In this article
You negotiate a large crypto OTC deal by treating the price as one term among six: the spread or discount, the size and whether it is filled in full, the timing and execution method, the settlement terms and who goes first, the fees, and what happens if the trade fails. A concession on one of those terms is usually paid for on another, so the work is to decide which terms matter most to you and to trade the others for them.
An over-the-counter (OTC) trade is a trade agreed directly between two parties, away from a public order book. One party is often a desk: a firm that quotes prices and trades for its own account or arranges trades for clients. Because nothing is posted on a screen, every term is open until it is agreed. This article is general information for funds, treasuries, family offices, large holders and the desks and brokers that deal with them; it is not legal, tax or investment advice, and the rules that apply to a desk differ from one country to the next.
Illustrative figures from the worked example in this article: 8,000 units sold at a reference price of $2,500. They are not market data.
What can you negotiate in a crypto OTC deal?
Six terms are negotiable in a crypto OTC deal: the spread, the size, the timing and execution method, the settlement terms, the fees and the consequences of a failed trade. Each one has a cost for the other side, and that cost is what you offer or ask for when you negotiate.
The spread is the distance between the price you are quoted and the reference price, the mid-market price both sides agree to measure against. It is quoted in basis points, hundredths of a percent; a seller sees it as a discount, a buyer as a premium. The other terms decide how much risk the quoting party carries, and a desk prices risk: a seller who asks for a firm price, on the full size, immediately, paid before delivery, is asking the desk to carry every risk at once, and the spread will say so.
The table below sets the six terms out as levers: what you can move, and what the other side needs in return.
| Term | What you can ask | What it costs the other side | What it is worth to you |
|---|---|---|---|
| Spread or discount | A tighter price against a named reference | Margin, and less cushion against a market move while it hedges | Direct money: basis points times notional |
| Size and fill | The full size in one trade, or a firm part and an option on the rest | Inventory risk that grows faster than the size | Certainty that the whole position is done |
| Timing and execution | A firm price now, or a worked order against a benchmark over hours | A firm price puts the market risk on the quoting side; a worked order hands it back to you | A lower cost if you carry the risk, a known result if you do not |
| Settlement and who goes first | Tranches, same-day payment, a neutral third party, your preferred asset or rail | Funding, operational work, credit exposure on you | A smaller unsecured exposure during settlement |
| Fees | All-in pricing, a cap on third-party costs, a clear split of an introducer's fee | Costs it would otherwise pass on | A net figure you can compare across quotes |
| Failure terms | A deadline, a cure period, who bears the cost of unwinding a hedge | An obligation it must honour when something goes wrong | A defined outcome instead of a dispute |
Size is not linear: a desk that can absorb a quarter of your position without moving the market may need a visibly wider spread for the whole of it. And fees only become negotiable once they are visible, so the first request on fees is always a quote that states everything you will pay.
What should you know before you ask for a price?
Before you ask for a price on a large crypto OTC trade, you need three figures of your own: the reference price and the depth of the market around it, your walk-away level, and what waiting costs you. Without them you cannot tell a good quote from a poor one, and the other side will notice.
The reference price is the mid-market price on the venues where the asset really trades, taken at a time and from sources both parties accept. Agree on the reference before you discuss the spread: 30 basis points against one source and 30 against another are two different prices. Depth is how much of the asset can be bought or sold near that price. If your order is several times the size resting within one percent of the mid on the main order books, a quote that looks wide may simply reflect what the market would charge you.
Your walk-away level is the result you can obtain without this counterparty. For a seller, it is usually the estimated net price from working the order yourself on exchanges, after slippage (the price moving against you as your own order consumes the book) and trading fees. In the worked example below, the seller's own estimate is 45 basis points, so any firm offer inside that level is already better than the alternative.
The cost of waiting is the market risk you carry while you negotiate. Take an illustrative asset whose price moves 3% on a typical day: on a $20 million position, that is a swing of $600,000 in either direction. A day spent holding out for 8 more basis points, worth $16,000 on that size, exposes you to a move 37.5 times larger than the prize.
- Fix the referenceChoose the price sources and the method (a mid at a stated time, or an average over a window) you will propose, and check that the counterparty can observe the same sources.
- Measure the depthCompare your size with the volume traded in a day and with the size resting near the mid on the main venues.
- Set the walk-away levelEstimate, in basis points, what the trade would cost if you executed it yourself. That figure is your limit, and it stays private.
- Price the waitMultiply your notional by a typical daily move. Compare the result with the value of the basis points you hope to gain.
- Rank the six termsDecide which two terms you will not give up and which two you are ready to trade. A negotiation without a ranking ends as a negotiation on price alone.
- Clear your own sideConfirm internal approval, signing authority, the wallets or accounts to be used and the documents the counterparty will ask for, so that a good quote can be accepted while it is still valid.
How do you ask for a quote without revealing your direction or size?
You ask for a two-way price on a size smaller than your full position, so that the party quoting does not yet know whether you are a buyer or a seller, or how much you have to trade. A two-way price is a bid and an offer given together: the price at which the desk would buy from you and the price at which it would sell to you.
A desk that knows you must sell a large amount can protect itself by lowering its bid. A two-way price on a standard size gives you a clean reading of the desk's spread before your own need shapes it, and the request is ordinary practice in wholesale markets.
Partial size works the same way. Ask for a price on a quarter of the position first, then ask how the price changes for double or for the whole amount. What you learn is the shape of the price as size grows, which is worth more than any single quote.
Concealment is a tool for the opening, not for the whole trade. A desk that prices a quarter of your position and then discovers three other desks were hit for the rest at the same moment has been given a risk it did not price, and it will remember. Before you execute, disclose the full size to the counterparty you have chosen and let it re-quote.
How many competing quotes should you ask for?
Two or three competing quotes are usually enough on a large crypto OTC trade, because each additional party you ask is another party that knows about your order. Competition tightens the price up to a point, and past that point the information you spread costs more than the tightening saves.
The mechanism has a name in wholesale markets: pre-hedging. The International Organization of Securities Commissions (IOSCO), in its 2025 final report on the subject, describes it as a dealer trading for its own account after learning of an anticipated client trade and before the client has agreed terms, to manage the risk of that trade and with the intent to benefit the client; the report recommends that dealers disclose the practice clearly and limit its market impact. When five desks are asked about the same large sale and each prepares for it, the market can move before you have traded with any of them.
Three practical rules follow. Ask every desk for a quote on the same terms (size, reference, settlement, validity), because quotes on different terms cannot be compared. Ask each desk, before you reveal anything, how it treats the information in a request for quote and whether it trades ahead of a quote being accepted. And say that the request is competitive without naming the others: desks quote more keenly when they know.
A competing quote is also only useful if you would act on it. Showing a desk a rival price you would never accept is a bluff, and it works once.
Is a firm price on the full size worth more than a better price worked over time?
A firm price on the full size is worth more when certainty matters to you more than the last basis points, and a worked order is worth more when you can carry market risk for a few hours. The choice is a transfer of risk: with a firm price the quoting party carries the market until it has hedged, and with a worked order you carry it until the order is done.
A firm price is a price the desk commits to for a stated size and a stated period, often seconds or minutes. A worked order is an instruction the desk executes for you over a period, with the result measured against a benchmark such as the time-weighted average price over that period, less the desk's fee. That fee is lower than the spread on a firm price, because the desk is no longer paid to carry the risk.
The chart below takes the middle offer of the worked example in the next section: instead of a firm price at 32 basis points on the whole $20 million, half is sold firm at 25 basis points and half is worked for a 12 basis point fee, for an expected saving of $27,000. The price of that saving is exposure on $10 million while the order is worked.
A move in your favour of the same size produces the same amounts as gains. Fees and spreads as stated in the worked example; no market data.
The market is as likely to move for you as against you over four hours, so the expected result is the saving. A seller who cannot explain a $100,000 shortfall to an investment committee should buy the firm price and regard the extra spread as an insurance premium; a seller with a long horizon and no deadline can reasonably keep the risk and the saving. Decide which you are before the call, because a desk will ask.
What does a negotiation look like with numbers?
A negotiation on a large OTC trade looks like a sequence of exchanges in which each move changes one term and has a dollar value that both sides can compute. The worked example below is illustrative: every figure is an assumption chosen to show the arithmetic, not a market rate.
The position. A fund wants to sell 8,000 units of a liquid crypto asset. The agreed reference price is $2,500, so the notional is $20,000,000 and one basis point is worth $2,000. The fund's own estimate of selling on exchanges over two days is a cost of 45 basis points, or $90,000: that is its walk-away level, a net price of $2,488.75.
The first offer. A desk quotes a firm price of $2,490.00 on the full size, a discount of 40 basis points, or $80,000. Its terms: the fund delivers all 8,000 units first, and the desk pays by bank transfer on the next business day. The offer is already $10,000 better than the walk-away level, and a seller who had not prepared would stop here.
| Stage | What changes | Cost to the seller | Value of the move |
|---|---|---|---|
| First offer | Firm, full size, 40 bps; seller delivers first, paid next day | $80,000 | Baseline |
| Counter-move 1 | A second quote on identical terms is shown; the desk improves to 32 bps | $64,000 | +$16,000 to the seller |
| Counter-move 2 | Half sold firm at 25 bps, half worked over four hours for a 12 bps fee | $37,000 expected | +$27,000 expected; seller carries the market on $10,000,000 |
| Counter-move 3 | Four tranches of 2,000 units, each paid the same day before the next is delivered; desk adds 2 bps | $41,000 expected | −$4,000 to the seller; exposure cut to one tranche |
| Agreed terms | 20.5 bps blended, expected | $41,000 expected | +$39,000 against the first offer |
Counter-move 1 is pure price: the fund holds a second firm quote at 33 basis points on the same size and terms, says so, and the desk moves to a price of $2,492.00. The concession costs the fund nothing but the effort of having run a comparable second quote.
Counter-move 2 is an exchange. The fund offers the desk something it values, relief from carrying the full size at once: 4,000 units are sold firm at $2,493.75, a cost of $25,000, and 4,000 are worked against the average price of a four-hour window for a fee of $12,000. The desk earns less but carries far less risk, and the fund accepts the exposure shown in the chart above.
Counter-move 3 runs the other way: the fund pays for a term. Delivering the whole position against a payment due the next day is an unsecured exposure of $19.9 million to the desk for most of a day; in tranches of 2,000 units it is about $5,000,000 at any moment. The desk asks 2 basis points on the full notional for same-day funding and the extra operational work. Expected proceeds settle at $19,959,000, against $19,920,000 under the first offer and $19,910,000 at the walk-away level.
The sequence matters as much as the numbers: price first because it was free to ask, the exchange of risk second, and the term the fund paid for last, when it knew how much of its gain it could afford to spend.
How does negotiating with a desk differ from negotiating with a direct counterparty?
With a desk you negotiate against a price it can produce at any time, and with a direct counterparty you negotiate with someone who has the opposite need and no obligation to trade. The desk sells immediacy and a process; the direct counterparty offers the chance of a price nearer the mid, at the cost of doing yourself the work a desk would do.
| Point | With a desk | With a direct counterparty |
|---|---|---|
| Who sets the first price | The desk, as a quote against a reference | Either side; often a discount or premium to an agreed fixing |
| What the other side wants | Margin for the risk it carries, and repeat business | The asset or the cash, on its own timetable |
| Room on the spread | Limited by the cost of hedging the size | Wider: both sides can share the spread a desk would have earned |
| Speed | Minutes to hours once onboarding is complete | Days to weeks; verification and documents start from nothing |
| Settlement terms | A standard process, adjustable at a price | Entirely open, and the heart of the negotiation |
| Typical sticking point | Size and validity of the firm price | Who delivers first, and proof that the other side can perform |
A desk is moved by information that lowers its risk: a clear size, a realistic timetable, a counterparty that settles when it says it will. Between two principals the best lever is usually the reference: trading at the mid of a fixing both can observe leaves both better off than either would be against a two-way price, provided both can settle safely.
Direct trades often involve a third person: the broker or introducer who brought the two sides together. The introducer's fee is a term of the deal, stated as a percentage or an amount, with who pays it and when. Where the fee and the seller's proceeds come out of the same payment, a payment router can divide it. Shaka is one such router: each party's share is set as a percentage and can be signed before the payer pays, the payer pays once, and every party is paid at the same moment, in one final payment that Shaka routes and never holds. It divides one payment among the parties; it does not exchange the two assets of the trade, and it is not a settlement or custody service.
What can you say at each stage of the negotiation?
At each stage of an OTC negotiation you say one thing clearly: what you are asking, on what terms, and until when. Short written messages are better than long calls, because they become the record of what was agreed. The wording below is a starting point to adapt.
- Asking for a quote. "Could you show a two-way price in 2,000 units, against the mid at the time of quote, firm for 30 seconds, all-in, settlement same day? This is a competitive request."
- Asking how size changes the price. "How does that price change for 4,000 and for 8,000, on the same terms?"
- Disclosing before you trade. "My interest is to sell, and the full size is 8,000. Please price the whole amount."
- Countering on price. "I have a firm 33 basis points on the same size and terms. I would prefer to trade with you. Can you improve on it?"
- Offering an exchange. "If I take a firm price on half now and let you work the other half over four hours against the average, what does each half cost?"
- Asking for settlement terms. "Before we fix the price, can you confirm your settlement sequence: who delivers first, in how many tranches, the cut-off times, and what you need from us in advance?"
- Paying for a term. "Same-day payment per tranche matters to us. What would you need on the price to do it?"
- Closing. "Done. To confirm: I sell 4,000 at $2,493.75 firm, and 4,000 worked from now for four hours at the window average less 12 basis points, plus 2 basis points on the full notional for settlement in four tranches paid same day. Please confirm in writing with settlement instructions."
The closing message repeats every term because the confirmation is where misunderstandings surface while they are still cheap.
What should you never negotiate away?
Two things are never traded for a better price on an OTC deal: verification of the counterparty, and written terms. A discount offered in exchange for skipping either one is a price for carrying a risk you cannot measure.
Verification means knowing who you are dealing with: the legal entity, the people authorised to commit it, its registration or licence where one is required, the source of the assets or funds, and control of the wallets and accounts that will be used. Reputable desks ask you for the same things under the customer due diligence obligations that anti-money-laundering rules place on them in most jurisdictions. Ask for a faster onboarding, never for a lighter one.
Written terms mean a confirmation that states the asset, the size, the price or the formula that produces it, the reference, the settlement sequence and addresses, the fees, the deadlines and what happens if either side does not perform. The form varies, from a master agreement with a short trade confirmation to a term sheet for a one-off trade. The terms exist before any asset moves, and settlement instructions are confirmed through a second channel agreed in advance.
Failure terms are negotiable in their detail, such as how long a party has to cure a late delivery, but never in their existence: a failed trade with no agreed consequence is settled by whoever holds the assets at that moment.
How do you keep the relationship for the next trade?
You keep an OTC relationship by being a counterparty whose word can be priced: you trade when you say you will, you disclose what the other side needs to manage its risk, and you settle on time. The price you are shown on the next trade depends on how the last one went.
Desks keep a view of every client, written down or not. A client who asks for quotes on five trades and executes one costs the desk time and, each time it prepared for a trade that did not happen, money. A desk that widens its price for such a client, or for one who spreads a large order across several desks without telling any of them, is pricing its experience, as it should.
Say no quickly and plainly when a quote does not work, so the desk can stand down, and do not let a firm quote expire in silence. Do not reopen terms after "done" unless a fact has changed. Send your side of the settlement at the agreed time and report any delay before the deadline, not after it.
The same applies between two principals, where there may be no next trade but there is always a reputation: a party known for renegotiating at the moment of settlement finds that its next counterparty asks for more protection and offers a worse price.
