Robinhood's Plan to Bring Global Finance Onchain | Johann Kerbrat

Summary

• Three weeks into its July 1 launch, Robinhood Chain has put up numbers that contend with Base: ~$750M market cap, roughly half a billion in TVL, 105M+ transactions, ~2M addresses, and over $560M in 24-hour DEX volume. Kerbrat hedges the DAU-flipping-Base chart — “the data always depend on how you slice it” — but confirms Robinhood is top-tier on daily active users and developers, crediting a launch ecosystem of perps, spot DEXs, Earn, and stock tokens rather than a bare chain launch.
• The chain is built on the Arbitrum stack on top of Ethereum for “liquidity, decentralization, security,” with no token planned right now — ETH is the gas asset. Kerbrat says: “we don’t have any plan for token… if tomorrow there is a real reason to have a specific token that will give benefits to the customer, that’s something that we could look into” — and he explicitly won’t optimize chain fees for revenue: “less about optimizing for revenue and more optimizing for adoption.”
• Crypto is one of 11 Robinhood business lines each doing at least nine figures in revenue, with its own P&L. Trading fees are still the largest crypto line, layered with staking, lending, Bitstamp’s institutional exchange, and a crypto-as-a-service white-label business already used by European banks including BBVA — plus Robinhood Canada, live since July 1.
• The partner playbook is often deep integration with one or a few names per category, not breadth: Morpho powers Earn at a deliberately stable ~7% rate, insured through Lloyd’s of London in “probably one of the largest insurance program for crypto.” There’s no rev share with Morpho; the Lighter perps rev share is “public. It’s on the contract.” Kerbrat’s candid leverage admission: “the most useful part of being Robinhood is that I can reach out to almost anyone, and they usually pick up the phone” — backed by 27 million funded accounts.
• Tokenized equities are “underhyped” in Kerbrat’s view, and the binding constraint is regulation, not technology. Not compelling for Americans buying US stocks, but transformative for global access — “if I ask you tomorrow to buy a stock on an Indonesian exchange… you can’t do it right away” — and eventually extending to real estate, art, and private equity, with the on-chain/off-chain distinction abstracted away: “most customers won’t care if it’s on chain or off chain.”
• The Wall Street story is all three at once: a revenue stream, an international-expansion shortcut, and a wedge into DeFi-native users Robinhood previously never touched. Building US-style Robinhood country-by-country “will take a long time”; a self-managed wallet in 120+ countries gets there faster, and Kerbrat says the on-chain DEX activity — about half a billion in 24-hour volume — was mostly from “pure DeFi customers” who came from other chains.
• On competition, Kerbrat declines the knife-fight: “we are still way too small to just focus on that. Let’s focus on increasing the pie.” Vlad is described as deeply into crypto and, with Baiju, as defending it during the 2018 “crypto is dead” era; the one internal tension Kerbrat mentions is pace — “he will always want us to move faster.”

Deep dive

1. Crypto is Robinhood’s international expansion vehicle, not a side bet

• Kerbrat’s arc: the vision was always that customers should be able to access “any type of financial instrument”; Robinhood launched Bitcoin and Ethereum trading in 2018, then saw blockchain as a way to go global, since “you can get access to a hundred of millions of people with just a wallet that people can manage themselves.” The wallet is now live in 120+ countries, carrying stock tokens and yield products.
• His swipe at rivals, worth keeping: “what a lot of competitors have done is creating a chain just for the sake of it” — Robinhood’s chain is meant as backbone infrastructure integrated into the main app, and his framing of the era: “we need to stop this question of DeFi TradFi. We just need to bring everything to the customer.”
• The concrete benefit of on-chain rails: permissionless composability means “we don’t need to create a separate FIX line with every single market maker” — extending Robinhood’s existing best-execution habit (Bitcoin orders already route across exchanges and market makers to the best price) to lending and everything else, with smart contracts reducing intermediary fees.
• On a 5–10 year horizon, Yanowitz asks if all of Robinhood moves on-chain. Kerbrat’s hedge: the worlds “will converge” — tokenized assets connecting to any exchange, portfolios as collateral globally — but full migration is “maybe a bit further than 5 years.”

2. A diversified P&L machine — and crypto’s own diversification project

• Robinhood has “11 different business lines that are making at least nine figures in revenue,” which cushions cyclicality: crypto soft this year, prediction markets up. Crypto runs as its own all-in P&L — headcount, engineering, and costs included.
• Revenue sub-items: trading fees remain the largest, plus staking, lending, and the Bitstamp institutional exchange acquired last year. Kerbrat won’t project line-item sizes — “we haven’t done this type of projection” — but the goal is a crypto business “not just based on volatility of trading.”
• The under-appreciated line: crypto-as-a-service. Institutions use Robinhood “a bit as a white-label platform” or for stablecoin on/off-ramping between entities; BBVA is a named European bank customer. Robinhood Canada went live July 1.

3. Chain design: Arbitrum stack, Ethereum security, fees balanced against spam

• Build-vs-buy resolved toward building from scratch — no “baggage” from acquiring a chain — on the Arbitrum stack atop Ethereum. Why Ethereum: “Liquidity, decentralization, security.” Why owning the stack: control over block time and gas — when launch demand spiked, “we made sure that we could reduce the gas fees,” which Kerbrat said would not have been possible without owning the chain and its decisions — plus Arbitrum’s Stylus multi-language support and priority ordering.
• Launch strategy was ecosystem-first: Earn seeded stablecoin TVL at launch, alongside perps via Slider and spot via Arcus. There was also meme-coin activity and $19M TVL in the Stock Token RWA product. The chain is deliberately “AI first,” with agent skills already deployed from Privy and Uniswap.
• On chain economics, an explicit balance: it “will generate revenue,” but “the last thing I would like to do is to optimize for revenue on the chain” and drive developers away — while making it completely free could invite “spam all over the place with fake activity.” The chain is “a back-end solution to build on top of.”

4. Partner leverage: announce early, pick few, keep them outside

• The sequencing was deliberate: announcing the chain the previous summer at Robinhood’s first event was a flare — “Robinhood is coming into the space. If you want to build with us…” — and the DAS testnet turned that into inbound from builders, plus infrastructure deals with Alchemy and Previ.
• Morpho won lending because “the way that they were thinking about building product really matched how we wanted to build products”: a roughly 7% rate held stable for a while rather than repricing hours after launch, wrapped in a Lloyd’s of London program under which some smart-contract losses would be covered. No rev share with Morpho; Lighter’s rev share is public on the contract.
• Kerbrat on Robinhood’s privileged position — make-or-break for integrated startups — lands on distribution, not extraction: almost anyone picks up the phone, and “if I put on Twitter or X tomorrow that I want to build XYZ, I will get a response… pretty quickly,” with 27M funded accounts as the prize.
• No appetite to acquire partner teams: “the startup is still the most exciting part… they can experiment, try things, iterate until they find product market fit” — inside Robinhood they’d inherit a public company’s requirements and rules. Hyperliquid builder codes were “discussed a few times,” but Robinhood wanted a platform “very unique” to itself — EU perps run through Bitstamp. On the DeFi side, Kerbrat cited Robinhood’s strong relationship with Lido, which it invested in last year, along with protocols such as Orcas; he also described a perps experience in which stock tokens can serve as collateral.

5. “We don’t subsidize” — and no token for its own sake

• Yanowitz’s pushback: the real lending rate is 3–4% and Earn pays ~7%, so “there’s some subsidy, right?” Kerbrat’s flat rejection: “we don’t subsidize” — Canada’s 90-day zero-fee crypto trading is customer acquisition, not subsidy: “instead of spending hundreds of millions” on naming a stadium — Robinhood has never bought one, though “you never know” — the money goes back to customers instead of being charged as fees, with “math behind it.”
• On the perennial token question: teams building on the permissionless chain will launch tokens and “they should decide for their own strategy,” but Robinhood itself is “more interested in this type of deep integration than necessarily just launching a token for the sake of launching one.”

6. Tokenized equities underhyped, competition premature, Vlad impatient

• Underhyped, with the caveat stated: “if you’re American… probably not the most exciting for you to have tokenized US assets.” The payoff is elsewhere — the Indonesian-exchange example, plus real estate, art, and private equity later — and the blocker is squarely regulatory: “if tomorrow we had a clear rulebook across the world, you could actually see the adoption.” End state: tokenized and traditional SpaceX could be abstracted in the user’s experience; customers just want 24/7 trading and collateralization while “all the complexity of private keys and wallets… we want to abstract that away.”
• On rivals who name Robinhood as their competition, Kerbrat won’t reciprocate: market-share tracking helps assess product-market fit, but “we are still way too small to just focus on that. Let’s focus on increasing the pie” — tokenized-asset holders remain “a fraction of the amount of people holding actual equity.” On Base’s experiment: “honestly, I don’t know exactly what went wrong,” paired with respect for Jesse, Kobe, and Brian and enthusiasm that Coinbase keeps experimenting.
• Tokens and equities converge in his view — tokens as a fundraising-plus-community primitive mean companies “maybe not going public on traditional exchanges, but actually deciding to… go public through a token.”
• The closing color: Vlad and Baiju backed crypto in 2018 when competitors said “crypto is dead,” and the one internal tension Kerbrat mentions is tempo — “he will always want us to move faster. And sometimes I’m more the engineer in the background being like, ‘Hey, I need a bit more time to build.’”