The SEC is about to hand crypto-native exchanges the US tokenized equities market on lighter terms than Nasdaq operates under. Combined with the institutional market makers now arriving on-chain, the convergence trade pays a specific kind of desk: the one built to run equities, commodities, FX and crypto from a single risk engine.
The story everyone is reading this week is that the SEC is about to release an "innovation exemption" allowing crypto-native platforms to offer tokenized US equities without full broker-dealer or exchange registration. Bloomberg's reporting on Monday 18 May 2026 suggested the framework could land within days. Paul Atkins's Project Crypto, two months in the making, was already signalled in his 21 April speech at the Economic Club of Washington. The exemption itself was not the surprise.
The surprise was one design choice buried in the reporting: the SEC is leaning toward allowing tokens that do not have the consent of the public companies whose shares they track.
If that holds in the final text, the regulatory map has inverted. Nasdaq spent eight months negotiating SEC approval (granted March 2026) to trade tokenized equities on its existing infrastructure, alongside NYSE in April and DTCC's October launch on Canton Network. Those approvals kept tokenized trading inside the existing market structure. The innovation exemption goes the other direction. It permits broader on-chain trading on crypto-native venues, including potentially DeFi protocols, under a lighter framework than the legacy market makers and exchanges have been fighting against since at least November 2025.
SIFMA and Citadel met with the SEC's Crypto Task Force in late January 2026 to argue precisely against this design. Their position, repeated in Citadel's December letter and SIFMA's 26 November submission: securities should not trade under different rules because they sit on blockchain rails. Exemptive relief should only address narrow administrative issues, not categorical exemptions from broker, dealer, or exchange definitions. Brett Redfearn, Securitize's president and former director of the SEC's trading and markets division, framed the third-party issuance question to Bloomberg this week. "If third parties can tokenize Apple or Amazon without the issuer at the table, there's no theoretical limit on how many wrappers of the same company exist at once. This could create a whole new level of market fragmentation and could leave investors less certain what their shares are actually worth at any moment."
The legacy market makers wanted equal treatment. They appear to have lost the argument, at least in the initial draft. This is the news.
Why this matters for microstructure
A non-US retail trader who wants to short Nvidia today has two paths. Path one is Interactive Brokers or a local equity broker: W-8BEN, wire transfer over two or three business days, market hours only, T+1 settlement. Path two is Kraken or Bybit: KYC in five minutes via automated biometric verification, USDC deposit in minutes, short NVDA-x perp at 20x leverage, 24/7, with stablecoins as collateral. That is it.
Ondato reports automated crypto KYC averaging 30 seconds. Entrust documents a major Indian exchange that cut onboarding from 24 hours to 5 minutes after deploying biometric verification. Sumsub runs non-doc verification in 4.5 seconds. For non-US retail and prosumer traders, the crypto-native exchange has been the most accessible global multi-asset venue available for nearly twelve months.
The empirical pattern is already in the listings data. CoinGecko reported that 45% of Kraken's new spot listings in Q1 2026 were RWA or xStocks-related. INTC-x and AZN-x landed in the exchange's top 15 best-performing spot assets. xStocks crossed $25bn in cumulative transaction volume by February, hit 100 tokenized equities in March, targeting 500 by year-end. Payward acquired Backed Finance in December 2025 to bring xStocks issuance in-house, then partnered with Franklin Templeton in May 2026 to integrate the $700m BENJI money-market fund.
The SEC's innovation exemption, if released this week, takes the trade that crypto-native exchanges already won outside the US and hands them the US market on the same terms. That is the inversion. For fifteen years the conventional sequence was: tradfi sets the rules, crypto either complies (slowly, expensively) or operates offshore. Project Crypto runs that sequence backwards. It explicitly recognises the crypto-native venue as the venue and adjusts the regulatory framework to fit, rather than forcing the venue to fit the framework. For non-US markets the regulatory question was already settled (MiCA, VASP licences). Project Crypto closes the US gap.
But the microstructure isn't there yet
The venues exist, the flow is migrating, and the liquidity is not where it needs to be. This is the honest part of the story that the press releases skip.
Ondo published data in August 2025 showing one tokenized version of AAPL with $195 of depth before 2% slippage, spreads at 64 basis points, and a 14.92% premium over the underlying at one point. xStocks have improved meaningfully since then. Gate Research's December 2025 piece noted the six biggest xStocks sit in Kamino with $3.5m in deposits, but documented the negative feedback loop the category is fighting: insufficient liquidity produces poor execution, which deters institutional participation, which leaves liquidity insufficient.
A January 2026 piece on tokenized gold compared PAXG and XAUT perp slippage against CME. Tokenized gold sees slippage rise exponentially with size, hitting roughly 150 basis points at $4m notional. CME gold futures at the same size? Near zero. A $20m ticket clears with less than 3 basis points of impact. Weekend xStocks liquidity drops roughly 70% versus weekday hours despite 24/7 availability. The assets inherit the behavioural rhythm of the underlying equities rather than producing genuinely continuous markets.
This is the gap that determines whether the Project Crypto experiment scales or stays a curiosity. The answer to that gap was published on 5 May 2026, two weeks before the Bloomberg report on the innovation exemption.
The market makers are now arriving
Jump Trading announced a partnership with Securitize on 5 May to make markets in tokenized stocks. Jump provides the liquidity, Securitize provides the issuance infrastructure. Carlos Domingo, Securitize's CEO, framed the goal as broadening the market using crypto-based technology.
Context for why this is the data point that matters. Jane Street produced $39.6bn in net trading revenue in 2025. Citadel Securities posted about $12.2bn. Jump sits in roughly the same tier. These are the firms whose participation defines whether a market has institutional liquidity or not. When one of Jump's caliber commits to making markets in a category, spreads compress, depth deepens, and capital from other top-tier MMs follows. It is the same pattern that played out in crypto perps between 2018 and 2022.
On the venue side, Kraken launched xChange in March 2026, an execution layer explicitly designed to bridge tokenized-equity trading with real-time traditional market depth, with atomic settlement that either fills in full at the quoted price or does not fill at all. xStocks now sits on Solana, Ethereum, and TON. Chainlink CCIP handles cross-chain mobility. The architecture is acknowledging the depth problem and importing tradfi liquidity rather than waiting for native depth to develop.
The trajectory is clear. In 2024 and most of 2025, tokenized equities were retail products with retail-grade liquidity. Through 2026, the venues are increasingly importing tradfi market-making expertise (Jump-Securitize), tradfi liquidity (xChange's bridge to traditional depth), and tradfi-grade infrastructure (DTCC's October launch on Canton Network). The category is being institutionalised in real time. The depth gap to CME is still real. It is closing faster than most tradfi participants expect, and the innovation exemption, by letting crypto-native venues operate in the US without legacy registration friction, accelerates the trajectory rather than slowing it.
Can crypto natives compete for tradfi infrastructure
The honest answer is: in three of the four dimensions, they already are. In the fourth, institutional depth on size, not yet, and possibly not for several years.
Dimension one is regulatory infrastructure. Crypto-native exchanges have spent five years building licensing footprints in roughly every jurisdiction outside the US. MiCA gave EU-licensed CASPs a unified passport. Circle's MiCA-compliant EURC went from 17% to over 50% of the euro stablecoin market in twelve months. This is paid-up work, not future tense. Project Crypto, if it lands as reported, removes the last major gap in that footprint.
Dimension two is execution infrastructure. Crypto exchanges run lower-latency, higher-throughput matching engines than virtually any equity exchange outside the major US venues. FPGA, colocation, low-latency networking. The technology was imported from HFT tradfi a decade ago. On speed and reliability, the gap closed somewhere around 2022.
Dimension three is multi-asset cross-margining. This is where crypto natives are actually ahead of tradfi. A single account holding USDC, USDT, BTC, ETH, tokenized treasuries, and tokenized equities can cross-margin all of it under one risk system. Equity brokers outside the very top tier of prime brokerage do not offer anything close. The structural barrier in tradfi is that equities, futures, FX, and treasuries grew up on separate rails and never got unified. Crypto exchanges built unified from day one because they had no legacy to preserve.
Dimension four is institutional depth. This is the gap. XAUT perps at $4m notional move 150 basis points. The depth that real institutional flow needs is not there yet, and importing it (Jump-Securitize, xChange) takes time. For retail and prosumer flow up to a few hundred thousand dollars per ticket, the existing depth is workable. Above that, traders still route to CME or to traditional venues for now. The innovation exemption changes who can compete to fill that gap, but it does not fill it by itself.
What tradfi convergence pays the crypto-native desk
For desks already operating with crypto-native infrastructure, multi-venue execution, multi-chain settlement, stablecoin treasury management, cross-margined risk, the combination of the innovation exemption and the market-maker arrival creates a set of strategies that were not viable eighteen months ago.
Basis trades are the cleanest. Tokenized equity perp funding rates are set by long/short imbalance on the tokenized venue, not by the underlying equity's borrow cost or implied repo. Those two numbers diverge often. A desk that can short tokenized NVDA perp while long the underlying via cash equities, or the reverse, at scale harvests that spread for as long as the venues remain fragmented from the cash markets. The equivalent trade in tradfi requires a prime broker, equity financing, and an exchange-traded derivatives account. In crypto-native infrastructure it is one account, cross-margined, with stablecoin collateral.
Calendar arbitrage extends the same logic to time. xStocks trade on chain through weekends and holidays, the underlying does not. Every Friday close into Monday open, every earnings print after the bell, every macro event landing on a Sunday creates a tradeable spread. Crypto-native desks are already running this. Tradfi market makers have not yet built the infrastructure to hold tokenized inventory or execute against an on-chain order book at 3am Sunday.
Cross-asset basis is the larger prize. Tokenized gold versus CME gold futures, with funding mechanics on the tokenized leg that respond to crypto liquidity conditions rather than precious-metals conditions. Tokenized S&P index perps versus CME E-mini futures. EURC/USDC on chain versus EUR/USD spot. Each is a known basis trade with new mechanics on one leg and dramatically lower capital requirements end-to-end. The funding cost of running a tokenized-perp versus CME-futures basis is materially below the same trade in tradfi, because tokenized treasuries posted as collateral earn roughly 5% while serving as margin.
Fragmentation across chains is the persistent edge. The same economic exposure trades at different prices across Solana, Ethereum, TON, and BNB Chain with bounded bridging costs. Multi-venue, multi-chain arbitrage is exactly what crypto-native trading infrastructure has been built for since 2017. Equity market makers have no operational footprint for it.
And the one most desks miss: collateral efficiency. A desk holding $10m in tokenized treasuries earns $400-500k a year in treasury yield and posts the same $10m as margin against an active trading book. For any strategy running at single-digit-percentage annual returns, that is the difference between viable and not. Tradfi prime brokers offer similar arrangements with higher minimums, more friction, and worse terms outside the largest funds. Crypto-native infrastructure offers it to anyone with a verified institutional account.
The innovation exemption, if it includes the third-party issuance provision, opens one more strategy class. Multiple tokenized wrappers of the same underlying (different issuers, different chains, different funding mechanics) will trade at slightly different prices. Desks that can monitor and arbitrage across that wrapper fragmentation will harvest spreads that do not exist in any other equity market. Redfearn's concern about market fragmentation is, from a trading-desk perspective, the opportunity description.
What the bitcoin equity correlation actually tells you
While the venue convergence has been playing out, bitcoin's 30-day rolling correlation with the S&P 500 hit 0.74 in March 2026, the highest of the year, with intraday r-squared touching 0.94. BTC fell from $126k to around $67k, tracking equity weakness almost tick-for-tick.
The mechanism is documented. CME research from May 2025 shows the 60-day correlation has run in a 0.0 to 0.6 band since early 2020, with a structural break at the January 2024 spot ETF approval (confirmed via DCC-GARCH and Chow tests in a December 2025 arXiv paper). US bitcoin ETF AUM reached $164-179bn by mid-2025, with nearly half of bitcoin's volume now flowing through institutional products. A January 2025 survey found 59% of institutional investors targeting more than 5% crypto allocations. The same money drives both books. When liquidity tightens, both sell.
The conventional read is that this kills the digital-gold thesis. The more useful read is that crypto and equities now sit inside the same institutional liquidity envelope at the capital-allocation level, which is exactly what the venue convergence is producing at the microstructure level, which is exactly what the innovation exemption is about to codify at the regulatory level. Three trends, one underlying movement. If you accept that bitcoin and the S&P now share a liquidity envelope, then the existence of venues where they cross-margin against each other is not a curiosity. It is the operational expression of the financial reality. And the SEC is now poised to formally recognise that operational expression as a legitimate way to trade US securities.
The desks positioned for the next cycle are not choosing between tradfi and crypto. They are running both inside one balance sheet, one risk engine, one custody footprint, one execution stack. That is where the structural margin lives, in being able to express any view across equities, commodities, FX, treasuries, and crypto without ever leaving the same risk system. It is also, not coincidentally, what Bequant has been built to deliver.
Tradfi is coming on-chain. The regulatory map is inverting, the flow is moving, the market makers are arriving. What that pays depends entirely on whether your infrastructure was built for the venues capturing that flow, or for the ones still arguing about whether they should be allowed to compete.