged representations carry the bridge's risk.

Concentration is the practical control most companies get wrong. A treasury policy should set limits per issuer, per chain and per custody arrangement, define which tokens are permitted at all, and set a target split between stablecoins and bank-held fiat with a floor of fiat sufficient to cover a defined period of operating expense without any crypto conversion.

Yield deserves particular caution. Stablecoin yield is generated by lending, and lending has counterparty risk regardless of how the product is packaged. Corporate operating reserves are not the appropriate capital for reaching for yield, and the failures of the last cycle were almost entirely companies that treated them as if they were.

The right posture is that stablecoins are a settlement and transit layer with real utility, held deliberately, sized by policy, and paired with a fiat banking capability that works before it is needed.

Converting digital assets to fiat is straightforward. Receiving that fiat into a bank account that will keep it is the difficulty, and it is the reason crypto companies with healthy balance sheets sometimes cannot pay salaries.

Banks decline or freeze incoming crypto-related fiat for predictable reasons: the sending institution is one they will not receive from, the flow does not match the account's stated purpose, the amount is materially outside the pattern they underwrote, the source-of-funds narrative is missing, or the account was opened without disclosing digital-asset activity at all. That last case is the most damaging and the most common.

The architecture that holds up has three properties. It is disclosed — the institution knows the business handles digital assets and underwrote it on that basis. It is layered — a payment institution or crypto-friendly bank receives conversion proceeds and a separate operating bank holds working capital, so a problem at one does not stop payroll. And it is documented — every material inflow has an invoice, contract or trade confirmation behind it, available on request.

Corridors and currencies matter more than founders expect. An institution that handles euro and dollar comfortably may not handle the currency your team is paid in, and correspondent routing for some corridors adds days and questions. This should be tested with small amounts before a large conversion is scheduled.

Finally, timing. Large conversions should be scheduled with the receiving institution informed in advance where the amount is outside the normal pattern. A proactive notification takes minutes; an unexplained large inbound triggers a review that can take weeks.

Approved assets and instruments, with explicit exclusions. Which stablecoins, which chains, which custody arrangements, whether staking or lending is permitted at all, and who may authorise an exception.

Custody and signing: where assets are held, the signing threshold and role separation between initiation, approval and release, address whitelisting procedure with out-of-band confirmation, and a tested recovery procedure with the recovery material held under dual control.

Counterparty framework: approved OTC desks and custodians with limits per counterparty, the diligence required before adding one, and settlement conventions permitted — atomic, tri-party, or delivery-versus-payment with defined sequence.

Liquidity policy: minimum fiat balance in bank accounts expressed in months of operating expense, target allocation between fiat and stablecoins, conversion schedule for incoming digital-asset revenue, and the authority thresholds for discretionary conversions.

Reporting and controls: valuation methodology and source, monthly reconciliation of on-chain balances to the ledger, accounting treatment agreed with your auditor in advance, and the tax treatment of conversions in each jurisdiction where the group operates. Companies that write this policy before they need it pass audits and bank reviews; companies that write it afterwards are usually writing it in response to a problem.

Executing size on a public order book and accepting several percent of slippage as unavoidable. Using a single OTC desk and never comparing quotes at the same moment, which is how above-market spreads persist for years.

Delivering first to an unregulated counterparty with no documented settlement convention. Confirming settlement addresses over a single channel and losing the transfer to an address substitution.

Holding treasury on an exchange account in a founder's name, with no segregation, no signing policy and no recovery procedure. Chasing stablecoin yield with operating reserves.

Opening a bank account without disclosing digital-asset activity, then receiving conversion proceeds into it — the fastest route to a frozen account and a closure that is then disclosable to the next bank.

Running a single banking relationship with no contingency, and scheduling a large conversion without informing the receiving institution. Both are avoidable with a phone call and a second account.

We start with the flow map: where digital-asset revenue arrives, in which assets and sizes, what has to be converted and on what schedule, which currencies the business actually pays out in, where the entities and beneficial owners sit, and what the sector and counterparty profile looks like to an underwriter. Everything else follows from that document.

On execution, we arrange OTC coverage rather than a single relationship — introducing and onboarding you to desks whose asset coverage, size capability, settlement conventions and fiat corridors match your flow, so you can put the same size to more than one desk and compare. We review the master terms and settlement conventions before you sign, and we do not recommend counterparties that require uncollateralised delivery first.

On banking, we build the layered architecture against a network of more than 120 banking and payment institutions whose current appetite for digital-asset flows we track: a disclosed operating bank, conversion-receiving institutions, multi-currency collection and payout capability for the corridors you actually use, and a contingency relationship so a single closure cannot stop payroll. The file is built to institutional standard — entity documents, ownership, trading profile, source of funds, flow-of-funds narrative — and we manage the information requests through to opening. Institutions always decide independently and we say so before starting.

On custody and controls, we help specify the custody arrangement and write the treasury policy: approved assets and issuer and chain limits, signing thresholds and role separation, whitelisting and out-of-band confirmation procedure, counterparty limits, minimum fiat runway, conversion schedule, reconciliation and reporting. Where an auditor or a regulator will review it, it is written to be reviewed.

And we run the connected pieces as one mandate rather than referrals: entity and group structuring so conversions land in the right company, corporate and founder tax treatment coordinated with qualified local counsel, licensing where the activity requires it, and liquidity or exchange access where the business is a token issuer or a trading operation.

Send us the flow — assets in, currencies out, sizes, entities and where the owners are resident — and we will come back with the execution and banking architecture we would build, what it costs, how long it takes, and where the risk sits. This page is general information, not investment, tax or legal advice.