Analysis by Korea IT Times Editorial Desk
Decentralized exchanges are often defined by the absence of a central intermediary. Their more fundamental innovation, however, lies in how prices are formed. While traditional exchanges discover prices through competition between buy and sell orders, automated market maker-based DEXs such as Uniswap calculate exchange rates through smart contracts and the changing balance of assets in liquidity pools.
Professor Hyoung Joong Kim of Kookmin University, a longtime researcher in cryptography, blockchain and digital assets, recently identified the central market question behind this structure. Writing on Facebook, Kim argued that AMMs have a relatively limited capacity for price discovery but a strong ability to track prices formed elsewhere.
When the price inside an AMM diverges from the broader market, arbitrageurs enter and trade until the gap narrows. In this system, part of the price-setting process moves away from competition among traders in an order book and toward an interaction between mathematical rules and arbitrage incentives.
Kim does not argue that AMMs have no price discovery function. He recognizes that they can provide an initial venue for pricing long-tail assets that lack sufficient activity in established order books. His distinction instead concerns how actively traded assets acquire their reference prices and how those prices are transmitted to on-chain liquidity pools.
Kim describes an AMM not as a mechanism for finding the perfect price, but as infrastructure that enables trading even in a world without sufficient liquidity. The observation shifts the standard by which AMMs should be assessed—from the completeness of their price discovery to their ability to keep markets operating.
This article takes Kim’s observations as its starting point. The subsequent analysis of Uniswap, liquidity-provider economics, Korean protocols and market policy draws on publicly available sources.
Mathematics sets the terms while arbitrage connects the prices
The foundation of Uniswap’s early design is the constant-product formula x × y = k. Here, x and y represent the quantities of two assets deposited in a liquidity pool. When a trader adds one asset and removes the other, the balance of the pool changes and the exchange price moves accordingly.
Consider a pool containing Ether and the dollar-denominated stablecoin USDC. If traders continue buying ETH, the amount of ETH in the pool declines while USDC increases, pushing the pool’s ETH price higher. As the size of a trade grows, the gap between the expected price and the average execution price—known as slippage—also widens.
An AMM does not simply import prices from an external market. Even if ETH rises first on a centralized exchange, the price inside an AMM pool may not adjust by the same amount until a transaction changes the pool’s asset balance.
That is where arbitrageurs enter. If ETH is cheaper in the AMM than on another exchange, an arbitrageur can buy ETH from the pool and sell it at the higher external price. The transaction reduces the pool’s ETH balance, increases its USDC balance and moves the AMM price closer to the wider market.
This is the price-tracking function highlighted by Kim. The mathematical formula establishes the trading conditions, while arbitrageurs transmit information between otherwise separated markets.
For highly traded assets such as ETH, reference prices are often established first in venues with greater liquidity and faster information flows. Arbitrage then carries those prices into on-chain pools. For a newly issued or long-tail token without an active order book, however, the AMM itself may become an important venue for initial price formation.
An AMM can therefore perform different functions depending on the asset and market. In one market it may help discover a price; in another it may primarily follow a price formed elsewhere.
The Bank for International Settlements has similarly explained that relative prices in AMMs are determined by the quantities of assets held in their liquidity pools. Arbitrage can correct deviations from external markets, but liquidity providers may be left worse off than if they had held the appreciating asset outside the pool.
Uniswap introduced concentrated liquidity in V3 to improve the capital efficiency of its earlier model. Instead of spreading capital across the entire potential price range, liquidity providers can place it within a selected interval around the current price. This creates deeper liquidity with the same amount of capital and can increase fee income where trading is concentrated.
The greater efficiency also brings greater management risk. If the market price moves outside the chosen range, the position becomes inactive and stops earning fees. A sustained move in one direction can also leave the position almost entirely composed of one asset. Uniswap’s explanation of concentrated liquidity
The cost behind LP fee income—and the signal for Korea
Liquidity providers deposit two assets into a pool and receive a share of trading fees. Their actual performance, however, cannot be measured by fee income alone. When the relative price of the two assets changes sharply, arbitrage removes more of the appreciating asset and leaves the pool holding a larger share of the weaker one.
As a result, the value withdrawn by an LP may be lower than the value of simply holding the same assets in a wallet. This difference is commonly called impermanent loss. The loss may narrow if relative prices return to their original level, but it becomes realized when liquidity is withdrawn, and there is no assurance that the price relationship will recover.
Recent market-microstructure research examines a related adjustment cost through the concept of loss versus rebalancing, or LVR. When an external market moves before an AMM has adjusted, faster arbitrageurs may trade against the pool at the previous price. Arbitrage performs the useful function of reconnecting markets, but part of the economic cost of that price correction can fall on LPs.
Kim has gone beyond explaining this problem to study the conditions under which impermanent loss and impermanent gain may arise. In a 2023 interview, Korea IT Times reported on his research combining origin-crossing, transaction reordering and batching to explore the possibility of producing impermanent gain under specified market conditions.
The research does not mean that losses disappear or gains are guaranteed in every market. Kim identified conditions including sufficient liquidity and a healthy balance between buying and selling. Its broader significance lies in challenging the assumption that impermanent loss must be accepted as an unchangeable feature of AMMs and exploring whether transaction processing and settlement design can improve outcomes for LPs.
KLAYswap, developed by South Korean blockchain company Ozys, is a domestic example of an AMM-based protocol. Operating within the Kaia ecosystem, it uses liquidity pools rather than an order book and applies a constant-product structure in which prices change with the quantities of the two pooled assets.
KLAYswap also explicitly identifies slippage and impermanent loss as material risks. Its documentation notes that LP losses may occur as differences between external prices and pool prices are corrected through the AMM mechanism. KLAYswap protocol documentation
South Korea’s digital-asset trading remains concentrated on centralized exchanges with Korean won markets. As stablecoins, real-world assets, tokenized securities and on-chain payments expand, however, AMMs could evolve beyond venues for exchanging speculative tokens. They may become infrastructure for digital-asset conversion, settlement, collateral valuation and liquidation.
That transition will require disclosure based on real LP performance rather than headline annual percentage yields. Fee income and token incentives should be presented alongside performance relative to passive holding, time spent outside an active price range, smart-contract audits, oracle dependence, liquidity concentration and administrative control.
Greater caution is also necessary when AMM pool prices are used for collateral valuation or forced liquidation. A low-liquidity pool can be moved sharply by a single trade or a deliberate attack. Time-weighted average prices and reference data from multiple markets may therefore be necessary instead of relying on a price recorded in one block.
AMMs have not eliminated price discovery. They have changed where it occurs and how external information reaches on-chain markets. Arbitrageurs connect prices across venues, while liquidity providers earn fees in exchange for bearing part of the market and adjustment risk.
Kim’s distinction invites the industry to assess AMMs not by whether they produce a perfect price, but by whether they can sustain trading where conventional liquidity is insufficient. It also raises the deeper questions of who keeps that market functioning, who bears the cost of price adjustment and whether technology can distribute that cost more effectively.
If South Korea is to capture the potential of on-chain finance, it must move beyond promoting high deposit yields and build infrastructure capable of measuring how prices are formed and what liquidity providers actually earn or lose. Kim’s distinction between price discovery and price tracking offers an important starting point for a more mature debate over market quality, transparency and responsibility.
