Last Monday, Hayden Adams broke a six-year silence. The Uniswap founder posted a blog arguing that Automated Market Makers — the same DeFi primitive that powered 2020's yield farming mania — will eventually dominate the world's largest markets: tokenized equities, ETFs, and index funds. Within 48 hours, a former XTX Markets trader fired back. His rebuttal was blunt: AMMs are going to zero in those markets. Zero. Not a squeeze. Not a correction. Zero.
That's not a debate. That's a declaration of war between two tribes. One tribe believes code can replace Wall Street's plumbing. The other believes that code is a toy until you've managed a billion-dollar book in a flash crash. I've been in both trenches. I've audited Solidity contracts that looked beautiful until a reentrancy attack drained them in three blocks. I've also traded options on IBIT, using Greeks to arbitrage retail FOMO. From where I stand, both sides are wrong — and both sides are right. The truth is somewhere in the cold, hard spread between their assumptions.
Context: The Battlefield
What's actually being argued here? Adams claims that AMMs are the natural infrastructure for a world where every asset is a token. In that world, you don't need a central order book to match buyers and sellers. You just need a pool of liquidity that lets anyone swap any two tokens — say, tokenized NVIDIA shares for tokenized SPY units. The pool's price curve does the matching. No human, no broker, no SEC-mandated exchange. Just math.
The former XTX trader counters with a question that cuts to the bone: "Who would want to sell their NVIDIA for SPY?" He's implying that the demand for such cross-asset swaps is a fantasy. In the real world, investors don't trade NVIDIA for SPY. They trade NVIDIA for dollars, and then dollars for SPY. The dollar is the numeraire. The order book is the king. The AMM is just a casino for shitcoins.
But that's too simplistic. The trader's real concern — and I've seen this firsthand in 2020 when I pulled my Uniswap V2 liquidity just before a flash loan attack — is that AMMs lack the microscopic control that professional market making requires. In an order book, a market maker can adjust quotes on a millisecond basis, hedge delta, manage inventory, and absorb large orders without moving the price. In an AMM, the price is a function of the pool's reserves. A large trade creates massive slippage. That's fine for a $1000 swap of a memecoin. It's catastrophic for a $10 million institutional trade of tokenized Apple stock.
Core: The Order Flow Code
Let's get technical. The core of this debate is market microstructure — the invisible mechanics of how orders become trades. I've spent years analyzing this from both sides: first as a cybersecurity student reverse-engineering vulnerable smart contracts, then as an options strategist pricing volatility on Bitcoin ETFs. The pattern is always the same. The winning infrastructure is the one that minimizes adverse selection for liquidity providers while maximizing execution quality for takers.
In an AMM, adverse selection is brutal. When a knowledgeable trader sees that a pool's price is stale (because the off-chain price has moved faster than the on-chain oracle update), they can front-run the pool. The liquidity provider gets picked off. This is why Uniswap V3 introduced concentrated liquidity — to let LPs choose their price ranges. But even that doesn't solve the fundamental latency problem. On-chain time is measured in seconds. In the equities world, high-frequency traders measure time in microseconds. An AMM cannot compete on speed. It can only compete on permissionless access and composability.
Here's the insight that both sides are missing: The debate isn't really about AMM vs. order book. It's about the assumption that tokenized assets will trade like crypto tokens. They won't. Tokenized SPY is not a new asset class. It's a representation of an existing security that settles through the DTCC. The underlying ownership is still tracked by a centralized custodian. The token is just a wrapper. And that wrapper introduces a new layer of risk: the custodian's solvency, the minting/burning mechanism, the regulatory compliance. The AMM is just the top layer of a stack. The real liquidity sits in the settlement layer, which is still TradFi.
From my experience during the 2022 Terra collapse, I learned that when the foundation is weak, the trading layer doesn't matter. Terra's liquidity pools on Curve were deep right up until the moment they weren't. The code was beautiful. The incentives were aligned — until the risk was repriced. The same will happen with tokenized assets if the underlying custody and compliance are not bulletproof. The AMM will be blamed for the failure, but the failure will be in the wrapper.
Contrarian: The Blind Spots
Now for the counter-intuitive angle. The conventional wisdom among crypto natives is that AMMs are the future because they are decentralized and permissionless. The conventional wisdom among TradFi pros is that order books are the future because they are efficient and professional. Both are looking at the wrong dimension.
The real blind spot is regulatory infrastructure. The former XTX trader isn't just a market maker. He's a representative of a world where compliance is a competitive advantage. The XTX's of the world have legal teams, broker-dealer licenses, and relationships with the SEC. When they say AMMs are going to zero, they're not just talking about slippage. They're talking about the fact that a permissionless liquidity pool for tokenized NVIDIA shares is likely an unregistered securities exchange under U.S. law. The AMM doesn't know who is trading. It can't enforce KYC/AML. It can't freeze assets if a regulator demands it. That's not a bug — it's a feature. But it's a feature that makes it illegal for the largest market in the world.
Adams' vision only works if the regulatory framework changes. But the SEC is not going to legalize anonymous trading of securities just because Uniswap's code is elegant. The pushback from the former XTX trader is a signal that the TradFi establishment is preparing to enter the tokenized asset space themselves — but on their terms. They will build licensed, KYC-compliant trading venues for tokenized assets, probably using a hybrid model: order books for large trades, with AMMs providing passive liquidity for small trades. The AMM will not win the biggest markets. It will be subsumed into them.
I've seen this pattern before. In 2024, when the Bitcoin ETFs launched, the most successful trading strategies were not pure crypto-native plays. They were hybrids that used on-chain data to inform off-chain derivatives trades. The ETF options market on IBIT was a prime example. The institutional players used their traditional infrastructure while the crypto-native traders were left chasing retail flow. The liquidity stayed cold for pure DeFi strategies. The same will happen with tokenized equities. The AMM will be a tool, not the platform.
Takeaway: The Silence After the Leverage
So where does this leave us? The debate between Adams and the XTX trader is a distraction from the real work ahead. The winning architecture for tokenized assets will not be pure AMM or pure order book. It will be a layered system: a compliant issuance layer, a settlement layer that talks to the DTCC, and a trading layer that combines AMMs for passive liquidity with RFQ (request for quote) systems for large block trades. The AMM's role will be to provide the initial liquidity for new tokenized assets, lowering the barrier to entry. But as the market matures, the order book will take over the high-volume pairs.
The code will bleed, but the liquidity will stay cold. The institutions will not trust a permissionless pool for their billion-dollar positions. They will trust a licensed venue that uses an AMM as a backend for small orders. The incentives will align only when the risk is priced in — and the risk of regulatory action is not priced into any current AMM model.
Watch for Uniswap's next move. If they announce a compliance layer, a permissioned pool, or a partnership with a regulated broker-dealer, then Adams is positioning for the hybrid future. If they stay pure, the XTX trader will be proven right. The AMM will not win the biggest markets. It will be regulated out of them. The silence after the leverage snaps will be the sound of a thousand founders realizing that code alone cannot override the law.
I don't have a crystal ball. But I have a scar from every trade I've taken based on narrative rather than infrastructure. The debate is interesting. The execution is what matters. And right now, the execution of tokenized asset trading is still in the hands of the people who own the licenses. The AMM is just a tool in their toolbox. They will use it, but they will not be ruled by it.