Crypto Market Makers Compared: How to Choose the Right One

Proprietary trading firms with a market making arm trade their own capital across many assets. They bring genuine balance sheet, sophisticated hedging, and strong venue standing, but small mandates compete internally for attention against far larger books.

Designated liquidity providers to token issuers are built around the issuer relationship. Service and reporting are usually better, capital is often smaller, and loan-plus-option structures are more common. Exchange in-house desks are convenient and cheap to onboard but create single-venue dependency. Algorithmic boutiques run lean technology-first operations, price competitively, and vary enormously in quality.

Capital deployed on your pairs and whether it is theirs or yours. Venue coverage and maker tier on each venue that matters. Technology, meaning quoting logic, hedging, and uptime architecture. Reporting quality and whether the data is verifiable. Contract fairness, particularly termination, cure rights, and conduct prohibitions. And attention — where your mandate sits in their revenue stack.

Score each provider one to five on all six, weight them for your situation, and the ranking usually diverges sharply from the ranking by brand recognition.

Strong desks tell you what they will not commit to. They price a low-float token conservatively, explain their hedging, produce data-level reporting without being asked, and accept explicit conduct prohibitions in the contract. They also decline mandates where the float cannot support the spec, which is the clearest quality signal available.

Weak desks lead with volume, resist measurement, present screenshots as reporting, and are unusually flexible on price while unusually rigid on notice periods and audit rights. The pattern is consistent enough to be diagnostic.

A useful practical test is to ask a candidate desk to walk through a hypothetical adverse scenario on your own token — a sudden large sell order, or a venue outage during an unlock — and describe exactly how their systems and people would respond. Strong desks answer with specifics: position limits, escalation contacts, hedging adjustments. Weak desks answer in generalities about their overall track record, which is a different question entirely.

If your mandate is small relative to a desk's book, you will be serviced by process rather than by people, and quoting will be adequate rather than attentive. If your mandate is large relative to a desk's capacity, you are a concentration risk to them and any inventory shock hits your book first.

The comfortable zone is a mandate that is meaningful but not existential to the provider. Ask directly where your fee sits relative to their client base. Desks that answer honestly are usually the ones worth hiring.

At launch, one. Coordination costs, venue conflicts, and the risk of two desks trading against each other outweigh redundancy benefits when volume is small. Once organic volume is established, a second provider on a distinct venue set introduces competitive tension, gives you a benchmark for performance, and removes single-point-of-failure risk on uptime.

If you do run two, define venue boundaries clearly and keep a single reporting standard so the numbers are comparable. It is also worth agreeing in advance how disputes over overlapping flow or a shared venue would be resolved, since retrofitting that agreement after a conflict has already occurred tends to be far more contentious than settling it while both mandates are still being negotiated.

We do not operate a market making book, which means we have no interest in which desk you choose. We write the specification, run the competitive process, benchmark the economics including modelled option value, take the reference calls, negotiate the agreement, and then review the monthly reporting against the schedule you signed.

For most issuers the value is concentrated in two places: the terms you avoid signing, and the month-seven conversation where somebody notices the quoting stopped.

A simple weighted matrix turns a subjective beauty parade into a defensible decision. List the six criteria — capital, venue coverage, technology, reporting, contract fairness, and attention — down one axis and each shortlisted provider across the other. Score every cell one to five based on the RFP response and reference calls, not on the pitch deck alone.

Weight the criteria before you see any scores, not after, otherwise the weighting quietly gets reverse-engineered to justify a favourite. A typical weighting for a small-cap issuer might put attention and contract fairness above raw capital, since a large balance sheet is of little use if your mandate never reaches the top of the queue.

Circulate the completed matrix internally to whoever signs off the mandate, including the numbers you did not like. A matrix that only ever produces the answer leadership already wanted is not doing its job, and keeping the full scoring on file is also useful evidence of process if the choice is ever questioned later.

Send an identical written question set to every provider on the shortlist, at the same time, with the same deadline. Cover capital committed to your pairs specifically, venue and maker-tier coverage, hedging approach for a low-float token, uptime target with defined measurement, reporting format with a sample, and full economic terms including any option or warrant structure.

Ask each provider directly for two reference issuers of comparable size willing to speak, and for one instance where a mandate did not go to plan and how it was handled — a provider that claims a spotless record on this question is not being candid. Score the responses on completeness and specificity as much as on content; vague answers to a precise question are themselves informative.

Quotes are only comparable if every provider priced the same thing. Before requesting economics, issue a written specification covering target spread by venue, minimum depth at defined bands, uptime percentage, reporting cadence and format, and contract term, and require each provider to price against that exact document rather than their own house template.

Convert every response into a single all-in annual cost figure, adding retainer, any token allocation valued at a defensible model, and modelled option value if warrants or loan-plus-option structures are involved. A headline retainer that looks cheapest often ranks poorly once the option component is priced properly, which is precisely why providers prefer to discuss retainer in isolation.

Running two or three providers through the same specification in parallel does more than produce a ranking — it creates genuine negotiating leverage that a single-provider conversation never generates. Providers who know they are being benchmarked against a specific written spec, rather than assessed informally, tend to sharpen both their economics and their contract terms without being asked.

Save the final negotiation for the top two candidates rather than trying to negotiate with everyone simultaneously, and be explicit with both that they are in a final round. This is a normal and expected part of a competitive process, not an aggressive tactic, and reputable desks respond to it professionally rather than taking offence.

Volume guarantees framed as a headline benefit rather than a risk disclosure are a reliable warning sign, since manufactured volume is a conduct issue rather than a service feature. Reporting that consists of dashboard screenshots rather than exportable trade-level data should also raise questions, as should reluctance to name reference clients or to put verbal assurances into the contract.

Unusual flexibility on price paired with unusual rigidity on notice periods, audit rights, or conduct prohibitions is a pattern worth taking seriously — it suggests the commercial terms are designed to be easy to sign and hard to exit. Proposals that do not engage with your specific float and unlock schedule, and instead read as a generic template, indicate the mandate will likely be serviced the same way.

A comparison is only as good as the shortlist feeding it. Three to five providers is usually the right range — enough to see genuine variation in economics and terms without the process collapsing under the weight of reference calls and negotiation threads. Include at least one designated liquidity provider and one algorithmic boutique alongside whichever proprietary desk your investors or advisors have suggested, so the comparison spans archetypes rather than three variations on the same model.

Source candidates from more than one channel — direct outreach, advisor introductions, and reference checks with other issuers of comparable size — rather than relying solely on inbound pitches, which skew toward providers with the most aggressive business development rather than the most disciplined desks. A shortlist built entirely from cold inbound interest is a biased sample before the comparison has even started.

Price comparisons are meaningless until termination notice, cure periods, audit rights, and conduct prohibitions are put on the same footing across proposals. A desk quoting a lower retainer alongside a ninety-day notice period and no audit rights is not actually cheaper than a desk quoting a higher retainer with thirty days' notice and full data access — it has simply moved the cost into terms that are harder to price.

Build a short checklist of non-negotiable terms — notice period, cure right, explicit wash-trading and self-matching prohibition, and data-level reporting access — and require every provider to confirm or reject each one in writing before economics are compared. Only once the terms are aligned does a straight price comparison become meaningful rather than misleading.