Liquidity provision as a service

Market making for hire: quoting a token or contract under an agreement with the issuer or venue. A real business on crypto and prediction markets, with its own incentives and risks.

The idea

What is liquidity provision as a service?

It is market making under a contract or incentive scheme rather than purely on your own book. A counterparty that wants a liquid market (an exchange, a token issuer, a new venue) pays a provider to quote it. The provider takes on obligations (size, spread, uptime) and earns incentives (fees waived, rebates, a retainer, tokens) on top of, or instead of, ordinary spread P&L.

Ordinary market making is "I quote because I think the spread is worth more than the risks." Liquidity-as-a-service adds a second income stream: someone who benefits from your quoting (because a liquid market attracts traders, listings or users) pays you to do it, and in return demands you actually show up and quote tightly even when you would rather not. The diagram above is the whole deal: obligation arrows flow one way, incentive arrows the other, and the balance between them is the bargain.

The mechanics underneath are unchanged (earn the spread, manage inventory, survive adverse selection) but the economics change. Your effective spread is widened by the incentives (a rebate or fee waiver shifts your break-even spread down, sometimes below zero), and your freedom is narrowed by the obligations (you cannot simply pull when flow turns toxic without breaching the agreement). Three arenas, increasingly open in 2026, are taken in turn below: regulated exchange programmes, crypto token-project market making, and bootstrapping new venues and prediction markets.

What are designated and supplemental market makers?

On regulated exchanges, designated market makers (DMMs) and supplemental or registered liquidity providers (SLPs) are firms that formally commit to quote a name (a minimum size, within a maximum spread, for a minimum share of the trading day) in exchange for fee rebates, lower fees or improved standing. The exchange buys a tighter, more reliable market; the provider buys an economic incentive.

NYSE DMMs are the modern descendant of the specialist: one firm per listed name with affirmative obligations to maintain a fair and orderly market, quote at the inside a required fraction of the time, and participate in the opening and closing auctions (continuous vs auction), in return for economic and informational privileges. Nasdaq, Cboe and other venues run market-maker and supplemental liquidity provider (SLP) programmes with quoting requirements traded for enhanced rebates.

In the EU, MiFID II formalised this: a firm pursuing a market-making strategy must enter a binding market-making agreement with the venue, quote on a continuous basis during a specified portion of trading hours (with stressed-market exceptions), and the venue must run an incentive scheme. The regime turned informal liquidity provision into a documented, supervised obligation.

The bargain in one line: affirmative obligations (be present, quote tight, quote size) traded for economic incentives (rebates, fee tiers, auction participation) and sometimes informational or queue advantages. Whether it is worth it depends on whether the incentives plus spread clear your obligations' cost, which, crucially, includes being forced to quote into toxic flow you would otherwise dodge. These obligations have teeth in stressed markets: the cases where makers want to pull are exactly when the venue most needs them present, which is why the agreements include limited stressed-market relief and why being a DMM or SLP is a real risk transfer, not free money.

How does token-project market making work in crypto?

A crypto token issuer pays a market-making firm to provide a liquid two-sided market in its token (tight spreads, real depth) so the token trades well on exchanges and attracts users. Payment is typically a retainer plus a token allocation or call-option, sometimes with exclusivity. The model is legitimate, but the incentive structure has produced well-documented abuses.

A freshly listed token has no natural two-sided flow: a thin book, a wide spread, easily manipulated. The project wants it to look and trade liquid, so it hires a market maker to quote it. The MM earns the spread on the (often thin, often toxic) flow, plus the contracted incentives, and the project gets a market. Two contract structures dominate:

In the retainer / "loan + fee" model the project pays a monthly fee and/or lends the MM tokens to quote with; the MM returns them (or their value) at term, and its incentive is aligned with tight spreads and uptime. In the token-option / "loan + call-option" model the project lends tokens and grants the MM call options struck above the current price; here the MM profits if the token rises, which can misalign incentives, since the MM may benefit from price appreciation or volatility rather than a genuinely tight, fair market. This structure is where most documented abuses originate.

The honest, recognition-framed warning (this is detection, not a how-to): the token-MM space has seen documented manipulation, including wash trading to fake volume, coordinated pump-and-dump around the MM's inventory, and "liquidity" that vanishes the moment it is needed. These shade into the conduct covered on the market-manipulation pages (pump-and-dump, spoofing and layering) and the market-abuse regimes. A legitimate token MM provides a real market; the structures that reward price-pushing over tight quoting are the ones to recognise and avoid.

How do you bootstrap liquidity on a new venue or prediction market?

A new venue, instrument or prediction market starts with an empty book: no liquidity, so no traders, so no liquidity (a chicken-and-egg problem). Bootstrapping breaks it: an anchor liquidity provider is incentivised (fee waivers, rebates, token rewards or a grant) to quote first, making the market usable enough to attract organic flow, after which the incentives taper.

Nobody wants to trade into a one-sided, wide, empty book, and nobody quotes into a market with no traders. Someone has to go first and be paid for the privilege, bearing concentrated adverse selection and inventory risk on thin flow until the market thickens. That is the anchor MM's job and why the venue pays for it.

New crypto venues and DEX liquidity mining. Exchanges and protocols run liquidity-incentive programmes (rebates, fee waivers or token emissions) to attract makers to new markets. On AMM-based DEXs the analogue is liquidity provision to a pool, a distinct passive mechanism with its own risk (impermanent loss); but for order-book DEXs and CEX listings it is recognisably contracted or incentivised market making.

Prediction markets (Polymarket). New markets open thin and event-driven, with bounded [0,1] payoffs and a terminal resolution. Bootstrapping liquidity here means quoting a two-sided market on an outcome few yet have a view on, bearing unique risks: adverse selection spikes as informed traders price the outcome, the book can go thin near resolution, and your inventory settles to 0 or 1. The incentive is often early-mover spread plus any platform rewards; the maths is the same A–S quoting, clamped to [0,1]. The economics: bootstrapping pays best when the incentive plus the thin, toxic spread covers the concentrated risk of going first, and worst when the market never develops organic flow and you are left as permanent, subsidised liquidity. Sizing that bet is the whole game.

How does being paid to quote change the economics?

Incentives shift your break-even spread down and your obligations shift your freedom down. A rebate or fee waiver means you can quote tighter, sometimes below the ordinary break-even, even at a gross-spread loss covered by the incentive. But mandatory uptime and maximum-spread rules force you to quote into toxic conditions you would otherwise sit out, raising your adverse-selection cost.

Mapped onto the P&L decomposition: incentives add a term: retainer, rebate, fee waiver or token value flows in regardless of (or on top of) spread P&L, lowering your break-even spread, the premise of rebate capture taken to its contracted extreme. Obligations remove optionality: you must be present and tight a minimum fraction of the time, so you cannot simply pull when VPIN flags toxic flow; that forced presence raises your adverse-selection cost, because you eat fills you would otherwise avoid.

So liquidity-as-a-service is a bargain over who bears which risk: the payer transfers the cost of a liquid market to you in exchange for an incentive, and the deal is good only when the incentive over-compensates the obligation. Pricing that trade-off is exactly the market-making skill, now with a contract on top. What AI changes in 2026: better toxicity detection lets a contracted MM meet its uptime obligation and manage adverse selection (widen within the maximum spread, lean inventory harder) squeezing more margin out of the same agreement. The structure of the bargain is unchanged; the provider's ability to honour it cheaply improves (see what AI changes for HFT).

Worked example

A simplified liquidity-provision-as-a-service arrangement, as of a 2026 worked snapshot. A new order-book venue wants an anchor market maker in one instrument; the deal pays a monthly retainer of 10,000 plus a maker rebate of 0.0002 per share, in exchange for quoting at least 500 shares within a 0.10 maximum spread at least 90% of trading hours.

Ordinary (un-incentivised) economics on this thin, toxic new market: gross spread capture about +30,000/month, but adverse-selection cost about −34,000 (the flow is informed because the instrument is new and few have a view) so ordinary net is about −4,000/month. On its own account the provider would not quote this: the break-even spread exceeds the maximum spread the venue will tolerate.

The incentive turned an un-quotable market into a profitable contract by lowering the effective break-even below the maximum spread. The obligation's cost is real: the 90%-uptime rule means the provider must keep quoting through the toxic bursts it would otherwise pull from; that is why the adverse-selection term is as large as −34,000. If the obligation were 50% uptime, the provider could dodge the worst flow, adverse selection might fall to about −20,000, and the deal would be far more profitable. The venue prices the uptime requirement precisely because that forced presence is what makes the market reliable for everyone else.

The bootstrapping bet. If organic flow arrives over six months and the book thickens, adverse selection falls, spread P&L rises, and the venue tapers the retainer, so the provider transitions to an ordinary profitable maker. If organic flow never comes, the provider is permanent subsidised liquidity and the deal is only as good as the retainer. Sizing that probability is the whole decision.

The numbers are illustrative and synthetic; real retainers, rebates, obligations and adverse-selection costs vary enormously by venue, instrument and contract (and token-MM terms are typically confidential). Educational only, not investment advice; no P&L is promised, and the token-MM structures noted above include arrangements that have been used abusively (see the recognition and detection pages).