Hayden Adams published his first blog in years. Within 48 hours, a former XTX trader replied with a prediction: AMMs are going to zero. This is not a disagreement. It is a collision of two market microstructure paradigms—one built on trustless math, the other on institutional muscle. The stakes are tokenized equities, and neither side has a single data point to back their claim.
Context: The Narrative Clash That Was Always Coming
The debate is simple on the surface. Adams argues that automated market makers (AMMs) will dominate the largest financial markets—stocks, ETFs, index funds—because they enable any trading pair without permission. The former XTX trader counters that professional market making, with its price discovery, inventory management, and risk hedging, renders AMMs structurally incapable of handling high-volume assets like NVIDIA or SPY. The underlying assumption: tokenization of traditional assets is inevitable, and the infrastructure that wins will capture trillions in volume.
But this is not a technical debate. It is a narrative battle fought in public, without a single quantitative benchmark. No one has published the average spread for a tokenized SPY trade on Uniswap v3. No one has modeled the capital efficiency of a concentrated liquidity pool for a low-volatility asset. The entire exchange exists in the realm of qualitative conviction—Adams’s faith in composability, the trader’s faith in professional risk management. Note: The market is overvaluing the tokenization narrative.
Core: The Microstructure Gap That No One Is Talking About
Let’s get specific. The core of the XTX trader’s argument is that AMMs cannot replicate the functions of a dedicated market maker: continuous quoting across multiple venues, delta hedging of correlated positions, and the ability to absorb large block trades without moving the price. In traditional markets, firms like Citadel Securities and XTX deploy algorithms that update quotes in microseconds, manage inventory risk across thousands of securities, and offer price improvement to institutional clients. AMMs, even with concentrated liquidity (v3), rely on passive liquidity providers who rebalance slowly and are vulnerable to adverse selection.
Adams’s counter is that for tokenized assets, the trading pairs are not “NVIDIA vs. USD” but “NVIDIA vs. SPY” or “NVIDIA vs. AAPL.” In a world where every asset is a token, the need for a common quote currency disappears. The AMM becomes a universal exchange—any pair, any time. This is a powerful narrative, but it ignores a critical structural issue: the proving cost of ZK rollups (a stance I have held since the Modular Thesis). If tokenized equities trade on an L2 that relies on zero-knowledge proofs, the settlement costs per transaction could negate the efficiency gains of AMMs. In a high-volume market, even a few cents of proving cost per trade adds up quickly. The XTX trader’s world—low-latency, high-frequency—is antithetical to the batch-processing model of ZK-rollups. Note: Sentiment turning bearish on L2s.
Furthermore, the oracle problem remains unresolved. AMMs require price feeds for valuation. For tokenized stocks, the underlying price is determined by traditional exchanges. Any latency in the oracle creates arbitrage opportunities that drain liquidity. Chainlink’s decentralized oracle network, which I have dissected in previous audits, introduces a centralized trust assumption at the point of data aggregation. For a high-volume asset like NVIDIA, a 1-second delay in the oracle could result in hundreds of dollars of slippage per block. The AMM model is not designed for this precision.
Contrarian: The Blind Spots in Both Camps
The contrarian view is not that AMMs will go to zero, but that the entire tokenization argument is premature. Both participants assume that tokenized equities will be traded in a permissionless environment. The SEC’s Howey Test would classify most tokenized securities as securities. That means the trading venue must comply with Regulation ATS (Alternative Trading System) or register as a national exchange. AMMs, by design, are permissionless—anyone can add liquidity, anyone can trade. This is incompatible with KYC/AML requirements for US investors.
The real blind spot is regulatory convergence. The XTX trader may be correct that AMMs cannot match professional market making, but that misses the point: the market for tokenized assets may never be permissionless. It will likely be a hybrid: a regulated order book for institutional investors, with AMMs serving as a secondary liquidity layer for retail. The former XTX trader’s firm is already exploring on-chain market making tools. They are not waiting for AMMs to fail; they are preparing to build on top of them.
Another blind spot: Adams’s vision assumes that all assets will be tokenized on a single blockchain. In reality, tokenized equities will be issued across multiple chains (Ethereum, Solana, Avalanche) and require cross-chain bridges. The security of those bridges is a known risk. A single bridge exploit could wipe out the liquidity that Adams’s AMMs are supposed to provide. The trader’s dismissal of AMMs as “going to zero” is hyperbolic, but it contains a kernel of truth: in a fragmented multi-chain world, the liquidity per pool is thin, and the risk of black swan events is high.
Takeaway: The Next Narrative Is Not AMM vs. Order Book—It’s Regulatory Scaffolding
The real question is not which technology wins, but which jurisdiction creates the first clear regulatory framework for tokenized equity trading. Singapore, Hong Kong, and the UAE are already moving. The US is paralyzed. The winner will be the infrastructure that can adapt to multiple regulatory regimes—not the one with the most elegant math.
If Uniswap launches a permissioned liquidity pool for tokenized assets that complies with ATS rules, Adams’s narrative will gain credibility. If the XTX trader’s firm launches a regulated on-chain order book, the professional market making model will dominate. Both are possible. The market is pricing in a binary outcome, but the reality is a spectrum.
Note: The market is underpricing the compliance cost of tokenized securities. The next 12 months will not be about AMMs vs. order books. They will be about which team can hire the best regulatory lawyers.
