Uniswap founder Hayden Adams has laid out a broad case for how tokenization could change market structure, arguing that the bigger shift is not just faster settlement or round-the-clock trading, but a rewrite of who makes markets and what assets should trade against.
In his essay, Correlated Pairs: How AMMs Win the Biggest Markets, Adams says tokenized markets built on a shared blockchain settlement layer could move away from the usual dollar-centered design of traditional finance. Under that model, NVIDIA stock would not necessarily trade mainly as NVDA/USD. It could develop a direct NVDA/SPY market instead. Oil company stocks could trade against oil ETFs or tokenized crude, while private credit could pair with tokenized U.S. Treasury funds.
Correlation sits at the center of the thesis
Adams argues that when two assets are highly correlated, liquidity providers take on less inventory risk. As that risk falls, the efficiency gap between passive automated market makers and professional market makers narrows.
His conclusion is that a large share of liquidity in correlated assets could eventually be handled by passive AMMs, while only a smaller set of bridge pairs connecting different asset clusters would still demand heavy involvement from sophisticated professional market makers.
He says the structure is already emerging onchain.
A parallel with index funds
Adams opens with the 50th anniversary of index funds this year. In 1976, when Jack Bogle launched the index fund, he had hoped to raise $150 million but brought in only $11.3 million. Rivals mocked the idea as “Bogle’s folly,” arguing that a fund making no active decisions could not beat professional investors charging high fees to pick stocks.
Fifty years later, passive investing has become one of the dominant forces in U.S. asset management. Adams says tokenization may now be going through its own folly moment.
He writes that markets often frame tokenization as a financial infrastructure upgrade: the same stocks, bonds, and funds moved onto blockchains, where they can trade faster, cheaper, and 24/7. In his view, that framing misses the larger point. “Tokenization makes markets programmable.” Once assets share the same blockchain settlement layer, they no longer have to follow the limits imposed by legacy financial plumbing, and the U.S. dollar no longer has to serve as the intermediary in nearly every market.
Uniswap volume and the rise of DEX trading
Adams founded Uniswap in 2018 with the idea that market making could be handled by code. The basic design is simple: anyone can deposit two assets into a shared liquidity pool, traders move prices along an algorithmic curve when they buy or sell, and liquidity providers collect fees from each transaction.
According to Adams, Uniswap has operated autonomously since launch and has now processed more than $4.6 trillion in cumulative trading volume. Over the same period, decentralized exchanges grew from less than 1% of centralized exchange spot volume to more than 20%.
Still, he says the more interesting story is not volume alone. It is the way liquidity has started to organize itself around correlated pairs.
Where AMMs first found product-market fit
Adams breaks AMM development into stages. The earliest clear product-market fit came in long-tail assets that traditional professional market makers had little interest in serving. For a professional market maker, trading systems, inventory management, hedging, and settlement all carry fixed costs. Small assets often do not generate enough volume to justify those costs.
AMMs changed that. Anyone can launch a market with a single onchain transaction, and token issuers or early supporters can become the first LPs themselves.
The next stage, in his telling, was stablecoin pairs. Using USDC/USDT as an example, Adams says passive LP strategies can come close enough to the efficiency of active professional market making when the two assets are tightly correlated. If LPs also require lower returns on capital than trading firms do, they can offer liquidity at lower cost.
That is why, he argues, professional trading firms rarely need to fight passive AMMs for stablecoin-to-stablecoin markets today.
Why traditional market making looks entrenched
In traditional finance, Adams says, market makers usually keep capital, strategy, execution technology, settlement, and distribution inside one firm. That vertical structure emerged for historical reasons. Different asset classes sit in different systems, settlement is slow, and each layer needs dedicated infrastructure. Large institutions became the most efficient way to tie those functions together, with scale helping spread fixed costs.
He points to Citadel Securities as an example, saying the firm handles about 25% of U.S. equity trading volume and generated a record $12.2 billion in net trading revenue last year using roughly $21 billion of trading capital.
Many observers might read those figures as a sign of market efficiency. Adams does not. “I read them as entrenchment,” he writes, arguing that they reflect high barriers to entry in legacy markets.
Blockchain unbundles the market maker
For Adams, the key change introduced by blockchains is that they separate the functions that used to be bundled inside a traditional market-making firm. Trade execution can move to smart contracts. Custody and settlement become shared public services. Infrastructure that once had to be built as proprietary systems can be replaced by open-source software.
In that kind of market, he says, capital becomes the scarce input that matters most. Competitive advantage shifts toward whoever can hold inventory at the lowest cost.
Professional trading firms need returns high enough to pay for staff, technology, and infrastructure. A liquidity provider willing to accept a lower yield can squeeze that margin. The same goes for hedging. Professional market makers generally do not want directional price exposure, so they pay to hedge it away. If an LP already wants to own the assets, that cost can disappear.
For token issuers, Adams says the cost of capital can even be negative, because issuers already hold large amounts of their own asset and, under the old model, may have to pay market makers to support liquidity.
Why crypto markets already cluster around ETH, SOL, and stablecoins
Adams says a recent conversation with a large global financial institution led to a simple question: what are the most common base pairs in DeFi?
His answer was that Ethereum ecosystem assets tend to trade against ETH, Solana ecosystem assets tend to trade against SOL, and stablecoins tend to form pairs with each other. A smaller number of highly liquid pairs then connect those separate clusters.
No one designed that structure from the top down, he says. It emerged naturally because LPs face less inventory risk when the two assets they hold move closely together, making them more willing to supply liquidity.
He argues that if stocks, bonds, commodities, and other real-world assets are widely tokenized, the largest financial markets in the world could reorganize along the same lines.
From NVDA/USD to NVDA/SPY
Adams argues that the fact nearly every asset in traditional finance is quoted against the dollar is not necessarily proof that this is the most efficient format. In large part, he says, it is a consequence of market plumbing. Stocks, commodities, and bonds live in separate systems, connected by institutions and rails such as SWIFT, Fedwire, and the banking system.
Blockchains offer a different setup. If assets are tokenized and live on a common settlement layer, any two assets can in theory form a direct market.
His example is a progression from NVDA/USD to NVDA/SPY to SPY/USD. Since NVIDIA and the S&P 500 ETF are more correlated with each other than NVIDIA is with the dollar, the relative price movement in NVDA/SPY could be lower.
He extends the same logic to other pairings:
- oil company stocks with oil ETFs or tokenized crude
- private credit with tokenized Treasury funds
- single-industry stocks with sector ETFs
These kinds of economically related cross-asset pairs are difficult to build, and often not economically viable, in traditional financial infrastructure, he says.
Delta-neutral market making carries its own cost
Traditional market makers generally try to stay delta neutral, meaning they minimize directional exposure to non-dollar assets on a balance sheet measured in dollars. If a market maker holds NVIDIA stock inventory, it may need options, futures, or other tools to hedge, and that hedge costs money.
Adams argues that in a more correlated market such as NVDA/SPY, an investor who already wants to own both NVIDIA and SPY can provide liquidity without rushing to neutralize every bit of price exposure in the way a professional market maker would.
That matters because the market maker already wants the underlying assets. In his view, this can become a major cost advantage. The stronger the correlation between the two assets, the smaller the efficiency gap between passive AMMs and the most advanced active market-making strategies, making it easier for low-cost LPs to undercut professionals.
The dollar does not disappear, but it becomes a bridge
Adams is not arguing that investors will stop buying NVIDIA with dollars. If many stocks build direct pairs with SPY, then trades starting in dollars or ending back in dollars could route automatically through Stock → SPY → USD.
That would turn SPY/USD into a bridge pair carrying heavy volume. In this structure, low-cost passive LPs could handle most correlated pairs, while a smaller number of bridge pairs connecting separate clusters would still justify complex strategies from professional market makers because they carry larger volume and more risk.
He says DeFi already shows this pattern. ETH/USDC is one of the deepest liquidity markets onchain because many assets and trading routes eventually pass through it. The same logic points to a future where passive LPs handle correlated pairs and active professional LPs compete in bridge pairs.
Early experiment on Robinhood Chain
Adams says the thesis is no longer purely theoretical. On Robinhood Chain, Uniswap pools now include 10 tokenized stocks paired directly with SPY.
Across their first 12 days, those markets generated $33 million in trading volume and drew more than 11,000 traders. He says a meaningful share of that activity took place while U.S. equity markets were closed.
Some traders have already gone directly from Stock A to Stock B without touching the dollar at all.
He also points to more extreme versions of “correlation trading” appearing onchain. Traders have paired Elon Musk-themed meme coins with Tesla stock, and even linked a hot dog meme coin with Costco stock. Adams jokes that he is not sure how correlated those prices really are, but maybe “vibes” count as a kind of correlation too.
Uniswap v4 Hooks and idle LP capital
Adams also says AMMs themselves still have a lot of room to improve. One of the tools he highlights is Uniswap v4 Hooks.
Hooks let developers customize the trading logic of a pool. He says Uniswap Labs recently introduced a DualPool Hook that allows passive capital sitting in an AMM to be deployed into lending markets for extra yield when it is not being used for swaps.
That means the same LP capital no longer has to sit idle while waiting only for trading fees. It can also increase capital efficiency at the same time.
Even after roughly $4.6 trillion in cumulative volume, Adams says he still sees AMMs as being in a very early stage.