T/SFAMI / JINQIAN Explore the data +Methods

How the FAMI Stock PumpFooled Memecoin Traders

A token called FAMI traded as much as 301% above the Nasdaq stock. The two weren’t connected.

The setup

The trade already had a playbook.

Over the past few days, a new meta has taken over Robinhood Chain: launch a memecoin against a tokenized stock instead of a stablecoin. The meme creates demand; the stock token gives that demand a familiar ticker, a dollar reference, and what looks like an arbitrage path.

We covered the first breakout version in our BONER/HIMS investigation ↗. BONER traded against Robinhood’s official HIMS Stock Token. That market had a real issuer, authorised minting, and a primary-market rail—even if thin liquidity and closed cash-market hours could still break the peg.

JINQIAN looked like the next version. Traders saw the “money mushroom” memecoin paired against FAMI and recognized the format before checking the contract.

FAMI is Farmmi’s Nasdaq ticker. The onchain token used the same name, the same ticker, and company-associated branding. In the middle of a stock-token memecoin boom, it looked like tokenized Farmmi.

When onchain FAMI traded as much as 301% above the Nasdaq stock, the apparent trade was obvious: buy the cheaper listed shares against the expensive wrapper. That is the arbitrage people thought they were trading.

Except the wrapper was fake. FAMI was an ordinary ERC-20, absent from Robinhood’s Stock Token registry, with no issuance, redemption, custody, or settlement path to Farmmi shares.

The new meta explains why that missing rail did not stop the cash trade. Traders did not need the token to redeem. They only needed to believe it represented the stock long enough to buy Nasdaq FAMI as the hedge.

The cash tape

First, the Nasdaq move was real

The stock-token narrative did not need a settlement rail to reach Nasdaq. It only needed traders to place the hedge they thought the trade required.

Farmmi entered September 2 with reported average volume of roughly 4.13 million shares. The session printed 837.6 million shares — about 203 times that amount.

Nasdaq TotalView also lets us reconstruct how the order book changed as that volume arrived.

Across the session, displayed liquidity increased and spreads tightened. Median cumulative depth across the top ten levels reached roughly 31,000 shares on the bid and 29,000 on the ask.

This is the cash market that the mistaken-arbitrage story directed traders toward: a low-priced, thinly capitalized stock whose displayed book changed as extraordinary executed volume arrived.

0.3×bid depth vs Sep 1 median

0.4×ask depth vs Sep 1 median

0.3×spread compression vs Sep 1

The price split

One ticker, two prices.

The peak same-bucket premium was 301%. That was the signal, not a curiosity.

At 10:30 AM, the last observed onchain price in the five-minute bucket was $0.9342. The Nasdaq median midpoint in that same bucket was $0.2331. The token was trading at roughly four times the cash reference.

In a real stock-token market, that spread tells arbitrageurs to sell or hedge the expensive token and buy the cheaper cash exposure. Traders familiar with the new Robinhood Chain playbook could read the FAMI chart exactly that way.

The interpretation was wrong, but the cash order it suggested was still executable. The trade could reach Nasdaq only because people acted on the wrong identity.

One ticker, two prices

onchain FAMI/USDN Nasdaq midpointJINQIAN volume

not deployedselected onchain FAMI/USD

$0.1129selected Nasdaq midpoint

not comparablesame-time price gap

The identity test

The FAMI token was not a Robinhood Stock Token

The first thing we checked was the token itself.

Its name is “Farmmi, Inc.” Its ticker is FAMI. Market interfaces also show branding associated with the company.

None of those things establish that it represents Farmmi stock.

Names, tickers and token metadata are fields written into a smart contract. They can be copied.

Robinhood identifies its official Stock Tokens by contract address. The FAMI address trading against JINQIAN does not appear in Robinhood’s canonical Stock Token registry.

Its issuance history looks different too.

The token supply was created in two genesis allocations contained in a single transaction. From that transaction through our September 2 snapshot, we found no additional mint and no burn.

We also found no public issuance mechanism connecting the ERC-20 to Farmmi shares.

The token used the same name and ticker.

It was not the same financial instrument.

Contract identity test · direct registry and log readsThe ticker matched. The instrument did not.

not a Stock Token

Observed ERC-20 contract“Farmmi, Inc.” · FAMI0x5D2e81cB3A6FECe856B824Dfd7e1d6D3dbaD8cd9

Canonical Stock Token registrychain 4663 · 0x5D2e…8cd9NO MATCH

Post-genesis issuance logTransfer(0x0, recipient, amount)0 EVENTS

Public issuance surfacemint() · burn()NOT FOUND

Cash settlement routecustody · subscribe · redeemNOT FOUND

name+ticker+uploaded mark≠economic claim on Farmmi

“Fake” here means a same-ticker token presented as FAMI without the canonical identity or settlement machinery.It does not identify the deployer’s intent.

The counterfactual

Why the 301% price difference wasn’t enough

A large price gap between two assets can look like an arbitrage without actually being one.

What matters is what a trader can do with the inventory afterward.

Robinhood’s real Stock Tokens are tokenized debt securities issued by Robinhood Assets (Jersey) Limited. They are not literally shares held inside a token.

But there is an issuer and a primary-market structure. An authorised participant can access issuance directly, creating a mechanism through which inventory can ultimately be managed across the token and the exposure it represents.

That is the machinery the new stock-paired memecoin trade depends on.

If a genuine stock-linked token trades too high, a market maker can sell the expensive instrument, hedge the underlying exposure and use the primary-market process to rebalance.

The JINQIAN/FAMI pair copied the surface of that trade while omitting its balance-sheet core. The observed FAMI token had no equivalent path.

A trader was free to buy the ERC-20 and buy Farmmi shares at the same time. But nothing allowed that trader to turn the token into Farmmi shares, or the shares into that token.

So the 301% spread was a price difference.

It was not, by itself, a 301% arbitrage.

Stock Token market structure · premium caseA premium can close only if canonical inventory can be created

RHJ primary market · AP only

Genuine Robinhood Stock Token

Secondary marketToken trades above reference

↓ authorised participant quotes and hedges

Primary inventory accessAP ⇄ RHJ issuerDirect subscribe / redeem · canonical mint / burn

New canonical inventory can reach the secondary market. The premium can compress.

Observed same-ticker FAMI

V4 poolFAMI ERC-20

×No mint / burn rail

NasdaqFAMI shares

RHJ issues tokenised debt securities; only authorised participants can mint or burn directly with the issuer.Secondary-market holders cannot convert the observed FAMI ERC-20 into Farmmi shares.

The causal test

Token supply gives us another test

There is another way to test the proposed mechanism.

Suppose the onchain FAMI really had been part of a stock-token arbitrage.

As JINQIAN activity increased and traders demanded more onchain FAMI, market makers would eventually need some way to create or source additional token inventory against their cash-market position.

That should leave evidence.

We scanned every FAMI transfer involving the zero address from deployment through our snapshot block.

The entire token supply had already been created before JINQIAN launched.

There was no subsequent mint.

There was no burn.

The supply stayed flat while Nasdaq FAMI went on to trade 837.6 million shares.

The flat supply proves the cash-market volume did not come from a just-in-time issuance loop. The onchain-to-offchain transmission ran through traders reacting to the apparent arbitrage, not through token creation.

Shared clock · independent ledgersToken supply and Nasdaq volume on one timeline

hypothesis failed

Ship gate: closed.No post-genesis mint event. No burn event. No public mint or burn selector.

The actual pool

What a JINQIAN trade actually did

We then followed the onchain trade itself.

The canonical JINQIAN/FAMI market was a Uniswap v4 pool.

When somebody bought JINQIAN, they sent the FAMI ERC-20 into the pool and removed JINQIAN. The balances inside the pool changed and the relative price of the two tokens moved.

That is where the transaction ended.

No Farmmi share was issued.

No Farmmi share was redeemed.

No custody instruction was sent to Nasdaq.

No message left the onchain system instructing somebody to hedge the trade.

The JINQIAN transaction changed an onchain liquidity pool. It did not change Nasdaq’s share ledger.

The pool itself was created in the same block and transaction window as JINQIAN’s deployment, shortly after the cash market opened. The timing is interesting.

But timing is not a settlement mechanism.

Canonical pool · Robinhood ChainWhat traders actually exchanged

Uniswap v4 · 0.30%

Pay FAMI→v4PoolManager→Get JINQIAN

Current five-minute bucket9:35 AM

Execution price139.4 JINQIAN / FAMI

FAMI turnover126.7K

Swaps13

Pool 0x48cf…d92386 five-minute buckets1 active position across the observed windowprice and turnover from decoded swaps

The fee path

The contracts show fees, not shares

We also traced JINQIAN’s deployment and fee path.

JINQIAN was created through a launchpad contract. Its canonical v4 pool exposes an LP fee, while the associated hook reports a protocol-fee constant and a fee-collector address.

Those contracts help explain how trading in JINQIAN was launched and where configured fees were directed.

They do not create a link to Farmmi.

They also do not tell us who ultimately controls the fee collector, and we do not infer beneficial ownership from the address alone.

Most importantly, none of the contracts we traced creates, redeems or settles Farmmi shares.

Deployment and fee call graphWhat the contracts expose—and what they do not

direct contract reads

JINQIAN side

creator0x3C74…B0B6↓launchpad0x160E…4898

PoolManager pool · 0x48cf…d923JINQIAN / FAMIfee tier 30 bps

↓ configured hook callbackpool hook0xd88D…40CCPROTOCOL_FEE_BPS() → 100↓fee collector0x3676…eDDf

FAMI token side

issuer0xd28d…7B0d↓ genesis onlysame-ticker token0x5D2e…8cd9no contract link to Nasdaq

Read methods: feeCollector() and PROTOCOL_FEE_BPS().Contract roles do not establish beneficial ownership or a cash-market link.

The verdict

The bridge was belief, not settlement.

JINQIAN did matter to the Nasdaq story, but not through the mechanism traders described.

The new Robinhood Chain playbook trained traders to read a memecoin/stock-token pair as two connected markets. JINQIAN/FAMI looked like that trade. Its copied ticker and branding made the false connection legible at a glance.

When the onchain token marked as much as 301% above Farmmi shares, the apparent arbitrage pointed in one direction: buy the cheaper Nasdaq stock. Those orders were voluntary, not forced by a mint or redemption, but they still landed in the real order book.

That is how a disconnected token could help produce a cash-market event. The contract supplied the resemblance. The stock-token narrative supplied the trade. Traders supplied the bridge.

Public data cannot identify every Nasdaq buyer or calculate how much of the 837.6 million-share session came from this narrative. It does establish the channel traders were discussing and that no automatic token-to-share channel existed.

Observed narrative pathThe contract stopped onchain. Traders carried the story to Nasdaq.

belief, not settlement

What traders saw

JINQIAN/FAMI

Same ticker and copied branding made the pair look connected.

Apparent arbitrageOnchain FAMI +301%Nasdaq FAMI looked cheaper → buy the stock

Contract pathStops here ×No public mint, redeem, or settlement rail.

Human pathTraders act →They choose to send cash-market orders.

Nasdaq FAMI cash bookReal orders. Real shares.

The contract supplied the resemblance. The stock-token story supplied the trade. Traders supplied the bridge.

Conclusion

The trade survived after the arbitrage disappeared.

This was not an unexplained coincidence between two unrelated tickers. It was a recognizable Robinhood Chain trade copied onto an asset that only looked like a Stock Token.

BONER/HIMS had a real receipt, an issuer, and a primary market. JINQIAN/FAMI kept the meme, the ticker, and the premium while removing the financial link that made arbitrage possible.

Traders still knew what the chart appeared to say: onchain FAMI was expensive and Nasdaq FAMI was cheap. Buying the listed stock looked like the other side of the trade—even though there was nothing to deliver, redeem, or settle afterward.

Farmmi’s fragile cash book made that category error consequential. A false wrapper did not need to own or control shares to move the stock. It only needed to persuade traders that the shares were the hedge.

The token contract did not buy Farmmi.The mistaken arbitrage story gave people a reason to.

Methods / reproducibility

How the evidence was built.

Frozen inputs

The published page reads two checked-in derived datasets. It does not query a market-data provider or Robinhood Chain from the reader’s browser.

The onchain/EOD snapshot was frozen at block 52,864,534. The order-book file was derived separately after both Nasdaq sessions were available from Databento. Credentials and raw Databento records are not included in either file.

Price comparison and cash fields

EODHD supplied the FAMI.US daily bars from August 20 through September 2, plus the provider’s delayed quote, average volume, reported float and shares outstanding. The 837.6 million event-day figure is the EODHD daily-bar volume; it is not the Nasdaq-only execution total below.

The reported peak gap compares the last observed onchain swap price in the 10:30 AM five-minute bucket, $0.9342, with that bucket’s Nasdaq TotalView median midpoint, $0.2331. The resulting 300.7% is a synchronized bucket comparison, not a simultaneous executable spread or an NBBO quote. The later stored snapshot was only 70.4% above the EODHD delayed last trade; it is not the event peak.

Nasdaq depth and executions

We requested XNAS.ITCH mbp-10 records for raw symbol FAMI from 09:30 to 16:00 ET on September 1 and September 2. The last valid top-ten book state received in each second becomes that second’s sample.

Five-minute depth frames are medians of those per-second states. Each step is cumulative displayed shares at a book level; horizontal distance is basis points from that frame’s midpoint. Empty intervals carry the last book state but retain zero executed volume. Volume bars sum only action T sizes, so they represent Nasdaq-venue executions rather than all reported US trading.

The depth replay uses one linear detail scale for both days. For this snapshot the view stops at 100,000 displayed shares; a larger observation is capped visually and labeled with its uncapped value. All ratios and session statistics use the uncapped data.

Onchain price and flow

An Alchemy-backed Robinhood Chain RPC supplied Uniswap v4 PoolManager logs through block 52,864,534. We filtered Swap events by exact pool ID: 131,134 events in the FAMI/dollar pool and 96,642 in JINQIAN/FAMI.

Those logs produce 94 and 86 non-empty five-minute buckets, respectively. Price is decoded from the final sqrtPriceX96 observed in each bucket and adjusted for token decimals. FAMI turnover is the sum of absolute FAMI amounts, not directional net flow. Empty buckets are not invented; chart lines connect observed closes.

Identity and supply

At build time, the exact FAMI contract address did not match any deployment in Robinhood’s Stock Token registry. We also searched the deployed runtime bytecode for conventional public mint(address,uint256) and burn(uint256) selectors; neither was present.

Separately, we queried every FAMI Transfer log involving the zero address from deployment through the snapshot block. One deployment transaction emitted two transfer-from-zero logs that created the full supply. We observed no later transfer from zero and no transfer to zero. The registry result and event history—not the copied name, ticker or image—support the identity finding.

Pool structure and fees

JINQIAN/FAMI is treated as a Uniswap v4 PoolManager pool, identified by its pool ID. The liquidity view nets ModifyLiquidity events by sender, tick range and salt from pool initialization through the snapshot block. Raw v4 liquidity units are not presented as token balances or Nasdaq shares.

The 30-basis-point LP fee comes from the recorded pool key. The hook’s displayed fee collector and PROTOCOL_FEE_BPS() value are direct contract reads at the snapshot state. These reads show configured addresses and values; they do not establish who ultimately controls an address or how much fee revenue was paid.

Causal limit

Contemporaneous posts and reporting establish that traders publicly circulated the tokenized-stock and arbitrage explanation while the markets moved. Contract data rules out the proposed automatic mint, redemption or settlement route.

Public data cannot connect an onchain wallet to a Nasdaq buyer, recover the motive behind an individual order, or allocate a share of the session’s cash volume to that narrative. The article therefore distinguishes the observed timing and public explanation from a mechanically proven causal transfer.

Sources

Robinhood Chain canonical token registry Robinhood Stock Token structure and primary market Robinhood Chain connection guide Robinhood Chain PoolManager logs FAMI token and transfer logs JINQIAN token and deployment JINQIAN/FAMI pool hook and fee reads Observed JINQIAN/FAMI pool Observed FAMI/dollar pool Databento Nasdaq TotalView-ITCH EOD Historical Data API Contemporaneous market report Transparent Space BONER/HIMS investigation Contemporaneous tokenized-stock attention-loop claim Contemporaneous issuer and arbitrage correction