The most insightful debate about automated market makers this year didn't happen on a stage—it unfolded in a blog post and a reply. Hayden Adams, Uniswap's creator, declared that AMMs will conquer the world's largest markets. A former XTX Markets trader responded with a single, devastating sentence: 'They're going to zero.'
Logic does not bleed, but code leaves traces. The XTX trader's retort came within 48 hours—a signal that the traditional market-making establishment is already watching this space. As someone who has spent years reverse-engineering DeFi exploits, I recognize this pattern: when a narrative shifts from 'what if' to 'who will win,' the real battle is not about technology but about market microstructure.
Context: The Battlefield
The debate centers on tokenized assets—stocks, ETFs, index funds—moving onto blockchains. Adams argues that AMMs will become the default trading infrastructure for these assets because they natively support any trading pair without needing a dollar-denominated quote. The XTX trader counters that professional market makers, with their sophisticated pricing, inventory management, and risk hedging, will always outperform AMMs in high-volume, liquid markets like NVIDIA or SPY.
This is not a new argument. I've seen it before, in 2020 when DeFi summer peaked. Back then, the question was: Can AMMs handle volatile crypto pairs? They did. But tokenized stocks are a different beast. They are low-volatility, high-liquidity, and regulated. The technical requirements are not the same.
Core: A Systematic Teardown
Let's break down the arguments into their component parts.
Adams' Thesis: AMMs eliminate the need for a base currency (USD) and allow any two assets to trade directly. In a world where stocks, bonds, and ETFs are all tokenized, this frictionless interchangeability is a structural advantage. Uniswap v3's concentrated liquidity already demonstrates that AMMs can optimize capital efficiency in specific price ranges.
The XTX Trader's Counter: Professional market making is about more than providing liquidity. It's about price discovery, risk management, and handling large orders without moving the market. A constant function formula cannot replicate the real-time decision-making of a human trader or a sophisticated algorithm. The trader's pointed question—'Who would want to sell NVIDIA to buy SPY?'—underscores a deeper issue: the demand for such pairs may not exist in the first place.
From my own work auditing DeFi protocols, I've seen that AMMs excel in two scenarios: high-volatility assets (where impermanent loss is acceptable) and long-tail assets (where liquidity is thin). Tokenized blue-chip stocks like NVIDIA fall into neither category. They are low-volatility, high-liquidity assets that require tight spreads and deep order books. The constant function approach introduces slippage that professional traders will not tolerate.
The Missing Data: Both sides rely on qualitative arguments. Neither provides quantitative evidence—no slippage comparisons, no volume data, no capital efficiency metrics. This is a red flag. When I analyzed the 2021 NFT bubble, I found that 60% of 'blue chip' volume was wash trading. The parallels are clear: narratives without data are noise.
The Regulatory Elephant: The article’s analysis flags a critical point: tokenized securities, if traded via AMMs, would likely violate U.S. securities laws. The Howey test applies to every step—issuance, trading, settlement. AMMs are permissionless, but securities trading requires KYC/AML and licensed intermediaries. The XTX trader, coming from a regulated environment, understands this implicitly. Adams' silence on regulation suggests that Uniswap may focus on jurisdictions with lighter frameworks, like Singapore or Hong Kong, before tackling the U.S.
Contrarian: What the Bulls Got Right
Despite the skepticism, the bulls have a valid point: AMMs are not static. Uniswap v4 introduces 'hooks' that allow for customizable liquidity management, potentially enabling features like limit orders or dynamic fee structures. This could bridge the gap between AMMs and order books.
Moreover, the XTX trader's dismissal ignores the long-tail market. Not all tokenized assets will be NVIDIA or SPY. There will be thousands of niche assets—real estate tokens, private equity shares, carbon credits—where liquidity is inherently thin. AMMs are ideal for these markets because they bootstrap liquidity without requiring professional market makers.
Based on my experience dissecting the Terra/LUNA collapse, I know that algorithmic mechanisms can fail spectacularly under stress. But AMMs are not algorithms in the same sense. They are passive, transparent, and auditable. The failure mode is not a death spiral but a liquidity crisis. In a regulated environment, that can be mitigated with circuit breakers or managed liquidity pools.
Takeaway: The Future Is Hybrid
The rug is not pulled; it was never tied. The debate is not about which technology wins but about how they coexist. The most likely outcome is a hybrid infrastructure: AMMs for baseline liquidity and long-tail assets, and professional market makers (via RFQ or order books) for high-volume, low-spread trading.
Gas fees are the price of truth. The truth here is that both sides are right in their domains. The XTX trader is correct about the limitations of AMMs for liquid markets. Adams is correct about the potential for AMMs in the long tail. The market will decide, but only after we have data—real on-chain metrics comparing slippage, spread, and finality across both models.
Until then, treat every narrative as a hypothesis. Check the contract, not the influencer. And remember: imagination is infinite, but liquidity is finite.
