How crypto markets actually work

The interesting part is not the assets — it is that the market maker was replaced by a formula. Automated market makers are a genuinely new answer to an old problem, and their failure modes are new too.

Part of how markets actually work.

Set aside what you think about the assets. The genuinely interesting thing here is structural: crypto markets solved the liquidity problem a different way, and it is the first new answer in a long time.

Recall the problem. In equities, you can always sell because a market maker quotes both sides and earns the spread. That requires a firm, capital, regulatory permission, and a reason to show up.

What if nobody shows up? A new token has no market maker. Traditionally it would be untradeable.

The automated market maker

The answer is to replace the firm with a formula and a pool of assets.

Anyone can deposit two assets — say a token and a stablecoin — into a shared pool. The pool then quotes prices automatically using a rule. The most common is a constant product: multiply the quantity of the two assets together, and keep that product constant as people trade.

If someone buys the token, the pool has less of it and more stablecoin. To keep the product constant, the token’s price rises. Sell, and it falls. The price is a function of the pool’s contents, computed rather than quoted.

The people who deposited assets are liquidity providers, and they earn a share of the trading fees. They have taken the market maker’s role: supplying the ability to trade, and being paid for it.

That is elegant. It means any asset can be tradeable immediately, with no firm’s permission, as long as someone will fund the pool.

What it costs the provider

The new structure has a new failure mode, and it is not obvious.

If the token’s price moves substantially in either direction, the pool automatically sells the asset that is rising and buys the one that is falling — that is what keeping the product constant requires. The provider ends up with more of the loser and less of the winner.

The result is that a liquidity provider can end up worse off than if they had simply held the two assets. The term of art is impermanent loss, which is a misleading name: it is only impermanent if the price returns to where it started, and often it does not.

So providing liquidity is not free income. It is being paid fees in exchange for accepting a specific and sometimes large exposure. Whether it is worth it depends on fee volume against price movement — the same trade an equity market maker makes, priced differently and frequently misunderstood by the people taking it.

Slippage is the same problem in new clothes

When you trade a large amount against a pool, you move the price as you go — the same effect as eating through an order book, arising from a formula rather than from resting orders.

Small trade in a deep pool: barely noticeable. Large trade in a shallow pool: you can move the price against yourself dramatically within a single transaction.

Which means the headline price is only available for small sizes. That is true in every market — it is just unusually visible here, because you can compute it exactly.

What is genuinely different

It never closes. No opening bell, no weekends, no circuit breakers. Circuit breakers exist in traditional markets to interrupt panics. Without them, moves run further and faster.

Anyone can provide liquidity. No licence, no firm. Genuine democratisation of a role that was closed.

The rules are visible. The pricing formula is public code rather than a firm’s private strategy.

And the counterparty may be the exchange. On a centralised crypto exchange, the venue may also custody your assets, run the order book, and in some cases trade against you. In regulated equity markets those functions are deliberately separated — and the failures in this industry have clustered precisely where they were not.

What the evidence says

Better data than most of the internet economy, because blockchains are public — every transaction is visible and researchable, which is unusual and genuinely useful.

What that data consistently shows: returns to retail participants in short-horizon trading resemble the day trading findings — most lose, and the losses go to more sophisticated participants and to fees. Public ledgers make this measurable in a way equity markets do not.

It also shows a structural extraction retail rarely accounts for: because transactions are public before they settle, faster participants can act on the knowledge that your trade is coming. You pay for that in the price you receive.

What to take from it

The mechanism is a real innovation. Replacing a licensed firm with a formula and a pool is a genuinely new answer to the oldest problem in markets, and it works.

Liquidity provision is a trade, not income. Fees compensate for exposure. It is the market maker’s bargain, available to anyone, and frequently taken by people who have not priced the exposure.

Every structural lesson from the other markets still applies. Liquidity is deepest when calm and thinnest when needed. Someone faster is on the other side. Costs are embedded in prices rather than itemised.

And ask who holds what. The separation of custody, execution and market-making exists in regulated markets because of what happened when it did not. Where the same entity does all three, that separation is the thing being economised on.

Next: the market that sets the price of everything.