The spread is only the starting point
From the outside market making looks like free money. You post a bid and an ask, someone trades against you on both sides, and you keep the difference. In a textbook the spread is your profit. In crypto the spread is your gross revenue, and there's a long list of costs behind it that decide whether you actually made money.
Say a maker quotes ETH/USDT with a bid at $3,499 and an ask at $3,501. That two-dollar spread on a $3,500 asset is about 0.057%. If they buy and sell equal amounts at those prices, they clear $2 per ETH round trip. Multiply by volume and it looks great. The catch is that buying and selling equal amounts at good prices is the hard part, not the easy part.
Inventory risk is the real business
Every fill on one side leaves you holding inventory. A wave of sellers hits your bid and now you're long a pile of ETH. Price drops and those holdings bleed. Inventory risk is the dominant cost of market making, and managing it is where the actual skill lives.
Pros hedge continuously. Accumulate a long ETH position on Binance and you might short ETH perps on Bybit to flatten out. That hedging isn't free either. You pay funding on the perp, eat slippage on the hedge, and carry the operational headache of running positions across a bunch of venues at once.
The firms that survive at this, names like Wintermute, GSR, and DWF Labs, built infrastructure specifically for the inventory cycle. Co-located servers near the matching engines, pre-funded accounts on dozens of venues, and quoting algorithms that shift their prices in real time based on how their inventory is tilted right now.
Adverse selection eats your edge
The second big cost is adverse selection. When someone trades against your quote, there's a real chance they know something you don't. An insider who knows a token is about to be delisted sells into your bid, and you've just bought something that's about to lose most of its value. The maker can't tell informed flow from noise, so they widen the spread enough to cover the expected losses from trading with people who know more than they do.
In crypto this is worse than in traditional markets. There's less regulatory disclosure, so the information gaps are bigger. Information spreads unevenly, with some participants seeing exchange announcements or on-chain data milliseconds ahead of everyone else. And insider trading, hard as it is to measure, is widely assumed to be more common than in regulated equity markets.
You can see it in the spreads. BTC/USDT on Binance trades at 1 to 2 basis points. A mid-cap altcoin on a secondary exchange might sit at 50 to 100. That gap isn't only about volume. It reflects the higher odds that whoever's trading the altcoin knows something the maker doesn't.
The token project revenue stream
A big chunk of maker revenue doesn't come from organic trading profit at all. It comes from deals with token projects. A new token listing needs liquidity. Without a maker the order book sits empty, spreads blow out, and the token looks dead on arrival. So projects hire makers, usually through a loan arrangement.
The standard structure goes like this. The project lends the maker a large allocation of tokens, often 3 to 5% of circulating supply, plus some stablecoins. The maker uses that to provide liquidity on agreed exchanges for an agreed window, typically 12 to 24 months. In return the maker collects a monthly retainer and, more importantly, call options on the tokens they were lent. If the price goes up, the maker profits by exercising those options.
That setup creates an alignment problem the industry hasn't really reckoned with. The maker wins if the token's price rises, and they happen to control a large share of the order book. Whether they lean on that position actively or just passively, the incentive to push the price up is sitting right there.
Cross-venue fragmentation
Crypto liquidity is scattered across centralized exchanges, DEXs, and OTC desks. A maker quoting on Binance has to worry about getting picked off by someone who saw a move on Coinbase first. That cross-venue latency arbitrage is a constant drain on the P&L.
The usual response is to quote on as many venues as possible and to internalize flow where the rules allow. Some makers run their own dark pools or OTC desks and match buyer against seller without ever posting to a public book. That lets them capture spread without eating the adverse selection that comes with quoting in the open.
What this means for regular traders
Knowing how maker economics work helps you in a few practical ways. When you see tight spreads and a deep book on a new token, that depth is almost certainly a contracted maker. It's real in the sense that you can trade against it, and it's artificial in the sense that it vanishes when the contract ends. That tells you more about the token's true liquidity than today's order book does.
When maker contracts expire or a maker walks away from a token, the fallout can be brutal. Spreads widen, depth evaporates, and suddenly you can't move size without moving the price. Watching for the tells, a book that thins out fast or odd patterns in trade timing, gives you some warning before those transitions hit. On Blockcircle I spend a lot of time looking at exactly those depth and timing signals across venues, and once you know what maker withdrawal looks like it's hard to unsee.