Pre-TGE Market Making Setup: The Twelve Weeks Before Your First Print

Pre-TGE market making is mostly not trading. It is a sequence of decisions and integrations that have to be finished before the first print, because none of them can be done afterwards without the market watching. The work divides into four blocks: deciding the venue set and the day-one float, selecting a desk and writing KPIs that are measurable, completing exchange and API onboarding while inventory is funded and segregated, and running a full rehearsal of the opening quote before the announcement goes out. Twelve weeks is comfortable, eight is workable, and under four weeks something in the list gets skipped, which is usually the rehearsal.

Most launch problems that look like market making failures are setup failures. The desk was appointed late, the exchange sub-account was not approved in time, the inventory arrived on the wrong chain, or nobody had agreed what the quote should look like in the first hour, so the first hour was improvised in public.

None of that is difficult. It is just sequential, and the sequence has dependencies that only reveal themselves when they are missed. What follows is the order the work has to happen in, working backwards from the opening print.

Twelve weeks out: the decisions the rest depends on

Three decisions have to be made before anything can be scheduled, and all three are commercial rather than technical.

The venue set. How many centralised venues, which ones, and whether there will be a decentralised pool at launch. Every additional venue divides the same liquidity budget across another book, so the question is not which exchanges would be nice to have but how many books the budget can hold to a tight spread simultaneously. Two well-quoted markets beat five thin ones, and thin markets on several venues produce visible price dispersion that undermines all of them. The trade-off between order books and pools is set out in CEX versus DEX market making.

The day-one float. The circulating supply at the moment trading opens determines almost everything else: how much depth is needed to absorb ordinary selling, what a normal price move looks like, and how much inventory the desk has to hold. A float that is too small produces a book where any real order moves the price, and one that is too large produces sell pressure the launch budget cannot absorb. The sizing method is in how much liquidity a token needs at TGE.

The pricing model for the engagement. Retainer, profit-sharing, or a loan-and-options structure, each of which puts the risk in a different place and has different consequences at the end of the term. This has to be settled before the desk is selected, because it determines which desks are even candidates. The economics of each are compared in retainer versus profit-sharing versus loan model.

Ten to eight weeks out: selecting the desk and writing the terms

Selection is a diligence exercise, and the useful questions are narrow. Which of your target venues does the desk already quote on today, and can it name the pairs. What spread and depth does it hold on tokens of comparable size and float. What does its reporting look like, and can you see a live example rather than a description. Who is the named operator responsible for your pair during your launch window, and in which time zone do they work.

Any desk that answers a depth question with a volume promise has answered a different question. Guaranteed volume, guaranteed price levels and opaque reporting are the three signals that most reliably precede a bad engagement, and the longer list is in market maker red flags.

The contract then has to convert those answers into obligations that can be measured. At minimum it should name the spread ceiling and how it is calculated, the depth commitment at defined price bands on each side, the uptime target and how downtime is evidenced, the venues in scope, who owns and controls the inventory and where it is held, the reporting cadence and format, and the exit terms including notice period, inventory settlement, option treatment and API key revocation. The full anatomy is in what belongs in a market making agreement, and the end of the relationship is worth reading about before the beginning, in what happens when a contract ends.

One clause is worth insisting on specifically for the launch: a defined quoting policy for the first forty-eight hours, distinct from the steady-state KPIs. The opening period behaves differently from every period after it, and holding a normal spread target through it is neither realistic nor desirable. What that period actually looks like is described hour by hour in the first forty-eight hours after TGE.

Six to four weeks out: accounts, keys and integration

This is the block that most often causes a delay, because it depends on third parties who do not work to your timeline.

Each venue needs a market making sub-account, which is approved rather than opened, and approval takes days to weeks depending on the exchange and the completeness of the corporate documentation behind it. API keys are then issued with trading permissions and, importantly, without withdrawal permissions, and are usually restricted to specific IP addresses. Fee tiers and any market maker rebate programme are negotiated at this stage rather than later, because they change the economics of tight quoting materially and cannot be applied retroactively.

The desk integrates and tests against the live venue with a small balance, which surfaces the details that documentation never covers: how the pair's tick size and minimum order size interact with your intended price, what the venue's rate limits mean for quote refresh frequency, whether the token's decimals behave as expected in the exchange's own systems, and how quickly deposits credit.

Two failure modes get caught here if anyone is looking. The first is a tick size that makes your target spread arithmetically impossible at the intended launch price, which is fixable only by changing the price or asking the exchange to change the tick. The second is a token contract feature, such as a transfer fee or a pausable transfer function, that the exchange's integration will not tolerate. Both are trivial eight weeks out and severe eight days out.

Three to two weeks out: inventory and funding

The desk needs both sides of the book. Tokens have to be transferred to the exchange sub-accounts in the amounts the depth commitment implies, and the quote currency, usually a stablecoin, has to be funded alongside them. If a decentralised pool is part of the plan, its seed amounts and the initial price have to be set, and the liquidity position locked or otherwise secured, with the lock verifiable by anyone who looks.

Custody arrangements need to be explicit. Whose keys hold the inventory, which addresses it sits at, what the desk can and cannot do with it, and how it is returned. Under a loan structure the tokens leave the project's control entirely, which is the point of the structure and also its principal risk, so the return conditions and the option strikes have to be written before the transfer rather than after.

Treasury policy belongs here too. If the project intends to convert any tokens to stablecoins in the first months, the mechanism should be agreed in advance and kept separate from the market making arrangement, because selling into your own quoted book through the same desk creates a conflict that is difficult to unwind later. The available approaches are described in selling treasury tokens without crashing the price.

Launch week: the rehearsal

The last thing to do before the announcement is to rehearse the opening, and it is the step most often skipped.

A proper dry run covers the exact sequence of the first fifteen minutes: when quoting starts relative to the trading enablement time, what the opening spread and depth are, what the desk does if the price gaps immediately in either direction, what it does if a venue's API degrades, and who has authority to change the parameters mid-session. It also covers the communication path, meaning the channel on which the project, the desk and the exchange's listing contact can all reach each other within minutes rather than hours.

Alongside it, agree the monitoring the project will actually watch. Spread, depth at the committed bands, uptime, and cross-venue price difference are the four numbers that matter on day one, and someone on the project side should be able to see them without asking the desk for a report. Motion Trade publishes these live, and how we do it is described in monitoring your market maker.

What the desk needs from you

The information a desk asks for is not administrative curiosity. Each item changes how it quotes.

• Total supply, day-one circulating float, and the full unlock calendar for the next twelve months, since every unlock is a scheduled depth requirement.
• The listing price or valuation the launch is structured around, and whether any private round is in the money at it.
• Every venue where the token will trade, including decentralised pools and any launch platform mechanics, because the desk has to keep prices consistent across all of them.
• The token contract address, decimals, and any non-standard transfer behaviour.
• The announcement schedule, including anything that will move the price on a known clock, such as an airdrop claim opening or an exchange campaign.
• The treasury policy, so the desk is not surprised by selling from an address it does not recognise.

The three ways this goes wrong

The desk is appointed too late, which compresses the integration work into the week when the project's attention is fully consumed by the launch itself, and the rehearsal is the first thing dropped.

The budget is spent on entry rather than maintenance, so the book is quoted properly for a fortnight and then thins as the retainer runs down. Monitoring at every venue measures daily and averages over consecutive-day windows, so a book that deteriorates in week three generates a record that a strong week five does not erase. The mechanism is described in why tokens get delisted.

The float and the depth are sized against each other rather than against reality. Depth is set to look adequate relative to a small float, the first genuine unlock arrives, and the same depth is now facing several times the sell pressure it was built for. Sizing should be done against the largest expected sell event in the period, not the average day.

FAQ

What is pre-TGE market making?

It is the preparation that has to be complete before a token trades for the first time: choosing the venue set and day-one float, contracting a desk with measurable spread and depth obligations, completing exchange sub-account and API onboarding, funding token and stablecoin inventory, seeding any decentralised pool, and rehearsing the opening quote.

How long before TGE should a market maker be appointed?

Eight to twelve weeks is the workable range. Exchange sub-account approval and API integration alone can take several weeks and depend on third parties, and the venue and float decisions that everything else rests on have to be settled before that work can begin.

How much liquidity is needed at launch?

It is a calculation rather than a fixed number, driven by day-one float, listing valuation, the venue set, and the largest realistic sell event in the first weeks rather than the average day. Sizing against the average produces a book that holds until the first unlock and then does not.

Should a token launch on a decentralised pool, a centralised book, or both?

It depends on where the token's audience already trades and whether centralised listings are part of the plan. A pool always quotes and is simple to seed, but it does not adjust to market conditions, does not satisfy exchange requirements, and exposes the provider to divergence loss. A quoted order book does the opposite. Many launches use both, with the split decided by budget rather than preference.

What KPIs should be in a pre-TGE market making contract?

A spread ceiling with a stated calculation method, depth commitments at defined price bands on both sides, an uptime target with evidence requirements, the venues in scope, inventory ownership and custody terms, reporting cadence and format, and exit terms covering notice, settlement and API key revocation. A separate quoting policy for the first forty-eight hours is worth adding, because that period does not resemble steady state.

Who should hold the inventory before launch?

Under a retainer the project normally retains ownership, with the tokens held in exchange sub-accounts the project controls and the desk holding trade-only API access without withdrawal permissions. Under a loan structure the tokens transfer to the desk, which is the point of that model, so the return conditions, option strikes and settlement terms have to be agreed in writing before the transfer happens.

What can still be fixed in the last week before TGE?

Quoting parameters, depth allocation between venues, and the announcement schedule. What cannot be fixed late is anything depending on a third party: exchange sub-account approval, fee tier negotiation, API whitelisting, and any change to the token contract itself.

Does a launch need more than one market maker?

Rarely at launch, and having several is usually worse rather than better, because two desks quoting the same book without coordination trade against each other and the project pays for both sides of that. Multiple desks make sense at a scale where venues and regions are genuinely separable, which is well beyond a first listing.

What does a launch dry run cover?

The exact sequence of the first fifteen minutes: when quoting starts relative to trading enablement, the opening spread and depth, the response to an immediate price gap in either direction, the response to an API failure at a venue, who can change parameters mid-session, and the channel on which the project, the desk and the exchange contact can all reach each other quickly.