Crypto Market Making Explained: The Complete 2026 Guide for Token Issuers.

Strip away the mystique and a market maker is a firm running automated systems that place and manage limit orders on both sides of an order book, continuously, across every venue where an asset trades.

At any moment, the MM's system maintains a ladder of buy orders below the current price and sell orders above it. When a trader sells into the book, the MM's bids absorb the sale and the MM now holds more of the token. When a trader buys, the MM's asks fill the purchase and the MM holds less. The MM earns the spread between those two prices, repeatedly, thousands of times a day, while trying very hard to end each period holding roughly the inventory it started with.

That last clause is the entire skill of the business. Collecting spread is easy in a flat market. The hard part is inventory risk: the token the MM accumulates while prices fall is worth less by the time it can be sold, and the token it sheds while prices rise was sold too cheap. Professional desks manage this with hedging where hedge instruments exist, dynamic quote skewing, and sophisticated models of short-term flow.

Beyond the core quoting loop, a competent desk also delivers multi-venue consistency (keeping prices aligned across exchanges), event coverage (maintaining orderly markets through listings, unlocks and announcements), and honest market intelligence about flow composition and venue quality.

What a market maker does not do, despite widespread founder fantasy, is make your price go up. A legitimate MM is directionally neutral: it sells into strength and buys into weakness by construction, because that is what quoting both sides means.

Bid-ask spread. The gap between the best buy order and the best sell order, usually in basis points relative to mid. A major pair might trade at 1–5 bps. A healthy mid-cap altcoin sits at 10–30 bps. A neglected small cap can sit at 100–300 bps, meaning a round trip costs a trader 1%–3% before fees.

Depth. How much can actually be traded near mid, quoted as the dollar value of resting orders within ±1% and ±2% of mid. Depth is what large holders and institutions actually look at, because spread without depth is a shop window with nothing in the store.

Slippage. The gap between expected price and achieved average price of an order — the direct consequence of eating through depth. It is what your community complains about, usually without knowing the word.

Uptime. The percentage of time the MM's quotes are actually live in the book within the agreed spread and size. Serious agreements specify uptime per venue, per side, with 95%–99% as the professional range. Uptime is where weak MMs cheat, quoting beautifully in calm markets and vanishing precisely when volatility makes quoting expensive.

Maker and taker. Makers place resting limit orders that provide liquidity; takers remove it. Exchanges charge takers more and often rebate makers.

Inventory, toxic flow, reference price. Inventory is the stock of token and quote currency the MM holds to run quotes. Flow is 'toxic' when traders are systematically better informed than the quotes. A reference price is the desk's blended fair value across every venue where the asset trades.

When you evaluate any market maker, the professional summary of your market is a small table: spread in bps, depth at ±1% and ±2%, uptime, on each venue, over time. If a desk cannot or will not report in those terms, that fact is itself the report.

Understanding MM revenue is not academic. Every structural conflict in the industry traces back to a revenue line. There are four.

Spread capture. Buy at the bid, sell at the ask, thousands of times a day. Honest, alignment-neutral, and on small tokens usually insufficient on its own to cover inventory risk.

Exchange rebates. On venues with maker rebate programs, high-volume MMs are paid to quote. On major venues this is a significant revenue line.

Fees and retainers. The client pays directly for the service. Fully aligned, fully transparent.

Trading PnL on the token itself. The MM profits from movements in the token price via inventory it holds or options it was granted. This is where alignment breaks — a counterparty with a large optioned position in your token has views about where your price should go, and the tools to express them.

In every trade, the MM's counterparty is your community and your investors. That is not sinister — it is what liquidity provision is. But it means the MM's information advantage and incentive structure operate on the people you most need to protect, which is why the engagement model is the most important commercial decision in this entire domain.

You pay the MM a monthly fee, typically in stablecoins, and you supply the trading inventory: an allocation of your token plus quote currency, deposited on the venues to be covered. The MM quotes to contractual KPIs — spread, depth, uptime, per venue. All trading PnL on your inventory belongs to you. The MM's compensation is the fee, full stop.

Pricing in 2026 spans a wide range because scope drives cost: single mid-tier CEX pair with modest KPIs runs $5k–$10k/month, multi-venue coverage with tight spreads and serious depth commitments from a reputable desk runs $20k–$50k+/month. Inventory requirements sit on top — commonly $50k–$500k total across venues for a small-to-mid cap program.

What is good about it: alignment. The MM has no position in your token's direction, no options to harvest, no incentive beyond keeping the contract by hitting KPIs. Reporting disputes are clean because the deliverable is measurable. Termination is clean because inventory is yours and comes back.

What is hard about it: cash and capital. The fee is a real operating cost paid monthly regardless of market conditions, and the inventory is treasury capital exposed to your own token's volatility. Post-2022, many projects run lean treasuries, and the retainer model's honest costs are exactly the costs lean treasuries struggle to carry.

The dominant structure in small and mid-cap crypto MM. You lend the MM a quantity of your token — commonly 0.5%–3% of supply — for a fixed term, usually 12–24 months. The MM uses this borrowed inventory to quote your markets. You pay no monthly fee. In exchange, the MM receives call options on the loaned tokens: the right, at the end of the term, to purchase those tokens at strike prices fixed today, typically set at or somewhat above spot, often in tranches at ascending strikes.

At maturity, one of two things happens. If your token trades above the strikes, the MM exercises, buys the tokens at agreed prices, and returns you the cash. You have effectively sold treasury tokens at prices you locked in long ago, below the market. If the token trades below the strikes, the MM returns the tokens and walks away, having had free use of your inventory for two years.

Now the uncomfortable part, stated plainly because almost nobody states it to founders: this structure converts your MM from a service provider into an options trader whose PnL is your token's volatility and price path. The desk holding those options has a hard mathematical incentive toward volatility itself, since options gain value with volatility. A market maker paid in options is being paid more, the more turbulent your market is.

Reputable desks manage these conflicts with internal controls, and the model is not inherently a scam. It solves a real problem: it lets unfunded projects obtain professional liquidity by paying in upside instead of cash. But the same structure that lets an honest desk defer compensation lets a predatory one borrow a huge slug of supply, quote thin cosmetic liquidity, trade the inventory aggressively, and harvest the option value it can influence.

The pricing illusion. Founders read 'no monthly fee' as 'cheap.' Price the options and the illusion dissolves. A two-year call option package on 2% of supply, struck near the money, on a volatile small-cap token, is routinely worth more than $500,000 of retainer fees would have been. There is nothing wrong with paying in upside. There is a great deal wrong with not knowing you are doing it.

Choose retainer if your treasury can carry $10k–$40k/month plus inventory, your token has real value and holders to protect, you answer to a DAO or serious investors, or your previous MM relationship ended badly and you now value auditability above all.

Accept loan + options if you are pre-revenue with a thin treasury, liquidity is existential for an imminent listing, and the realistic alternative is no professional liquidity at all. Then negotiate hard: smaller loan size, strikes meaningfully above current price, shorter term, European-style exercise at maturity only, explicit KPIs, and reporting that lets you see what is being done with your inventory.

The hybrid — reduced monthly fee plus a smaller option package — is often the adult answer. Desks rarely lead with it. Ask.

The model choice is also not permanent. A common healthy arc: loan at launch, hybrid at first renewal once the token has value and data, then pure retainer at maturity. Write the arc into your planning.

Before listing. Every serious CEX evaluates liquidity readiness as part of listing diligence: which MM will support the pair, what depth and spread will be maintained from minute one, what inventory will be deposited before trading opens. Many venues require a named, approved market maker as a listing condition. The MM engagement must be signed and funded weeks before the listing date.

After listing. The exchange monitors your pair. Persistent wide spreads, empty books, and dying volume put a pair on internal delisting watchlists. Maintaining professional liquidity is listing insurance for the fee you already paid.

Designated market maker programs. Major venues run formal programs granting fee advantages, rebates, and API privileges to firms that commit to quoting obligations. A desk with top-tier status on your target exchange can deliver the same KPIs at lower cost than one without it.

Sequencing venues. Each new venue is a fixed increment of inventory and coverage cost, so the professional sequence is depth-first: one primary market with genuinely good metrics before you scatter across five thin ones. Announcement-driven listing sprees are the most expensive marketing strategy in crypto per unit of durable benefit.

Most tokens in 2026 live on both centralized order books and decentralized AMM pools, and teams routinely fund one while neglecting the other.

On a CEX, liquidity is active — thousands of limit orders being placed and refreshed by the MM's systems. Quality is a function of the desk's skill, capital, and contractual obligations.

On a DEX, base liquidity is passive — an AMM pool prices along a curve and requires nobody to actively quote. Your decisions become: how much capital to seed into pools, on which chains, in which fee tiers, across which price ranges if using concentrated liquidity.

Protocol-owned liquidity vs the rental problem. Incentivized liquidity rents depth that leaves the moment a better farm appears. Protocol-owned liquidity converts that permanent rent into a balance sheet asset that earns fees instead of paying them. For most projects in 2026 a core of owned liquidity plus targeted, time-boxed incentives beats pure emissions renting on any horizon longer than a quarter.

Your liquidity budget has three components: inventory (token plus stables committed to quoting), service cost (retainer fees or option value you are granting), and DEX capital (pool seeding).

Start from target depth, not from vanity. For a newly listed small cap, $25k–$50k per side at ±1% on the primary pair is a respectable floor. For a token courting institutional attention, $100k+ at ±1% across venues is the conversation starter. Working multiple: 3–5x per-side depth per venue.

Spread targets set the skill bar, not the capital bar. Adequate depth at 30–50 bps is a healthier young market than cosmetic tightness at $200 of depth.

Budget it as OpEx with a floor, not a launch line-item. Liquidity is an operating expense for the life of the traded asset. A program you can only afford in good conditions disappears exactly when holders need it.

Worked example. A project listing on one mid-tier CEX with an existing DEX pool, targeting $40k per side at ±1%: roughly $150k–$250k of working inventory on the CEX, a retainer in the low-to-mid five figures monthly or option package priced accordingly, and $100k+ of DEX pool depth. Meaningful liquidity for a small cap is a mid-six-figure capital commitment.

Scope and venues. Named pairs on named venues. Vague scope lets the desk deliver its KPIs wherever cheapest for it, not where you need them.

KPIs. Maximum spread in bps, minimum depth per side at defined bands, minimum uptime %, per venue, with measurement methodology written down. An agreement whose KPI section fits in one sentence is not a liquidity agreement — it is a token custody transfer with vibes.

Reporting. Monthly minimum, weekly better, with per-venue KPI compliance, inventory positions, and trading summary. Best practice in 2026 is dashboard access showing near-real-time compliance rather than PDFs of the desk grading its own homework.

Inventory and custody. For retainer deals: whose accounts, what controls, what happens to inventory at termination. For loan deals: exact loan size, term, and the complete option schedule — quantities, strikes, exercise style, expiry, settlement mechanics.

Use-of-inventory restrictions. The clause most often missing and most worth fighting for: restrictions on what the desk may do with your tokens beyond quoting. May they lend them onward? Short them on derivatives venues? Silence on this clause means yes to all three.

Term and termination. 6–12 month initial terms with termination for KPI breach after a cure period, and termination for convenience with 30–60 days notice. Watch option packages that survive termination in full — they remove your leverage over a non-performing desk.

Price talk. Any mention of supporting, defending, or growing your token price, or volume targets detached from real order flow.

Guaranteed volume. Organic volume cannot be guaranteed by an honest liquidity provider. Guaranteed volume means manufactured volume.

Loan asks above 3% of supply, or option strikes at or below current price, or American-style options exercisable anytime.

No KPIs offered. A desk that resists specifying spread, depth, and uptime is planning not to be measured.

Opaque or self-graded reporting — monthly PDFs with no raw data access, and a dispute window measured in days.

Silence on use of inventory in a loan structure, combined with resistance when you ask to add restrictions.

Unverifiable reputation. No referenceable clients, no exchange program status you can confirm, entities registered nowhere in particular, teams allergic to naming principals.

Pressure and listing-week urgency. Desks circling listing announcements with expiring offers are counting on you having no time to read the contract properly.

Signing well is half the relationship. The other half is verification, and the tooling for it is better than most teams realize.

Independent monitoring is non-negotiable. Third-party liquidity analytics services track your pairs' spread, depth, and uptime across venues independently. A junior team member with a script, or an analytics subscription in the low hundreds per month, converts your MM relationship from faith-based to evidence-based.

Review on a cadence. Monthly KPI review against the contract, quarterly relationship review covering venue performance, inventory usage, upcoming events, and whether the coverage map still matches where your volume actually happens.

Watch the event windows. Pull uptime and spread data for macro selloffs, your own bad news, unlock days. A desk that hits 98% uptime overall but was absent during your token's three worst hours delivered the letter of the contract and none of its purpose.

Keep the exit warm. Know your termination mechanics, keep at least one alternative desk relationship alive, and never let inventory or option structures accumulate to the point where leaving is unaffordable.

The pitch arrives eventually, from a 'volume support' desk: guaranteed daily volume, trending placement on trackers, a livelier chart. What is being sold is wash trading — the desk trading with itself in your pair to fabricate activity.

Detection is easier than buyers believe. Exchanges, analytics firms, and increasingly regulators run pattern detection tuned for self-matching flow, uniform trade sizing, and volume with no corresponding depth or holder growth. Tokens flagged for manufactured volume are delisted, and the delisting announcement says why.

The audiences you are trying to impress discount reported volume by default and look at depth and holders instead, so the purchase buys nothing even while it works.

Wash trading is market manipulation, plainly illegal in every serious jurisdiction, and enforcement in this area has been active for years — reaching desks and issuers who commissioned the trading. The legitimate version of everything fake volume promises is real liquidity: tight spreads and honest depth attract real traders, real volume follows.

Selecting a market maker is a procurement exercise, and running it like one changes the outcome more than any single negotiating tactic.

Longlist from evidence, not inbound. Build a list of 5–8 desks from tokens whose markets you admire (their teams will usually tell you who covers them), exchange program participant lists, and advisor networks.

Standardize the ask. Send each desk the same brief: token profile, venues and pairs, target KPIs, inventory availability, model preference. The variance between proposals for identical scope is the most educational document you will read this quarter.

Diligence the shortlist. Reference calls with 2–3 current clients of comparable size, focusing on reporting honesty, behavior during drawdowns, and termination experiences.

Normalize the economics before comparing. Convert every offer to a single figure — estimated total cost over 24 months including priced option value, against the identical KPI table. The normalized comparison routinely reorders the shortlist.

Negotiate two finalists in parallel. Competitive tension does more than any clause-by-clause cleverness, and the runner-up becomes your warm alternative.

Composite drawn from post-mortems that arrive on our desk after the fact.

Month 0. A project three weeks from its first CEX listing is approached with a polished deck. The offer: no fees, 2.5% of supply as a 24-month loan, options struck at spot, 1.25x, and 1.5x. The team, burning runway, signs a nine-page agreement whose KPI section reads 'commercially reasonable liquidity across supported venues.' Nobody prices the options. Priced properly, the package is worth roughly four years of what a retainer would have cost.

Months 1–3. Listing goes smoothly. Monthly reports arrive as two-page PDFs with venue volume totals and no depth or uptime data.

Months 4–8. The market cools. Spreads drift from 25 bps to 120 bps. Depth at ±1% quietly halves, then halves again. When the team asks, the desk cites 'market conditions' and points to the agreement's absent KPIs. On-chain watchers notice loaned tokens moving to derivatives venues.

Months 9–14. A modest recovery lifts the token toward the first strike. The market develops a strange ceiling: rallies stall in the same zone repeatedly. Whether the desk is actively managing price toward its strikes or merely hedging its option book mechanically is unknowable from outside — and this is the central lesson: the structure makes the benign and malignant explanations produce identical charts.

Months 15–24. At maturity the token sits above the first two strikes. The desk exercises, acquiring 2% of supply at prices 40% below market, and sells methodically into the book it still controls. Total cost of the 'free' market making, in dilution, forgone upside, and market damage: a mid-seven-figure sum, for liquidity that stopped being real in month five.

The work happens before you sign.

Most liquidity problems are blamed on market makers and were actually committed months earlier, in the tokenomics spreadsheet. Market quality is downstream of supply design.

Circulating float. A token with 2% of supply circulating trades like a penny stock regardless of who quotes it. Desks quoting ultra-low-float tokens charge for the privilege, in fees or in option terms.

Unlock schedule. Every cliff unlock is a scheduled liquidity event: a known date when supply lands on holders whose cost basis is far below market. Projects that smooth unlocks into linear vesting and pre-arrange event coverage experience unlocks as volatility. Projects that do none of that experience them as craters, and then blame the desk.

Treasury behavior. Nothing poisons an MM relationship faster than a treasury that trades against its own market maker: quiet OTC sales that land as surprise inventory, undisclosed selling into strength. Desks watch client treasuries on-chain. Disclose treasury operations in advance.

Quote currency choice. For most projects, a stablecoin pair should carry primary liquidity. Stablecoin pairs dominate 2026 volume and give holders a clean risk-on/risk-off decision.

Everything above can be executed by a founding team with time, market structure literacy, and negotiating patience. The honest case for advisory help is that listing timelines rarely allow the first, this guide is a floor rather than a ceiling on the second, and desks negotiate these agreements weekly while you will do it twice in your life.

What Xavion Capital does in this domain is matchmaking and deal support on the issuer's side of the table: mapping your project's profile, venues, treasury, and constraints to desks in our network whose model, tier, and appetite actually fit; running the standardized selection process so proposals arrive comparable; pricing the option structures so you see the real cost of every loan-model offer next to its retainer equivalent; and stress-testing the agreement against every failure mode in this guide before you sign it.

We sit on your side. We are not a market maker, and we hold no position in the outcome except the engagement itself.

Tell us where your token trades, what your liquidity looks like today, and what you are trying to fix. We will tell you plainly whether you need a market maker, a better market maker, or a better agreement with the one you have.