How Much Does a Crypto Market Maker Cost?

The first layer is cash: a monthly retainer, sometimes with a setup fee. The second is float: tokens you lend the provider as working inventory, which are out of your control for the term and which, in aggregate, represent supply that can reach the market. The third is optionality: call options granted over the loaned tranche, which transfer real economic value if the token performs.

Founders routinely compare only the first layer. A provider quoting a low retainer alongside a large loan and aggressive strikes can be several times more expensive than a higher cash quote with no options, once you value what you have given away.

Retainer pricing scales with venue count, depth commitment, pair count, and how demanding your uptime and spread targets are. A single-venue mandate with modest depth on a low-volatility pair sits at the bottom of the range; multi-venue coverage with tight spreads and deep books on a volatile low-float asset sits well above it. Providers also price the operational load — more venues means more reconciliation, more reporting, more incident handling.

Because ranges are wide and confidential, the only reliable way to know whether a number is fair is to run a competitive process on one written specification. Three quotes on identical depth, spread, uptime, and venue terms will cluster; if one is an outlier, the difference is usually in what it excluded.

To value a loan mandate, model it. Take the loaned quantity, the strike ladder, and the expiry, and calculate what the provider receives at several price outcomes — flat, up fifty percent, up three hundred percent. Then compare that against the retainer you would otherwise pay over the same period. Founders are frequently surprised: in a strong scenario the option value can dwarf a year of cash fees.

Also model the supply effect. Exercised options mean tokens leaving your treasury into the market at defined prices, which can act as a ceiling. If those strikes cluster near a psychologically important level, you have effectively pre-sold your first rally.

Exchange trading fees on the pairs being quoted, and who bears them. Stablecoin working capital you must supply alongside the token tranche, which is real treasury locked up. Custody and transfer costs. Legal review of the agreement. Internal time for reconciliation and reporting review. And the opportunity cost of a mandate you cannot exit because the notice period runs to ninety days.

Then there is the cost of a badly specified deal: paying for depth you never verified, on venues that never mattered, while the pair that carries your volume was out of scope.

Write one specification and send it unchanged to every provider. It should state the venues, pairs, target maximum time-weighted spread, minimum depth at ±0.5%, ±1% and ±2%, minimum uptime, reporting format and frequency, term, and notice period. Ask each provider to price both a pure retainer and a hybrid so you can see how they value the option component.

Then normalise. Convert every quote into an annual all-in figure that includes cash, the modelled value of options at a reasonable price scenario, and any exchange fees you would carry. Compare that number, not the retainer.

Narrow the venue list to the pairs that matter. Relax spread targets that your float cannot support anyway — paying for a one percent spread on an illiquid asset is paying for a promise nobody can keep. Shorten the initial term to three or six months with an extension, so pricing resets once the provider has real data. And ask for a fee step-down tied to organic volume growth, so the mandate gets cheaper as the book becomes self-sustaining.

Finally, competitive tension. Providers price differently when they know the specification went to three other desks.

A growing number of retainer agreements include a step-down clause: the fee reduces after a defined number of months if organic volume and depth have grown past a threshold, on the logic that a self-sustaining book requires less active support from the provider. This is worth requesting explicitly, since providers rarely offer it unprompted.

The reverse also exists and is worth watching for — step-up clauses that increase the fee if depth targets are raised later, or if a new venue is added. These are legitimate provided they are tied to a genuinely expanded scope rather than used to claw back an initially discounted quote once the mandate is underway.

Ask for the full fee schedule including any setup cost, not just the headline monthly figure. Ask whether the quoted depth and spread numbers assume calm markets or are meant to hold through defined volatility, since a spread target that only survives on quiet days is not worth much. Ask what proportion of the quote is fixed versus contingent on a token loan, and if a loan is involved, ask for the strike ladder and expiry in writing before agreeing to the headline number.

Ask, too, what the provider's cost structure looks like on its side — how many venues it is quoting across for other clients with the same infrastructure, since a desk with genuine operating leverage should be able to price more competitively than one building bespoke infrastructure for a single mandate.

Two providers quote the same specification. Provider A proposes a retainer at a moderate monthly rate with clearly defined KPIs and no token loan. Provider B proposes a lower headline retainer plus a 1% supply loan with three strikes between 75% and 250% above the reference price. On paper Provider B looks cheaper every month.

Modelled over twelve months at a moderate price appreciation scenario, Provider B's option value alone can exceed the entire annual cost of Provider A's retainer, before accounting for the float that was locked up and unavailable to the treasury for other purposes. Neither structure is wrong in isolation, but comparing only the monthly invoice would have led to the wrong decision.

A full invoice, whether explicit or implicit, tends to break into: the base retainer or its option-value equivalent; a setup or integration fee for sub-account and API provisioning, usually charged once; exchange trading fees on the quoted pairs, which may sit with the provider or be passed through depending on the agreement; stablecoin working capital supplied to fund the bid side, which is real treasury even if it is never spent; and legal review costs for the agreement itself, particularly where a token loan is involved.

Reporting and monitoring tooling is sometimes bundled and sometimes billed separately for multi-venue mandates with bespoke dashboards. Ask for each of these as a separate line even where a provider prefers to present a single bundled number, since bundling makes it materially harder to compare one quote against another on equal terms.

Very early-stage tokens with a small circulating float and limited treasury cash typically cannot support a large cash retainer and gravitate toward loan or hybrid structures almost by necessity — the question then becomes how tightly the strike ladder and tranche size can be negotiated rather than whether cash is affordable at all. Mid-stage projects with a funded treasury and a tier-one listing in view should budget for a genuine multi-venue retainer, sized to depth and pair count rather than to headline market capitalisation.

Later-stage or already-listed tokens with substantial organic volume sometimes find that a much smaller mandate, focused on maintaining an already-tight book during quiet periods, is sufficient, and can negotiate a fee step-down against that history. In every case, budget against the specification you intend to write, not against a market capitalisation benchmark someone else quotes informally — depth commitments and venue count drive cost far more directly than the size of the token itself.

Volatility is the single largest driver: a desk quoting a thinly traded, high-volatility asset carries more inventory risk per unit of depth than one quoting a stable, liquid pair, and prices accordingly. Tight spread targets on an illiquid float compound this, since the provider must hold a wider inventory buffer to defend a narrow quote through sharp moves.

Multi-venue coverage, especially across time zones and API architectures that do not share infrastructure, adds real operational cost that shows up in the retainer. Aggressive uptime requirements, twenty-four-hour coverage with strict penalty clauses, and demanding reporting cadences all add cost too, as does a short initial term, since providers price in the risk of being unable to recover setup costs over a brief engagement.

A longer initial term lets a provider amortise setup costs and generally earns a better rate than a short trial. A narrower, honestly scoped venue and pair list — quoting only where volume genuinely exists rather than everywhere the token happens to be listed — reduces operational load and cost together. Realistic spread and depth targets, set against the actual float rather than an aspirational number, avoid paying for a promise the market cannot support.

Competitive tension from a genuine multi-provider process reliably brings pricing down, as does a track record of clean, organic volume that reduces the provider's perceived inventory risk. A step-down clause tied to measured performance is a further way to ensure cost falls as the mandate matures rather than staying fixed regardless of outcomes.

Withdrawal or transfer fees when moving working capital between the treasury and exchange sub-accounts, which are small individually but recur monthly across multiple venues. The cost of internal time spent reconciling the provider's reporting against on-chain and exchange data, particularly in the first few months before a routine is established. Currency conversion or slippage incurred when funding stablecoin working capital from a different asset.

And, less obviously, the cost of a mandate that under-delivers but sits inside a long notice period — every month spent renegotiating rather than replacing a poor provider is a month of thin liquidity that carries its own cost in lost trust with exchanges and holders, even though it never appears as a line on any invoice.