TriportRPCDashboard

Robinhood Chain first quarter on mainnet: what was reported and what it means for fee handling

A primary filing, a research firm and a third-party analyst give three different pictures of the same quarter. Fee swings of that size are the practical lesson for anyone sending transactions on the chain.

Robinhood Chain went live on mainnet on 2026-07-01. Four months later there is a first full quarter to read, and three sources that disagree about it. Robinhood's own filing is the primary one. A research firm's quarterly and a third-party analyst's note are the others. For an engineer who has to send transactions to the chain or index what happens on it, the disagreement itself is informative, and one of the swings has a direct effect on how you set fees.

This is a news post and a dated one. We fetched the filing on 2026-11-01. Everything from the other two sources is secondary and marked as such.

What Robinhood reported

On 2026-10-27 Robinhood published results for the quarter ended 2026-09-30. The press release exhibit, filed with the SEC, says that in its first quarter on mainnet "Robinhood Chain ranked among the top five chains by revenue and volume, 'generating over $80 billion in DEX volume'" and over $50 million in chain fees. In the forward-looking statements section it lists "the introduction of products such as Robinhood Chain and Stock Tokens globally" among sources of regulatory, legal and operational risk, and it names the continued offering of "Classic Stock Tokens (formerly 'Robinhood Stock Tokens')" in the European Economic Area as a risk area alongside perpetual futures there.

Two details are worth noticing. The wording is "over", so both figures are lower bounds and there is no breakdown in the text. And the product name changed: the EEA offering is now "Classic Stock Tokens", which matters if your code or your documentation matches on the old display name. Contracts are identified by address, not by marketing name, and the name you read in a token's metadata is whatever the issuer set.

The same exhibit says about 200 thousand customers have opened agentic trading accounts, driving over $1.6 billion in notional trading volume to date, and that Robinhood announced 17 new products at its September summit. Neither number says anything about on-chain activity.

What the other two sources say

ARK Invest's quarterly DeFi report, as relayed by a secondary outlet on 2026-10-22, puts Robinhood Chain's first-quarter network revenue at about $51 million, DEX volume at about $65 billion and tokenized equities at about $13.7 billion of that. We could not retrieve ARK's report itself, so these are second-hand. The revenue figure is close to Robinhood's "over $50 million." The volume figure is not close: $80 billion against $65 billion is a gap of $15 billion, or 23 percent. Different counting windows, treatment of internal swaps or double counting through routers could each explain a gap like that, and we cannot tell which from outside.

A research piece by Four Pillars from 2026-10-20, as reposted by ChainCatcher on 2026-10-21, adds daily figures: chain fees of about $8.27 million on 2026-09-04, a spot trading volume of about $3.27 billion that day, and mid-October fees of roughly $30,000 to $40,000 with daily volume around $600 million to $700 million since early October. The piece attributes the September fee spike to congestion on the Arbitrum technology stack, where the base fee rises when usage exceeds target capacity. We could not reach the original, so treat the daily numbers as reported, not verified.

Arithmetic on the three numbers

The quarter from 2026-07-01 to 2026-09-30 has 92 days. "Over $50 million" in fees is an average of about $543,000 a day. The reported peak day of $8.27 million is about 15 times that average. The reported mid-October range of $30,000 to $40,000 is roughly 6 to 7 percent of it.

Divide fees by volume, with every caveat above, and you get a rough fee rate. For the quarter: $50 million over $80 billion is about 6 basis points. For 2026-09-04: $8.27 million over $3.27 billion is about 25 basis points. For mid-October, using a midpoint of $35,000 over $650 million, it is about 0.5 basis points. The same trade cost about 47 times more in fees on the peak day than in mid-October if these inputs are right. We would not publish that ratio as a fact, because the inputs come from different sources and may be defined differently. It is still the right order of magnitude to plan for.

What this means for fee handling

The point of the arithmetic is not market analysis. It is that a fee-per-transaction constant tuned in October is wrong in September, and the reverse. Our gas guide records the model: the chain is an EVM-compatible Orbit chain that follows the Arbitrum Nitro gas model, so the gas a transaction pays covers both its execution and the cost of posting its data to the parent chain. Ask the chain for the numbers instead of reusing an Ethereum mainnet limit. That means eth_estimateGas for the limit, eth_gasPrice for the price and eth_feeHistory for a window of base fees, all read at send time. Re-estimate if you hold a signed transaction for a while.

The failure you see when you get this wrong is specific. If the cap you signed is below the current base fee, the node answers max fee per gas less than block base fee, which is the node's own JSON-RPC error passed through unchanged. If the account balance cannot cover gas limit times max fee per gas plus value, you get insufficient funds for gas * price + value. In a spike the second error can appear on an account that was fine an hour before, because the worst-case cost rose. Treat both as signals to refresh the fee inputs and retry, not as fatal.

Fee history also tells you something about ordering. Our eth_feeHistory reference notes that it is useful for budgeting and that it does not predict ordering, because ordering here is arrival order rather than an auction. A higher tip does not buy a better position, so in a spike, raising the cap protects you from rejection but does not move you forward.

Counting stock-token activity yourself

If the volume figures from three sources cannot be reconciled, reconcile your own. Tokenized equities on the chain are ordinary ERC-20 contracts, as our stock token transfer guide puts it: a position moving between wallets is a Transfer event, and there is no separate stock-token API. Filter eth_getLogs by contract address, derive the topic from the event signature in code instead of pasting a literal, and backfill after a disconnect. Counting transfers is not counting DEX volume, but it gives you a number you can audit end to end, with the block range and the method on the record.

If you need to answer a question like "how busy was the chain on a given day", the block number is the wrong clock for some purposes. Our block number explainer covers why eth_blockNumber on this chain is not the parent chain's number.

When this is the wrong approach

Do not build capacity or pricing decisions on a quarterly figure. A first-quarter number with a launch spike in it says little about a steady state, and Robinhood's own filing does not break the quarter down. Do not build fee logic on an average at all: the point of the daily range above is that the average is a poor guide to the day you send. If you are not sending transactions on the chain and only read it, the fee discussion matters less than the reorg and finality behaviour of the feed you use. And if you need volume figures for a report, take them from a source that publishes its method, and cite the date. Our own endpoint limits for the chain are listed as unmeasured operational defaults on the limits page; they are not capacity claims, and nothing in the filing changes them.

Sources

All fetched 2026-11-01.