Lighter vs Hyperliquid: two regulatory bets on US perps
Quote1,249 words
Argues the Clarity Act is not the perp switch, since perps were always CFTC territory and the US unlock is agency-led. Lighter licenses the venue, while Hyperliquid keeps the protocol neutral and needs the CFTC to let a registered DCM run on onchain rails.
Key claims
Fiduciary capital (RIAs, ERISA plans, 40 Act funds) is the pool that cannot touch an unregistered venue, not prop money
Hyperliquid Policy Center and Phantom asked the CFTC to let a registered DCM use an onchain protocol as its matching layer
HL has 20x Lighter's revenue and over $11B OI
Good take by my fren @Louround_ and he's asking the right questions.
Want to add some nuance on the regulatory mechanics tho, because that's where most takes on this trade blur things together.
饾煭. "饾棖饾椆饾棶饾椏饾椂饾榿饾槅 饾棓饾棸饾榿 饾椊饾棶饾榾饾榾饾棽饾榾" 饾椂饾榾 饾椈饾椉饾榿 饾榿饾椀饾棽 饾椊饾棽饾椏饾椊 饾榾饾槃饾椂饾榿饾棸饾椀 饾椊饾棽饾椉饾椊饾椆饾棽 饾榿饾椀饾椂饾椈饾椄 饾椂饾榿 饾椂饾榾.
Perps were always CFTC territory. Retail derivatives having to flow through regulated venues is Dodd-Frank era law, not something the Clarity Act introduces or changes, and the US perp unlock is already happening without it: Coinbase has listed perp-style futures since July 2025 + got a no-action letter to route US customers to Deribit, while the CFTC also approved Kalshi's BTCPERP in May. The Clarity Act matters as the statutory umbrella (plus the DeFi carve-outs), but the perp door is being opened by the agency, venue by venue, for registered entities. "clarity passes and 28m americans trade perps" compresses two separate tracks into one event.
饾煯. 饾棯饾椀饾棶饾榿 饾棶饾棸饾榿饾槀饾棶饾椆饾椆饾槅 饾椀饾棶饾椊饾椊饾棽饾椈饾榾 饾榿饾椉 饾棝饾槅饾椊饾棽饾椏饾椆饾椂饾椌饾槀饾椂饾棻 饾椂饾棾 饾椂饾榿 饾椈饾棽饾槂饾棽饾椏 饾椏饾棽饾棿饾椂饾榾饾榿饾棽饾椏饾榾? 饾棦饾椈 饾棻饾棶饾槅 饾椉饾椈饾棽: 饾椈饾椉饾榿饾椀饾椂饾椈饾棿.
US persons are already restricted. Nothing that's legal today becomes illegal, and enforcement in this space has historically gone after operators, not retail users. That means the pressure on HL is competitive, not punitive. Or more specifically: it's about fiat rails, brokerage UX, tax docs, and (admittedly) most importantly institutional and advised capital that legally cannot touch an unregistered venue (nothing new tho and not a change that the Clarity Act will introduce).
And to be precise about what that capital is: prop money is already there. Crypto funds, MMs, HFT shops are all running through offshore entities with mandates that allow it. The pool that structurally cannot touch an unregistered venue is fiduciary capital. RIAs bound by the SEC custody rule, pension money under ERISA prudence standards, 40 Act funds, and any FCM or broker handling client accounts. That money doesn't compare on the level of orderbooks or UX, bc the legal wrapper kills any attempt before the investment committee ever sees it kek. That is indeed a huge market (US retirement assets alone sit at >$47T per the latest ICI data), but it's not money that "is on HL now and then leaves".
Outside of that highly regulated capital tho, nobody switches venues just for the sake of it while HL has the deepest books and the best onchain UX. But also outside of that segment, there could be an argument around HL forgoing the marginal cohort if they don't embark on the US path too. What I mean: the next wave of US retail that onboards where it's legal, and institutional size that needs a registered counterparty. As mentioned above tho, that's again more of a potential growth ceiling, not an existential threat.
The actual existential vector is different imo: CME and ICE lobbying to force registration onto HL, with CME literally suing the CFTC over the perp approvals. The real fight is whether the US framework fully extends to onchain protocols or stays CEX-only.
饾煰. 饾棩饾棽饾棿饾椂饾榾饾榿饾椏饾棶饾榿饾椂饾椉饾椈 饾棻饾椉饾棽饾榾 饾椈饾椉饾榿 饾椀饾棶饾槂饾棽 饾榿饾椉 饾椀饾棶饾椊饾椊饾棽饾椈 饾棶饾榿 饾榿饾椀饾棽 饾椂饾椈饾棾饾椏饾棶 饾椆饾棽饾槂饾棽饾椆. 饾棫饾椀饾棶饾榿'饾榾 饾榿饾椀饾棽 饾棽饾椈饾榿饾椂饾椏饾棽 饾椊饾椉饾椂饾椈饾榿 饾椉饾棾 饾榿饾椀饾棽 饾棸饾槀饾椏饾椏饾棽饾椈饾榿 饾棖饾棛饾棫饾棖 饾棾饾椂饾棿饾椀饾榿.
Right now, Lighter and HL are taking different approaches:
- Lighter: License the venue. Delaware-based incorporation + pending CFTC application, and per Novakovski that license would cover front-ends like Robinhood sitting on top of it.
- Hyperliquid: Keep the protocol neutral, put licensed wrappers on top. Phantom already got a CFTC no-action letter saying a non-custodial front-end connecting users to registered venues is not an introducing broker. That solves the interface half of the stack. Only half tho, because the venue at the end of the pipe still has to be registered, and HL isn't (which is why Phantom's own HL integration stays blocked for US users). So last week the Hyperliquid Policy Center + Phantom formally asked the CFTC to solve the other half: (a) confirm that publishing protocol software is not operating an exchange and (b) let a registered DCM use an onchain protocol as its matching and execution layer. That second ask is the whole game. If it lands, a licensed entity (think @tradexyz, @phantom, or someone new) holds the registration while HyperCore stays the neutral engine underneath. Exactly the credibly neutral financial infra layer HL always aspired to be.
However, to be clear on feasibility: Under today's rules this path does not exist. The CEA requirement sits at the transaction level, US retail derivatives must trade on a registered venue, and licensed front-ends can't replace an unregistered engine. The Clarity Act doesn't change that either. The DeFi carve-outs exempt people (devs, validators, UI providers) from registration, not trades from the venue requirement, and the perp rulebook lives in the existing CEA that Clarity mostly leaves untouched. So the wrapper model needs the CFTC to act further: guidance or rulemaking confirming a registered DCM can run on rails it doesn't control. The HPC letter itself names the open blockers: system safeguards, segregation, recordkeeping, all written for custodial systems.
Also note that @HyperliquidX itself registering was never on the table anyway. A DCM has to enforce rules, halt trading, control its systems. Hence, any entity registering on HyperCore's behalf would need control over the protocol, which breaks the entire design. Which ultimately means the endgame is a binary outcome: either the CFTC greenlights licensed venues on neutral rails (Selig has said on record he wants to onshore onchain perp markets like Hyperliquid, and the HPC argues this sits within the agency's existing authority, no new law needed), or there is no compliant US retail path for HL at all (which is no death sentence either tho, as explained previously). Lighter's model works within today's rules. HL's model needs new ones. That's the real asymmetry in this trade, and it's exactly the rule CME and ICE are lobbying to prevent.
饾煱. 饾棦饾椈 "饾棩饾椉饾棷饾椂饾椈饾椀饾椉饾椉饾棻 饾椃饾槀饾榾饾榿 饾棷饾槀饾椂饾椆饾棻饾榾 饾椂饾榿饾榾 饾椉饾槃饾椈 饾棽饾椈饾棿饾椂饾椈饾棽": 饾棶饾棿饾椏饾棽饾棽 饾椉饾椈 饾榿饾椀饾棽 饾椂饾椈饾棸饾棽饾椈饾榿饾椂饾槂饾棽 饾榿饾椉 饾棻饾椉 饾榾饾椉, 饾棷饾槀饾榿 饾槃饾棽 饾椈饾棽饾棽饾棻 饾榿饾椉 饾棻饾椂饾棾饾棾饾棽饾椏饾棽饾椈饾榿饾椂饾棶饾榿饾棽 饾椆饾棽饾棻饾棿饾棽饾椏 饾棾饾椏饾椉饾椇 饾棽饾槄饾棸饾椀饾棶饾椈饾棿饾棽.
Robinhood Chain is an @Arbitrum Orbit L2. Literally a general purpose EVM with a single Robinhood-run sequencer and an optimistic rollup (in terms of settlement) with a handful of whitelisted challengers. Perfectly fine as a distribution rail and ledger for RH's tokenized stocks. BUT it is not exchange infrastructure. That's also why perps run on a dedicated Lighter, instance with a reported 50/50 rev share.
HL had to build a purpose-built L1 with a fully onchain CLOB to get CEX-grade execution. @Lighter_xyz spent 3 years on custom ZK circuits for verifiable matching and liquidations. That is not a feature you bolt onto an Orbit chain next quarter.
So if @RobinhoodApp does attempt to internalize, it's likely either a centralized engine (doable via Bitstamp and their derivatives arm, but then the onchain story they just started selling will die) or an outright acquisition of smth like Lighter. Given the Ventures stake, Vlad advising, and a relationship Novakovski calls 12 years in the making, an acquisition looks at least as likely as replacement. But only time will tell, and as we know anything is possible in crypto lol. Yet, building a CEX-grade onchain exchange is not as easy as ppl think sometimes.
In conclusion: This is less of a"Lighter vs HL for the same prize" kinda situation and more two different regulatory bets. License the venue vs neutralize the protocol. Both can pay, on different timelines. But given HL's absolutely dominant position in the market (20x Lighter's revenue, over $11B in OI, etc), they actually do have that time. At least for now.
Hyperliquid.
QUOTING @Louround_
Could Lighter overtake Hyperliquid if all the stars align? Let's say clarity act passes, Robinhood keeps Lighter as its perp engine, 28m users (Americans and EU++) start trading perps on crypto and stocks, what does $lit actually earn from that? After writing this article, debu
Could Lighter overtake Hyperliquid if all the stars align? Let's say clarity act passes, Robinhood keeps Lighter as its perp engine, 28m users (Americans and EU++) start trading perps on crypto and stocks, what does $lit actually earn from that? After writing this article, debu
What came next 路 since 13 Jul 2026
Since then
Zero has written 5 more pieces on Hyperliquid since this one.
- State of Hyperliquid: two years after the airdrop (A1 Research report)On Hyperliquid
- Hyperliquid's US regulatory path: HyperCore, HIP-3 deployers and buildersOn Hyperliquid
- Hyperliquid ecosystem briefing: 14 days of regulation, HIP-3/4, RWAs, revenueOn Hyperliquid
- Why profitable onchain businesses deserve attention: Hyperliquid and PumpOn Hyperliquid