Is StonkFun dumping on its own holders? We followed the fee wallet's $56 million.
On 21 September an X account posted a Solana wallet tagged STONK FEE DRAIN and said it had sold almost $50 million of coins launched on StonkFun. The images spread fast, and so did the replies. We traced it on-chain for a month, sale by sale, and checked what StonkFun's backers said in its defence.
- $56.3M
- $56.2M
- 69%
- 20%
- 58%
- $1.41M
01 · The postA wallet tagged fee drain
On 21 September an X account called MidCurveMortal posted pictures of a Solana wallet that someone had tagged STONK FEE DRAIN. One showed it selling $25,000 of a tiny coin called BUDDY while buying none. The post said it had sold almost $50 million of coins launched on StonkFun, $8 million of it in ZCAT, and was running small caps into the ground on top of a 5% tax.
Other traders piled in. One said the wallet seemed to hold an infinite supply of every StonkFun coin, and dropped 80 SOL at a time into every breakout. Another posted that 60% of one coin's supply had been market dumped. Then came a Grok summary on X. It said MidCurveMortal had accused the wallet of "reverse market making". StonkFun's backers, it said, called the selling standard tax mechanics and pointed to $2.09 million of revenue that had gone into buybacks.
StonkFun is a Solana launchpad, a site where anyone can launch a meme coin in a few clicks. The wallet in those posts is real. It runs the platform's reward system, and StonkFun's own API names it as the address that collects tax on reward coins and pays holders. Bitquery indexes the trades and transfers on Solana, so we could follow it for a month, from 23 August to 22 September. We counted every sale, every payout and every payment that went anywhere else. Then we checked the buyback claim against the chain.
The sales are real, and a little bigger than the post said. Almost all of the money went back out to holders. But the tax behind the selling still hurts charts, and a slice of the money went where StonkFun's own ledger does not show it.
02 · The platformA launchpad that pays you in the pair
StonkFun opened in late July and works a lot like pump.fun, with one twist. A new coin doesn't have to trade against SOL. It can trade against a tokenized stock such as SPYx, against ZEC, or against another StonkFun coin. By its own count the platform has launched more than 105,000 coins, most of them through Raydium's launch tools. On 9 September pump.fun switched on the same kind of stock pairs, and we have measured that market too.
The platform has its own coin, STONK. StonkFun takes a cut of the trading fees on its pools and says about 60% of that money buys STONK on the open market to burn it. So far 17% of STONK's supply is gone.
Most coins on StonkFun are what it calls reward coins. Every time a reward coin moves, whether in a buy, a sell or a plain send, 1% or 3% of it is held back as tax, at a rate set at launch. StonkFun collects and sells it, and pays each holder a share in whatever the coin is paired with. Hold ZCAT, which trades against ZEC, and you get paid in ZEC.
03 · The machineEvery trade feeds one wallet
At first the tax goes nowhere. It sits in the account of whoever received the coins, locked, where they can't spend it, until one wallet sweeps it up, the same one for every StonkFun reward coin. It is the one in the posts.
It sweeps around the clock. What it sweeps gets sold back into each coin's own pool, and the money goes out to holders in batches. Nothing is minted along the way, and these coins can't mint new supply at all, so the "infinite supply" people saw is simply the tax, pulled in from tens of thousands of coins at the same time.
04 · The selling$56 million in 30 days
The critics had the size right. Across the 30 days we tracked, the reward wallet sold $56.3 million of StonkFun coins, spread over tens of thousands of them. ZCAT alone came to $8.47 million, a little above the $8 million in the post.
Treat that as a floor. A few hundred small coins traded on venues our trade data does not price, and the selling started before our window opened.
The pace jumped on 6 September, as StonkFun's new launches moved onto Raydium's LaunchLab. From then on sales ran between $1.6 million and $5.7 million per day, each sale priced at the moment it happened using our price data.
| Coin | Tax | Sold | Supply lost to tax |
|---|---|---|---|
| ZCAT | 3% | $8.47M | 61% |
| KNOTS | 3% | $2.35M | 35% |
| PURR | 3% | $1.86M | 81% |
| GP | 3% | $1.27M | 71% |
| RAYCAT | 3% | $968K | 91% |
| LEVERCAT | 3% | $927K | 56% |
| ALLINU | 1% | $830K | 27% |
| BUDDY | 3% | $41K | 69% |
05 · The payoutsAlmost all of it went back out
Selling is only half the job. Over the same 30 days the wallet paid holders $56.2 million in ZEC, STONK, tokenized stocks, SOL and a few hundred other assets, which is almost exactly what it sold.
Day by day, selling and payouts rise and fall together. Some days it paid out more than it sold and some days less, but the gap never grows.
StonkFun also publishes a ledger of its reward payouts, so we checked it against the chain, one payout asset at a time. More than nine in ten agree within 10%, and overall our count runs about 4% above StonkFun's. The money goes out in batches, to about 14 to 19 holders at a time. Since 20 September a second wallet, topped up by the first, has sent them.
So the wallet is not quietly keeping the tax. It does what StonkFun says it does.
06 · The dragThe tax is the seller
None of this means the critics imagined the damage. A 3% tax on every buy and every sell turns trading into a steady stream of selling, and the wallet is where that stream comes out. On ZCAT, KNOTS and BUDDY it made about 6% of all the selling, twice the tax rate, because it hits both sides of every trade. On ALLINU, which has a 1% tax, it made 2.2%. Other traders still do almost all of the selling. The wallet's share is small, but it comes out of every trade.
The drain is worst on the coins people churn hardest. BUDDY, a small coin paired with a token called AMC, lost 69% of its supply to tax in about two weeks, and ZCAT lost 61%. The median coin, the one in the middle of the pack, lost 2.5%. But 160 coins lost more than half their supply, and nearly all of it was sold back into the pool.
Holders are paid for all this, in the pair asset. The price still takes every one of those sales. Anyone who held reward tokens on BNB Chain in 2021 will know the deal.
07 · The first hourSnipers pay the most tax
The heaviest tax lands right after launch. Sniper bots buy in the first seconds, flip, and buy again. Every flip pays. Within a minute of one launch on 20 September, 522 wallets had traded 3.4 times the coin's entire supply between them. The reward wallet swept up its first 149 million coins of tax and sold them 45 seconds later.
ZCAT shows what that costs. It lost 20% of its supply to tax in its first hour and 39% by the end of its first day. The 17.5% of supply sold as tax in that first hour fetched about $18,000, and at ZCAT's peak, going by StonkFun's own figures, the same slice was worth roughly $30 million.
This is where the "80 SOL clips on every breakout" come from. The wallet sold in at least 662 fills of 80 SOL or more, and its biggest fill was $130,000. What it does not do is sit on tax and dump it into pumps. On every day we measured, what it held unsold was never more than 1.6% of all it had swept up. It sells as it collects, usually within seconds, so when trading spikes its sales spike too.
The payouts go to whoever holds when they are sent, snipers included. One wallet that bought ZCAT in its very first second has since been paid in 281 different assets.
08 · The defenceWhat the backers got right and wrong
StonkFun's backers made two points. The first was that this is standard tax mechanics from earlier meme cycles. That one holds. In 2021, reward tokens on BNB Chain took a tax in their own coin, sold it, and paid holders in BNB, BUSD or CAKE. StonkFun runs the same machine on Solana.
The second was that StonkFun put $2.09 million of its 21 September revenue into STONK buybacks and burns. StonkFun's own ledger says otherwise. It booked $2.11 million of revenue that day and spent $1.23 million of it, or 58%, buying STONK. The rest it kept, which matches the platform's stated policy.
The buybacks themselves are real. That day the platform wallet bought about $1.23 million of STONK through Jupiter, and burned a little more than it bought, topping up with STONK it had taken in pool fees. DefiLlama shows only $0.43 million of buybacks for the same day.
The defence also blurs two pots of money. Revenue is StonkFun's cut of trading fees, and that is what pays for the buyback. The $4.5 million of tax sold that day is a separate pot. It went to each coin's holders in the pair asset, and none of it was burned.
09 · The side doorPayments outside the batches
Most payouts go out in batches to many holders at once. A few went somewhere else. Between 6 August and 5 September, single payments went again and again to the wallet StonkFun's API lists as STONK's creator. From 5 September the same kind of payment went to a treasury instead. The treasury also takes in SOL and STONK from StonkFun's platform wallet, and passes STONK on to the creator. And on 8 August, 1,000 SOL went to a third wallet, which swapped it for USDC and sent it on.
Together that comes to at least $1.41 million. Over our 30 days, payments like these were about 2% of everything paid out. StonkFun's reward ledger leaves them out, and we found no stated reason for them. The chain shows where the money went. It cannot tell us why.
The reward wallet also holds a power it has never used. On 2,178 reward coins launched through LaunchLab, KNOTS among them, it can still change the tax rate, as high as 100%. A change would take a few days to kick in, and none has ever been made. On older coins such as ZCAT the rate is locked for good.
| Wallet | Address |
|---|---|
| Reward wallet ("fee drain") | 5KXDF6…K6tD |
| Payout wallet, from Sep 20 | HuBMeY…i8Ga |
| Platform wallet | 5CEbue…SPAG |
| Treasury | 458aGt…Ukre |
| STONK's creator wallet | H6qoWz…acRQ |
| Got 1,000 SOL on Aug 8 | 4GwDmF…LSWg |
10 · The answerSo, is StonkFun dumping on its holders?
Not in the way the viral posts meant. The "fee drain" is StonkFun's reward wallet. It sells what the tax hands it, pays that money back to holders, and holds no hidden stack. It could not mint one if it tried.
| Claim | What the chain shows |
|---|---|
| Sold almost $50M | True. $56.3M in 30 days |
| $8M of it in ZCAT | True. $8.47M |
| A 5% tax | No. It is 1% or 3% |
| Holds infinite supply | False. These coins cannot mint more |
| 80 SOL clips on breakouts | True. The tax spikes with trading |
| Dumped 60% of a coin | True. BUDDY 69%, ZCAT 61% |
| Reverse market making | Mostly false. $56.2M went back out |
| Standard tax mechanics | True. Same as 2021 reward tokens |
| $2.09M into buybacks | No. $1.23M, 58% of revenue |
The design still sells into every chart, all day long. A 3% tax on a coin that trades many times its supply hands a big share of that supply back to the pool, and while holders get paid in something else, the price takes the hit. On the busiest coins, and in the first hour after a launch, that drag is heavy.
Two questions came up in none of the posts. Why do some reward payments go to a treasury and to STONK's creator? And will the tax switch on those LaunchLab coins ever be flipped? Both can be followed on-chain with the Bitquery MCP.
Follow the money in plain English
Most figures above come from Bitquery's Solana data. The Bitquery MCP server puts that data behind an AI assistant, so you can ask what a wallet sold, who it paid, and where the money went next, without writing the query yourself.
This article describes on-chain activity and does not claim that StonkFun, its team or any holder broke a law or a promise. Each wallet is described by what it did on-chain. The reward wallet's link to StonkFun rests on StonkFun's own API, which names it; the others are tied to the platform only by the payments shown. The sales cover 23 August to 22 September 2026 and are a floor, because a few hundred small coins traded on venues we do not price. Payouts are counted from transfers and valued at each asset's price on the day, so they carry some pricing error, and our count runs about 4% above StonkFun's own ledger. We make no claim about why the payments outside the batches were made, or who knew of them. Market caps are StonkFun's own figures. The social media posts quoted are summarised, and their figures are theirs.
This article is provided for informational and educational purposes only. It reflects analysis of publicly available on-chain data as of the dates given, and does not constitute legal, financial, compliance, tax or investment advice, nor a recommendation or offer to buy, sell or hold any asset. Blockchain addresses are pseudonymous: a transaction between two addresses does not by itself establish the identity, intent or knowledge of any party, and every entity attribution here is an inference that may be incomplete or wrong. Readers should verify independently before acting on anything above, and Bitquery accepts no liability for loss arising from reliance on this material. All trademarks and company names are the property of their respective owners. Corrections and right-of-reply requests go to support@bitquery.io.