Tokenized stocks on Robinhood Chain: equities as ERC-20 tokens

1. What tokenized stocks are

A tokenized stock is a blockchain token whose value tracks a real-world equity: one token contract per instrument, with the token supply representing tokenized exposure to that instrument. On Robinhood Chain the naming is literal. The Tesla token is TSLA, "Tesla • Robinhood Token". NVIDIA is NVDA, the S&P 500 ETF is SPY, and the list extends past public equities to private companies such as SpaceX, whose token is deployed as SPCX.

What makes them interesting as a data subject is the standard they use. Each one is an ordinary ERC-20 contract with 18 decimals. There is no bespoke settlement format to reverse-engineer: a tokenized share of Tesla emits the same Transfer event as any token on Ethereum, with the amount in the data field and sender and recipient in the indexed topics. Fractional shares are native to the representation, and in practice most transfers on the chain are fractions: 0.01 TSLA is a typical amount.

2. The chain they live on

Robinhood Chain is an Ethereum Layer 2 built on the Arbitrum Orbit stack, chain ID 4663, with ETH as the gas token. Its public mainnet opened in mid-2026 with tokenized equities as the headline use case, and the early ledger reads exactly like a young general-purpose chain: as of July 2, 2026 the chain stood at 323,791 blocks carrying 128,950 Transfer events from 720 contracts. 114,386 of those are ERC-20 token transfers across 690 token contracts; the rest is NFT activity (Robinhood Punks among them). The single busiest contract is WETH, with 69,940 transfers, 61% of all ERC-20 movement, and most of the remainder belongs to memecoins with names like Lady Marian and Cash Cat. The tokenized instruments are the sliver: 24 stock and ETF tokens have moved so far, 313 transfers between them, 0.27% of the total, and even the full ERC-8056 family, adding the tokenized T-bill fund WEEK, stays under half a percent. That proportion is worth knowing if you plan to index the chain: the equity story sits inside ordinary EVM noise, and filtering by contract address is what separates the two.

The other permanent resident is USDG, the Global Dollar stablecoin, the third most active token on the chain with about 17.9 million USDG minted onchain. Stock tokens are the asset leg and a dollar stablecoin is the cash leg; seeing both in the same transfer stream is what makes portfolio- and flow-level analysis possible from one dataset.

3. The token contracts

These are deployed contract addresses of selected Robinhood stock tokens on mainnet, verified against the chain:

The set grows as Robinhood tokenizes more instruments. Because every stock token is an ERC-20, discovery does not depend on an announcement: aggregate all Transfer events on the chain by emitting contract and new tokens surface as soon as they move. SPCX is the current example of the opposite case, a contract that is deployed and named but has yet to record its first transfer.

4. A real trade, read from the chain

The query below asks SQD's Portal for every Tesla token transfer since the chain's genesis. It is a standard EVM log query, filtered to the TSLA contract and the ERC-20 Transfer signature:

The first event it returns is the token's genesis: at block 143,822, on Monday June 22, 2026, 0.1 TSLA was minted from the zero address. One of the transfers that follows is worth reading in full:

The data field is 0x2386f26fc10000, which is 10¹⁶, or 0.01 TSLA at 18 decimals: a fractional share on the move. The timestamp decodes to Friday June 26, 2026 at 23:53 UTC, hours after the closing bell in New York. Eight minutes later, at block 247,131, the receiving wallet passed the same 0.01 TSLA onward, and by then the clock had crossed into Saturday. A stock changing hands on a Saturday does not exist in traditional market data; on this chain it is just the next row in the stream.

That midnight transfer is not an outlier. Bucket every stock-token transfer by the day it settled and nearly half the activity lands outside the NYSE session: 144 of the 313 transfers in the tokens' first week, 46%, settled on nights, before the open, or on the weekend.

One caveat before you turn this stream into a trade tape: a Transfer on its own proves that custody moved, not that a trade happened. Sweeps between a user's own wallets, operational rebalancing, and internal settlements all emit the identical event, and the midnight transfer above carries no payment leg in its transaction. Classifying every token movement as an execution is how analytics built on transfer data go wrong.

The proof pattern lives in the same dataset. When a purchase settles onchain, the stock leg and the cash leg share one transaction, so query the stock contracts and USDG together and group by transactionHash: rows with both legs are provable trades, complete with a price. Seven of the 286 stock-token transactions on the chain so far are exactly that, and one of them is how the wallet above got its Tesla in the first place:

The same pattern prints on the weekend. At block 261,397, on Saturday June 27 at 10:54 UTC, one transaction moved 6 USDG against 0.0174 GOOGL, about 345 dollars per share for an Alphabet trade settled while every stock exchange on earth was closed.

The same query shape scales sideways. Put several contract addresses in the address array (it is OR logic) and one request returns the transfer history of the whole tokenized equity market; group by contract to rank activity. Drop the address filter entirely and you get every token on the chain, stock tokens, USDG, and the memecoins alike.

5. Issuance and redemptions

Supply changes are the part of tokenized equities that market data cannot show at all, and onchain they are the easiest thing to isolate. Minting, when new shares are tokenized, is a Transfer whose sender is the zero address. Redemption burns tokens with the zero address as recipient. Both are one topic filter away:

topic1 pinned to zero returns only mints; move the same constant to topic2 for burns. Sum the mints, subtract the burns, and you have the tokenized supply of each instrument over time, per token contract, from genesis. For TSLA that whole history is currently thirteen events: nine mints totalling 0.683 TSLA, four burns totalling 0.25, and a net supply of 0.433 tokenized shares. Plotted, it even shows the operators warming up, two mint-and-burn round trips back to zero before supply starts compounding:

That series (how much Tesla exposure exists onchain, and when it grew) is the fundamental chart for the whole tokenized-equity category, and it is derivable from nothing but transfer logs. The same rollup across all 24 traded contracts puts today's total tokenized float at about 36 shares' worth of exposure, over nine of them in the rare-earth company USAR, which tells you exactly how early this market is.

One wrinkle to encode before a corporate action makes it matter: the sums above are in raw token units, and Robinhood stock tokens implement ERC-8056, the scaled-UI extension. On a split or a stock dividend the contract's uiMultiplier() changes while raw balances and Transfer values stay fixed, so the effective share amount is the raw amount times the multiplier. The adjustment is visible in the same log stream: every stock token pairs each Transfer with a TransferWithScaledUI event carrying both the raw and the adjusted value (for TSLA the two series match one for one, 50 and 50), and multiplier updates get an event of their own. As of writing every live stock token reads a multiplier of exactly 1.0, so the chart above is exact in share terms; after the first split, scale raw sums by the multiplier, or sum the adjusted values instead. The full mechanism, including the first real multiplier changes on the chain, is unpacked in the ERC-8056 guide.

6. A portfolio in one query

Because sender and recipient are indexed topics, wallet-level views need no address list at all. Filtering topic2 to a wallet (left-padded to 32 bytes) returns every token that ever arrived there, stock tokens and stablecoin alike:

One decode-time filter before netting: ERC-721 transfers share this exact topic0 and also put the recipient in topic2, so a wallet that has collected Robinhood Punks gets NFT logs back from the same query, with the tokenId as a fourth topic and empty data. Drop logs with four topics before treating the data field as an amount. This wallet has none, so its ledger nets cleanly.

Run it once more with the wallet in topic1 for outflows, net the two, and the result is a holdings history: which tokenized equities a wallet accumulated, when, and how its USDG balance moved alongside. Net flows alone do not say which USDG transfer was payment and which was a deposit; for cost basis, pair the legs by transactionHash as in section 4, which is exactly how this wallet's first Tesla lot arrived (0.0129 TSLA against 4.9575 USDG). For the wallet above, the recipient of the Friday-night transfer from section 4, the two queries net out to this:

A fraction of a Tesla share, a USDG balance on the side, and the full timeline of how both got there. That is the raw material for portfolio trackers, holder analytics, and the compliance-style questions that follow real-world assets wherever they go.

7. Tokenized stock data with SQD

Everything above ran against the robinhood-mainnet dataset on SQD's Portal, which serves the chain from genesis with real-time streaming at the head: open-ended queries follow the chain tip, which matters for an asset class that settles on weekends. The dataset is queried with the standard evm type, so pipelines built for any EVM chain (decoding transfers, joining prices, sinking to Postgres or ClickHouse with the Pipes SDK) work on Robinhood Chain without modification.

One thing to know before you build: Robinhood Chain is currently a private SQD dataset. It is available to enterprise clients with a provisioned API key rather than through the free public Portal. If you want access to Robinhood Chain data, reach out to the team and we will set you up. For the dataset details, see the Robinhood Chain page.