Back to All posts

What Should Be in a Crypto Market Making Agreement? KPIs, Terms, and Exit Clauses

WRITTEN BY
Helen Juhan
Marketing Team Lead at Motion Trade
Helen is Marketing Team Lead at Motion Trade with 4+ years in Web3 and crypto marketing. Before joining Motion Trade, she built and led the marketing function at CLS Global and managed social media campaigns for a portfolio of crypto clients at Ninja Promo. She specializes in turning complex trading products into clear stories.
Visit LinkedIn Profile
A crypto market making agreement should specify, at minimum: maximum bid-ask spread and minimum order book depth as numbers, the venues and trading pairs each target applies to, quoting uptime, the commercial model and its full economics, inventory and capital ownership, real-time reporting rights, conflict-of-interest disclosure, the term and review points, and exit mechanics covering notice, inventory return, option survival, and API key control. An agreement missing the numerical KPIs is not enforceable in any practical sense, however professional the rest of it reads.

Most market making agreements founders sign are drafted by the desk, and desk-drafted agreements are naturally generous about scope and vague about obligation. That is not usually bad faith — it is what any commercial party's first draft looks like. The problem is that founders frequently have no reference point for what a balanced version contains, so the vagueness passes unchallenged and only becomes visible when performance is disputed months later.

This is a clause-by-clause account of what a workable agreement covers. It is written from the issuer's side and is not legal advice; have a lawyer familiar with digital asset contracts review anything before you sign it.

1. Scope: venues, pairs, and what is actually included

The agreement should name every exchange and every trading pair covered, rather than referring to a number of venues. A desk quoting your token on eight exchanges is providing a materially different service from one quoting it on three, and the difference should be legible in the contract rather than assumed. Specify also whether the scope can change, who decides, and what happens commercially if you add a venue mid-term after a new listing.

Where the engagement includes listing support as well as liquidity, keep the two sets of obligations separate in the document. They are different services with different success criteria, and blending them makes both harder to enforce.

2. Liquidity KPIs, expressed as numbers

This is the clause that determines whether everything else in the agreement means anything. Four figures belong here, defined per venue.

  • Maximum bid-ask spread — as a percentage, measured how and over what averaging window. A spread target with no measurement methodology is a target that can be reported as met almost regardless of conditions.
  • Minimum depth — the notional value resting on each side of the book at defined distances from mid price, commonly at 0.5%, 1%, and 2%. Depth is what actually absorbs a sell order, and it is the figure most often omitted from desk-drafted agreements.
  • Quoting uptime — the percentage of time two-sided quotes are live, with the exclusions stated explicitly. Exchange outages are a fair exclusion. Volatility is not, and a desk requesting a volatility carve-out is asking to be absent precisely when you need it.
  • Volume — if included at all, framed as genuine two-sided market activity rather than a target to be hit. Volume commitments have a way of turning into manufactured activity, which is a listing risk you do not want written into your own contract.

Add what happens when targets are missed. Remedies range from fee reduction to a cure period to termination rights, and any of them is better than the common position of no stated consequence at all.

3. Commercial terms and the full cost picture

State the model explicitly: retainer, profit-sharing, loan-and-options, or a combination. Each carries obligations that belong in writing.

For a retainer, cover the fee, the billing cycle, what triggers a change, and whether capital for the book is provided by the desk or the project. For profit-sharing, define how profit is calculated, over what period, and how losses are treated. For a loan structure, the agreement must state the loan size, the term, the option size relative to circulating supply, strike levels, the exercise window, and the return mechanics. We have compared how these structures behave in practice in our breakdown of market making pricing models.

4. Inventory and capital ownership

Establish who owns the tokens and the quote currency in the trading accounts, where they are custodied, and who controls the exchange sub-accounts and API keys. In market-making-as-a-service arrangements the project retains ownership and the desk trades under permissioned keys, which is a materially safer position than transferring inventory outright. Whichever structure applies, the agreement should say so plainly, and API key permissions should be limited to trading rather than withdrawal.

5. Reporting and audit rights

Specify the format, the frequency, and the underlying data. A monthly summary produced by the desk is the weakest acceptable standard and the most common one. What you want stated is continuous access to spread, depth, uptime, inventory, and fill data across every venue in scope, plus the right to request raw trade data for a defined period. Our own approach to real-time client monitoring exists because this clause is where most disputes are ultimately settled.

6. Conflicts of interest and token positions

If the desk holds, will hold, or may acquire a position in your token — through options, an investment, or otherwise — the agreement should disclose it and state the constraints. Reasonable provisions include notice before material disposals, a cap on the proportion of daily volume the desk's own selling may represent, and lock-up or vesting terms on options. A desk that quotes your book and also sells into it without constraint has an incentive structure worth pricing in.

7. Term, review points, and renewal

Twelve months with a formal review at three or six months is a sensible default. The review point matters more than the length: it is the moment when performance against the KPIs is assessed while you still have commercial leverage. Automatic renewal clauses should require affirmative action rather than passive rollover, and any notice period for non-renewal should be short enough to be usable.

8. Exit mechanics

Write the ending while the relationship is good. The clause should cover the notice period for both parties, the timetable for returning loaned tokens and capital, whether options survive termination and on what terms, the transfer or revocation of API keys, the handover of historical performance data, and any obligation on the desk to maintain quoting during a transition to a successor. Add a termination-for-cause provision tied to sustained KPI failure, so that persistent underperformance is a route out rather than a grievance.

9. Compliance, confidentiality, and jurisdiction

Cover the regulatory representations each party makes, restricted jurisdictions, confidentiality in both directions, and the governing law and dispute forum. Digital asset engagements are frequently cross-border, and a dispute clause specifying an inconvenient forum can make an otherwise enforceable agreement practically unenforceable.

A short checklist before signing

Read the agreement once looking only for numbers, and note every obligation expressed as an adjective instead. Read it a second time asking what you would do if the desk simply stopped quoting for a week, and check whether the document gives you a remedy. Read it a third time as though the relationship has ended acrimoniously, and check whether you can retrieve your tokens, your keys, and your data. If all three passes produce clear answers, the agreement is doing its job.

FAQ

What KPIs should be in a crypto market making agreement?

Maximum bid-ask spread as a percentage, minimum order book depth at defined distances from mid price, quoting uptime as a percentage with exclusions stated, and the specific venues and pairs each target applies to. Measurement methodology and averaging windows should be defined alongside the figures, because a target without a measurement method cannot be assessed.

How long should a market making contract run?

Twelve months with a formal review point at three or six months is a common structure. The term needs to be long enough for liquidity quality to be assessed across different market conditions and at least one unlock cycle, and the review point ensures that assessment happens while renegotiation is still possible.

Who should own the tokens in a market making arrangement?

Where the structure permits, the project retaining ownership with the desk trading under permissioned API keys is the safer arrangement, since it avoids transferring inventory to a counterparty. Loan structures necessarily transfer tokens, in which case the loan size, term, collateral or guarantee position, and return mechanics all need to be explicit.

What happens if a market maker misses its KPIs?

Whatever the contract says, which in many desk-drafted agreements is nothing. Workable remedies include a cure period, a proportionate fee reduction, and a termination right triggered by sustained failure across a defined number of reporting periods. Negotiating this clause before signing is considerably easier than invoking goodwill afterwards.

Should a market making agreement include a volume target?

Usually not as a hard commitment. Volume follows genuine demand, and a contractual obligation to produce it creates pressure toward manufactured activity that exchanges detect and penalise. Spread, depth, and uptime are the targets that describe liquidity quality and are within a desk's control.

Can I terminate a market making agreement early?

That depends on the termination clause. Look for a defined notice period, a termination-for-cause provision tied to KPI failure, and clear mechanics for returning inventory and capital. Where options have been granted, check whether they survive termination, because that single detail changes the cost of leaving substantially.

Who controls the exchange accounts and API keys?

The stronger position is for the project to hold the accounts and issue the desk trading-only API keys with withdrawal permissions disabled, so access can be revoked immediately if the relationship ends. The agreement should state who holds what and how keys are handled at termination.

Does the agreement need to address the desk's own token position?

Yes, if a position exists or may arise. Disclosure, notice before material disposals, a cap on the share of daily volume the desk's own selling may represent, and any lock-up or vesting on options are all reasonable provisions, and a desk that resists all of them is worth a second look.

Is a market making agreement the same as a listing agreement?

No. A listing agreement concerns getting a token onto an exchange, while a market making agreement concerns quoting it once it is there. Where one provider delivers both, the obligations should still be documented separately, since they carry different deliverables and different measures of success.

Do I need a lawyer to review a market making agreement?

Yes. Options terms, cross-border regulatory representations, custody arrangements, and dispute forums all carry consequences that are difficult to assess without specialist input, and the cost of a review is trivial relative to the token supply frequently at stake.

August 14, 2026
9 mins