Everything this runs on, written down

Robinhood Chain is an Arbitrum Orbit L2 settling to Ethereum. It went to genesis on 30 April 2026 and opened publicly on 1 July 2026. Gas is paid in ETH.

Every stock token described here lives there and nowhere else. The June 2025 launch issued tokens on Arbitrum One behind a transfer allow-list; those contracts are not what this reads, and no reputable source publishes their addresses.

Robinhood publishes its asset registry unauthenticated. It is the only safe way to turn a ticker into a contract address, and this app resolves every token through it before touching a price feed.

That is not caution for its own sake. The chain carries memecoins wearing stock tickers — an AMC that is "A Meme Coin", a "BANDIT STOCK", a "StonkBroker" — at addresses entirely unrelated to the equities they imitate. Resolving a symbol against a price API returns the wrong asset. Address first, always.

These are tokenised debt securities issued by Robinhood Assets (Jersey) Limited, giving economic exposure to the underlying equity rather than title to a share. They are ordinary ERC-20s at 18 decimals, and most carry a published ISIN.

The primary market is gated: only Authorised Participants can mint or burn against the issuer, after onboarding. The secondary market is not. Transfers are unrestricted, which is why the tokens trade on Uniswap v4, v3 and v2, on RFQ venues, and on Lighter — and why a public price exists at all.

Splits and dividends are handled under ERC-8056. Raw balances never change; the adjustment is carried in a multiplier held against the asset. A holder of 100 raw units after a four-for-one split still holds 100 raw units, and the multiplier reads 4.

Any figure derived from raw supply is therefore wrong until the multiplier is applied. CRWD currently sits at 4.0, so its market cap reads a quarter of its true size if you trust the raw number. Market-data feeds report supply raw, so every supply, market cap and float figure on this site has the multiplier applied from the value the registry publishes.

Prices are aggregated from the decentralised exchanges where these tokens actually trade — the Uniswap pools, RFQ venues and orderbooks named above — rather than from the issuer. A price here is what someone paid, not what someone quoted.

A whole table is priced in one batched request rather than one per row, and charts come from OHLCV candles over the same source. 193 of the 194 registry assets return a live price; the one that does not is shown as pending rather than filled with a guess.

Credentials are server-side only. Every market-data request is made from a route handler, and nothing sensitive reaches the browser.

Holdings are not taken from any aggregator. Balances are read from the chain itself, at a block you can name, and valued at market afterwards. Nothing about a reported AUM depends on a third party agreeing with us about what a wallet holds.

Reading 196 balances as 196 separate calls gets rate-limited, and batching that many into one JSON-RPC array is rejected. Multicall3 is deployed here at its canonical address, so the entire scan — every registry asset plus the curated ones — is a single eth_call. Two upstream requests value a wallet of any size.

The tokens page ranks every registry asset by live liquidity and shows the head of that list. Ranking rather than a hand-kept list means the page follows where trading actually is, and nobody has to maintain it as depth moves.

The assets page is the other side of the book: the stablecoin and wrapped ETH that the stock pools quote against. Because no registry enumerates those, they are an allowlist keyed by address, and anything else the chain surfaces is shown without a verified mark rather than hidden. You can see what the chain holds and what we vouch for at the same time.

A strategy is a stated rule for when to hold and when to stand aside, run against one market you choose. It is not a fund and not a manager’s discretion: the rule is written down before you allocate, and it is the same rule for everyone running it.

Each strategy declares three things up front. A thesis — what it believes and therefore when it acts. A risk band, Low through High, which describes how violently it can move against you before its own limits trigger. And an epoch, the period over which its performance is measured and settled: three, seven, fourteen or thirty days depending on how quickly the thesis is supposed to resolve.

Faster epochs settle more often and pay more per period; slower ones ask you to be right over a longer horizon. Neither is safer by construction — a thirty-day epoch simply means you find out less often.

Tokens are emitted against realised performance at the end of each epoch, never in advance and never for simply having capital parked. Every strategy publishes a base emission — tokens per epoch per 1,000 allocated — and that base is multiplied by how the epoch actually went.

The multiplier is the whole mechanism. An epoch that finishes below par pays nothing at all: the multiplier is zero, not a reduced rate. At par it pays the base. Above five percent it pays 1.75×, and above fifteen percent it pays 3×. So the emission for an epoch is base × multiplier × allocation ÷ 1,000.

Worked through: Momentum posts a base of 12 on seven-day epochs. Allocate 10,000 and an epoch that clears par emits 120 tokens — 1.2% of what you put in. There are 52.1 seven-day epochs in a year, so an unbroken year at par compounds to roughly 86%. The same year at the Strong band would be 2.1% a period instead.

That last figure is arithmetic, not a forecast. It assumes every epoch of the year clears at the same band, which no strategy does — and because a losing epoch pays zero rather than a negative, a run of them shows up as nothing earned rather than as a loss to claw back. Read the annualised column as the shape of the schedule, not as a projection of what you would take home.