How to Hire a Crypto Market Maker: Selection, Terms and Red Flags

The single biggest determinant of outcome is whether the issuer defined the requirement before speaking to providers. Without a brief, each proposal arrives in a different format with different assumptions, comparison becomes impossible, and the decision defaults to whoever was most persuasive on a call.

A workable brief is one page. It lists the venues in scope and their priority, the target spread and depth per venue, the required uptime, the term, the reporting expectations, and the compensation model the issuer is willing to consider. Sending the same document to every provider forces like-for-like responses and immediately reveals which desks read carefully and which paste a template.

Be explicit about what is not in scope. If the issuer will not consider a token loan, say so in the brief rather than discovering three weeks later that every proposal is built around one.

Three to five providers is the right number. Fewer and there is no pricing tension; more and the process stalls. The shortlist should be tiered to the token: a desk that principally serves major-pair flow will either decline a small mandate or price it as a nuisance, while a small operation cannot credibly cover six venues with institutional depth.

Source candidates from exchange listing teams, from other issuers at comparable size, and from public venue leaderboards where they exist. Exchange business development teams see provider performance across dozens of tokens and will usually give a candid read on who actually maintains their books.

Verify independently before the call. Look at the order books of tokens each desk claims to cover, at a random hour, and see what the spread and depth look like when nobody is watching. That five-minute check disqualifies more candidates than any reference call.

Reduce every proposal to a single comparison table: monthly cash cost, tokens loaned, option strikes and quantities, spread commitment per venue, depth commitment per venue, uptime, term, notice period, and who retains exchange rebates. Proposals that look wildly different in a deck usually converge once expressed this way, and the outliers become obvious.

Price the loan structures properly rather than treating them as free. A loan of tokens with call options attached transfers both float and upside; modelled across plausible price paths it frequently costs multiples of the equivalent retainer. The comparison table should include a modelled cost of the option package, not just the headline cash figure.

Performance obligations must be numeric and per-venue. 'Best efforts to maintain liquidity' is unenforceable. 'Maximum 40 basis point spread with 50,000 USD depth within one per cent, 95 per cent uptime, measured monthly from exchange sub-account data' is a standard the issuer can hold someone to.

Termination is the second priority. Insist on a 30-day notice right without cause, and define precisely what happens on exit: loaned tokens returned within a fixed window, unexercised options cancelled or clearly surviving, sub-accounts closed and final data delivered. Mandates that are difficult to leave are where issuers lose the most.

Then reporting. Monthly, per venue, with spread, depth, uptime, volume excluding the provider's own activity, and any incidents. Sourcing from exchange sub-account exports rather than a provider dashboard removes the argument before it starts.

A proposal that leads with volume figures rather than spread and depth is selling an appearance. Reported volume can be generated between a desk's own accounts; spread and depth cannot be faked to an observer reading the book.

Refusal to provide sub-account level reporting, refusal to accept a 30-day termination right, insistence on a loan structure when the issuer has stablecoins available, option strikes clustered just above the current price, and a term longer than twelve months on a first mandate are each individually a reason to slow down.

So is vagueness about who the counterparty actually is. Ask for the contracting entity, its jurisdiction, and who signs. A desk that will not put a named legal entity on a term sheet is not a counterparty for treasury assets.

A three-month initial term on one or two venues, with a defined review at the end, gives both sides a clean way to test the relationship. Providers that are confident in their performance rarely object; those that push hard for twelve months upfront are usually pricing in the difficulty of retaining clients who can leave.

Use the pilot to establish the reporting rhythm and to build a baseline. At review, the discussion is then about measured performance against the brief rather than impressions, and expansion to further venues can be priced with real data on both sides.

Liquidity mandates fail quietly when nobody inside the issuer owns them. Assign one person responsible for reading the monthly report, checking two or three order-book snapshots independently each month, and raising variances within the notice period rather than at renewal.

That role does not require a trading background. It requires someone who will open the exchange, look at the book, compare it to the contract, and ask a question when the two do not match. Most underperformance persists because nobody looked.