Hayden Adams posted his first blog since 2019. Within 48 hours, a former XTX trader fired back. The question: can AMMs handle tokenized stocks? The debate is not academic. It defines the next infrastructure layer for crypto.
I’ve spent the last five years auditing smart contracts and implementing zero-knowledge proofs. I’ve seen AMMs thrive in volatile, long-tail assets. I’ve also seen them fail under high-frequency, large-order flows. The tokenized asset market is not a theoretical playground. It’s where two techno-economic paradigms collide.
Context: The Tokenized Asset Thesis
Hayden Adams argues that AMMs will win the biggest markets—tokenized stocks, ETFs, index funds. The logic: in a world where every asset is a token, the “quote in USD” paradigm breaks. AMMs natively support any trading pair. Uniswap v3’s concentrated liquidity can theoretically provide tight spreads for liquid assets like NVIDIA or SPY. This is the core of the “AMM as universal settlement” narrative.
But the former XTX market maker disagrees. He says AMMs will go to zero. His point: professional market making is about price discovery, inventory management, and risk hedging. A constant product formula cannot replicate that. He asks: “Who would sell NVIDIA to buy SPY?” The question undermines the entire use case.
Core: The Code-Level Breakdown
Let’s look at the math. Uniswap v3 uses a concentrated liquidity model where LPs provide liquidity within a price range. For a tokenized stock like NVIDIA, the price is tied to the real-world market. The AMM must rely on an oracle to update the price. This introduces a dependency: the oracle’s correctness and timeliness. I’ve audited oracle-based protocols before. The 2021 LUNA crash taught me that a single oracle failure can cascade into a death spiral. Math doesn’t negotiate. The AMM’s invariant is fixed. The real world moves faster.
Professional market makers, on the other hand, use algorithms that incorporate latency, order book depth, and cross-asset hedging. They can quote tighter spreads because they manage risk across multiple venues. An AMM cannot dynamically adjust its inventory based on volatility. It uses a static fee structure. For large orders, the price impact is catastrophic.
I built a minimal zkSNARK proving system in Rust during the 2022 bear market. That experience taught me that cryptographic proofs are powerful but slow. AMMs are fast because they are simple. But simplicity is a double-edged sword. In a high-volume market, the AMM’s simplicity becomes a bottleneck. Code is law, but bugs are reality. The AMM’s code is audited, but the economic model is not tested at scale for tokenized assets.
Contrarian: The Blind Spot No One Talks About
The debate misses the real issue: compliance. Tokenized stocks are securities. In the US, trading them on an unlicensed AMM violates the Securities Exchange Act. The Howey test applies to every tokenized share. The AMM’s permissionless nature is a liability, not a feature. Professional market makers have the licenses and relationships to operate within the regulatory framework. They can offer ATS (Alternative Trading System) services. The AMM cannot.
I audited a custodial wallet solution for a major asset manager in 2024. The key-shares distribution protocol had a critical flaw. But the bigger problem was the legal gap: the system claimed to be decentralized, but the settlement still required a trusted custodian. Tokenized assets face the same issue. The AMM executes the trade, but the actual stock transfer happens on a centralized registry. This fragmentation creates counterparty risk that no smart contract can fix.
Takeaway: The Future is Hybrid
The AMM will not vanish. The market maker will not be replaced. The real outcome is a hybrid: AMMs provide baseline liquidity, and professional market makers layer on top using Uniswap v4 hooks or RFQ mechanisms. The winner will be the infrastructure that bridges the two worlds. I’m watching for Uniswap’s next move—a tokenized asset pool with a permissioned hook. That would be a signal. Until then, the debate is a distraction. The code will decide.