Author: Hayden Adams, Founder of Uniswap
Compiled by: Jiahuan, ChainCatcher
I have been working at the forefront of DeFi for 9 years. It's a fascinating field with nearly unlimited potential and depth, and the ability to change capital markets.
I have always believed that AMMs have enormous potential, but a question has persistently nagged me for the past decade: Could this new market structure really become the core engine for all financial markets?
After years of evolution and growth, a path for AMMs to dominate a broader range of financial markets is becoming increasingly clear. To explain this, it's best to start with 1976.
Tokenization Changes Who Does Market Making
The index fund celebrates its 50th anniversary this month. In 1976, when Jack Bogle launched the index fund, he hoped to raise $150 million but ended up with only $11.3 million. Competitors called it 'Bogle's Folly' and even made posters accusing the index fund of being 'un-American'.
They argued that a fund making no investment decisions could never beat highly paid professionals dedicated to that task. Today, the majority of fund assets in the US are in passive investment vehicles.
I have been thinking about this lately because the phase where tokenization is seen as 'pie in the sky' is also ending. The SEC has approved Nasdaq and NYSE to trade tokenized stocks. The DTCC, which handles the settlement of nearly all US securities, also conducted a real-world test of a tokenized trade in July. Almost all these changes are described in the same way: tokenization is an infrastructure upgrade.
The same markets, just faster, cheaper, and operating 24/7. All of that is true, but I think the phrase 'infrastructure upgrade' obscures a bigger story. Tokenization makes markets programmable. It will change which markets can exist, who makes markets, and what assets in those markets can trade directly with what.
In 2018, I created Uniswap, an automated market maker protocol. Anyone can deposit two assets into a shared pool and earn fees from each trade; prices automatically adjust along a curve as users buy and sell. Uniswap has run autonomously since day one, processing over $4.6 trillion in cumulative volume, and helped push decentralized exchange spot trading volume from less than 1% of centralized exchange volume to over 20%.
As Uniswap and other AMMs have grown, their liquidity has formed a pattern most financial market participants haven't yet noticed: correlated trading pairs.
Which Markets AMMs Won First
To win a bigger market, you first have to win a part of it. AMMs first found product-market fit in long-tail markets, because most such assets couldn't attract the attention of professional market makers. On Uniswap, anyone can create a market with a single transaction, and asset issuers and early supporters can become the initial LPs.
Next came stablecoin pairs. For pairs like USDC/USDT, a good passive strategy is close enough to optimal, and the lower cost of capital is enough to offset the efficiency gap. This is why professional trading firms today largely don't make markets for these stablecoin swap markets: they are being pushed out by passive AMMs with lower costs and lower return requirements.
Scale Barriers for Traditional Market Makers
Traditional financial markets largely belong to market-making firms. They integrate capital, trading strategies, execution technology, settlement, and distribution into the same vertically integrated business. There are good reasons for this architecture: assets exist in separate systems, settlement is slow, and every step must be done by someone, so having one company handle all functions was the most natural way.
As long as the scale is large enough, all these fixed costs can eventually be spread out. Citadel Securities handles about 25% of US stock trading volume, and last year, with about $21 billion in trading capital, generated a record $12.2 billion in net trading revenue.
Most people see these numbers as proof that the system works effectively. What I see is a market structure that is already firmly occupied.
Blockchain is Unbundling the Traditional Market-Making System
Blockchain enables competition at each layer, unbundling what was previously bundled together. Execution is done by code, custody and settlement become shared services anyone can access, and what once required proprietary infrastructure now becomes open-source software.
For AMMs, capital is the scarcest input, and the advantage goes to those who can hold inventory at the lowest cost. A trading firm needs a relatively high return to cover its costs, so an LP willing to accept a lower return can compete at a lower cost.
Most market makers hedge out almost all price risk, and hedging itself has a cost, so an investor who already holds these assets can take on that price exposure for free. The cost of capital for an asset issuer might even be negative, because issuers typically pay professional market makers to provide liquidity for new assets.
In short, DeFi and AMMs lower the barriers to market making, allowing many new participants into the market. Their advantage could come from many places, like lower capital costs, a willingness to hold asset exposures that professional firms would typically hedge, or even because they are the asset issuer themselves.
But all of this ultimately depends on one question: Are automated strategies good enough for these advantages to actually hold?
Correlated Trading Pairs
Recently, I was on a call with one of the world's largest financial institutions. They asked me what the most common primary paired asset in DeFi is. I explained that Ethereum ecosystem assets often trade against ETH, Solana ecosystem assets typically trade against SOL, stablecoins trade against each other, and these liquidity clusters are connected by a small number of highly liquid trading pairs.
No one specifically designed this structure. It emerged naturally, partly because LPs typically perform better when the two assets they hold move in sync. Correlation between assets means liquidity providers take on less inventory risk, leading to deeper liquidity.As more assets become tokenized, the world's largest financial markets will reorganize in the same way.
This isn't possible today because in traditional finance, almost all trades must be settled in dollars out of practical necessity. Different assets exist in separate, siloed systems, and fiat infrastructure like SWIFT and Fedwire is the glue that binds them together. But blockchain is a more flexible, programmable connecting layer. Once assets are tokenized, they share a common settlement layer, so any asset can trade directly with any other.
NVDA/USD could become NVDA/SPY, with SPY/USD serving as the bridge back to USD. Oil company stocks could trade against an oil ETF or tokenized crude, private credit could trade against a tokenized US Treasury fund. Tokenization can also create markets across different asset classes, something extremely difficult or impossible in traditional financial infrastructure.
Delta Neutral is an Inefficiency
Traditional market-making firms typically aim to stay 'Delta neutral'. In trader terms, this means using the US dollar as the numeraire and minimizing any non-dollar risk. When making markets for highly volatile assets, they pay to reduce their non-dollar risk, i.e., hedge, often via options. This is one of the higher-cost components of traditional market making.
Grouping assets into lower-volatility 'correlated trading pairs' and connecting them with a small number of higher-volatility 'bridging pairs' can unlock many efficiency gains. But the most important one is this: Market making is cheaper and more efficient when done by someone who is already willing to hold the underlying assets.
And the higher the correlation between a pair of assets, the smaller the efficiency gap becomes between today's passive AMM strategy and the most advanced active strategies, making it easier to compete with the latter based on lower holding costs.
Concretely, if someone holds NVIDIA long-term, they likely also hold SPY long-term. Therefore, compared to NVIDIA/USD, the efficiency gap between a passive AMM and an active strategy is much smaller on the NVIDIA/SPY pair.
How Correlated Pairs Connect the Whole Market
If stocks primarily traded against SPY instead, then all trades that start or end in dollars would go through a single pair: SPY/USD. These bridging pairs would still require highly specialized market-making expertise, but there would be far fewer of them, and the trading flow through them would be large enough to justify professional resources.
DeFi has already proven this structure. ETH/USDC is one of the deepest on-chain liquidity markets because trades between different liquidity clusters are routed through it. Passive LPs provide liquidity for correlated pairs, while active LPs compete on bridging pairs.
Investors could still buy and sell any asset using dollars because routing between different pools would happen automatically. At the same time, liquidity would concentrate where the risk is lowest, not where traditional infrastructure mandates it must be. This would drive the deepest markets toward correlated trading pairs—precisely the area where AMMs already hold the greatest advantage.
Tokenized Stocks Have Already Started Trading This Way
On-chain correlated liquidity first came from crypto-native assets. But today, the first correlated markets for tokenized stocks have emerged: there are currently 10 tokenized stocks trading directly against SPY via Uniswap pools on Robinhood Chain.
In the first 12 days after launch, these pools saw $33 million in volume from over 11,000 users, with a significant portion of that volume occurring during US stock market closures. Some trades even go directly from one stock to another, never touching dollars.
Notably, we are even starting to see Meme coins paired with so-called 'related' stocks: Elon-related Meme coins paired with Tesla stock, hot dog Meme coins paired with Costco stock. How correlated they are in price remains to be seen, but I suppose 'vibes' could be a form of correlation.
AMMs May Ultimately Prevail
Correlated trading pairs are just one piece of the puzzle. The other part is the design and customizability of AMMs.
Uniswap v4 Hooks enable full market customization and significantly improve LP returns. For example, our recently launched DualPool Hook allows idle capital in passive AMMs that isn't being used for trading to earn lending yield.
Although Uniswap has processed about $4.6 trillion in volume, I believe AMMs are still in a very early stage, and there are many ways to make them even more competitive. Both internally at Labs and among our partners and the broader ecosystem, more methods to improve LP returns are being explored. More developments are coming.
In 1976, the argument against index funds was that a fund making no decisions couldn't beat paid professionals dedicated to making investment decisions.
Fifty years later, a 'decision-free' fund has beaten roughly 90% of professional investors. More importantly, index funds democratized investing and improved the lives of ordinary people.
I believe passive liquidity will prevail along a similar path, with even greater impact because it can dramatically lower the barrier to creating and participating in markets.
