Traditional exchanges work on a simple premise: buyers and sellers meet, haggle, and eventually agree on a price. The New York Stock Exchange employed specialists to facilitate this dance; currency markets relied on banks quoting bids and offers. Then in 2018, a mechanism emerged that dispensed with counterparties altogether, replacing human judgment with a mathematical formula that determines prices automatically based on supply and demand.

Automated market makers represent one of the few genuinely novel financial mechanisms to emerge from cryptocurrency — not a digital recreation of existing infrastructure, but something architecturally different. Understanding how they work reveals both the elegance and the limitations of algorithmic finance.

The constant product formula

The dominant AMM design uses a deceptively simple equation: x × y = k, where x and y represent the quantities of two assets in a pool, and k is a constant. When a trader wants to swap asset X for asset Y, they deposit X into the pool and withdraw Y, but the product of the two quantities must remain unchanged.

This creates an automatic pricing curve. As traders buy more of asset Y, its quantity in the pool decreases while X increases. The formula ensures that the price rises as supply dwindles — buy enough of an asset and the price approaches infinity. Sell enough and it approaches zero. No human sets these prices; they emerge from the interaction between the formula and trader behavior.

Liquidity providers deposit equal values of both assets into the pool and receive tokens representing their share. In exchange for locking up capital, they earn a fraction of every trade — typically between 0.05% and 1% depending on the pool's configuration. The model transforms passive capital into active market-making infrastructure.

The impermanent loss problem

This elegance comes with a catch that took many early liquidity providers by surprise. When prices move significantly in either direction, liquidity providers end up with a less valuable portfolio than if they had simply held their original assets. The mechanism that enables automatic pricing also ensures that pools accumulate more of whichever asset is falling in value and less of whichever is rising.

Consider a pool containing equal dollar values of two assets. If one asset doubles in price relative to the other, arbitrageurs will buy the appreciating asset from the pool until prices align with external markets. The pool now holds more of the depreciating asset and less of the appreciating one. Liquidity providers would have been better off holding their original split.

This loss is called "impermanent" because it reverses if prices return to their original ratio — but for volatile pairs, prices rarely cooperate. Trading fees can offset this drag, but only if volume is sufficient. The calculation between fee income and impermanent loss determines whether providing liquidity is profitable, and it often is not.

Concentrated liquidity and capital efficiency

Later AMM designs addressed the capital inefficiency of the original model. In a basic constant-product pool, liquidity is spread across all possible prices from zero to infinity, even though trading typically occurs within a narrow range. Most of the capital sits idle, earning nothing.

Concentrated liquidity allows providers to specify price ranges where their capital will be active. A provider confident that two stablecoins will trade between 0.99 and 1.01 can concentrate their entire position in that range, earning fees on a much larger effective position. The tradeoff is increased impermanent loss if prices move outside the specified range — the position becomes entirely one asset at exactly the wrong moment.

This innovation dramatically improved capital efficiency but also transformed liquidity provision from passive income into active management. Providers must now monitor positions, adjust ranges, and compete with sophisticated actors running automated strategies.

Our take

Automated market makers solved a genuine problem: how to create liquid markets for thousands of token pairs without requiring dedicated market makers for each one. The mechanism is clever, transparent, and permissionless. But the industry's marketing of liquidity provision as passive income obscured real risks that cost retail participants dearly. The math works exactly as advertised — it just works against providers more often than the yield figures suggest. Understanding AMMs means appreciating both the innovation and the fine print.