StonkFun has done the one thing most “revenue” tokens never do — routed a majority of its cash into buying back and burning its own token, on-chain, where anyone can check it. The catch is everything you cannot check: the business is barely two months old, thinly traded, run by a team nobody has named, and the token has just gone parabolic into the claim. STONK is the native token of a Solana launchpad that clones the pump.fun model with one twist — new coins are quoted against a tokenized stock (an xStock like SPYx or NVDAx) instead of SOL or a stablecoin. The mint has no mint or freeze authority, so supply can only fall: ~102M of 1B was verified burned on-chain on 29 Aug and feeds put it near ~130M (~13%) by 8 Sep. Then on 6 Sep StonkFun became the first partner to bring custom-quote-token launches live on Raydium's LaunchLab, STONK spiked ~250% to an intraday record ~$0.212, and the platform printed ~$1.5M of revenue in a day against a pre-integration run-rate of ~$12M annualised. This report takes both halves seriously: the accrual is real and the entry is expensive.
The launchpad that changed the denominator
Every launchpad since pump.fun has quoted new tokens against SOL or a stablecoin. StonkFun changed one field of the pool configuration — the quote asset — and priced launches against tokenized shares. That single choice is the whole company. Anyone can mint a coin and have a liquidity pool opened automatically, exactly as pump.fun made standard, and the pitch is “launch coins paired with anything”: memes, xStocks, pre-IPO equity (PreStocks), currencies, commodities, leveraged assets and collectibles. STONK itself launched into a STONK/SPYx pool on Raydium's concentrated-liquidity engine on 23 July 2026 at 19:07 UTC, with 1B minted and mint and freeze authorities null from the start.
The economic hook is a revenue-funded burn. StonkFun automatically claims the trading fees on the locked liquidity behind each launch and uses ~60% of that revenue to buy STONK on the open market and burn it, retaining the rest; two adjacent mechanisms round out the design — an Ecosystem Flywheel that recycles a share of pool fees into buying and burning the platform's top-10 tokens, and a Rewards system that pays a launched token's holders in whatever quote asset it is paired against. The token's history is short and steep: aggregators only began listing it in early August, the week of 21–28 Aug closed +88.3%, 1 Sep printed $131,343 of revenue and 3.05M STONK burned in a single day, and then came 6 September.
A memecoin priced in S&P 500 shares — and where the design bites
The technology is mostly borrowed: Solana, Raydium's pools, Backed Finance's xStocks. The invention is a design choice. On a conventional launchpad, a STONK/SOL pool asks “how many SOL is one STONK worth?” A STONK/SPYx pool asks “how many index-shares is one STONK worth?” — a different question with a different answer whenever SOL and the S&P diverge. Pairing against a tokenized equity turns a launch into an implicit relative-value trade against a real-world asset: not tokenized money, but tokenized markets used as the unit of account for on-chain speculation. Since the 6 Sep integration, launches run through LaunchLab — a bonding curve first, then `migrate_to_cpswap` seeds a Raydium CPMM pool at the 0.25% fee tier — with fees harvested through Raydium's Burn-&-Earn program and the buyback executed as swaps on Jupiter.
Three structural frictions follow from the idea, and none of them can be patched away. First, equity beta: a flat token and a rising index produces a losing position inside the pair even though nothing happened to the token — a trader long STONK against SPYx is quietly short the S&P 500. Second, market-hours risk: xStocks track an underlying that stops at the US close and all weekend, but the pool never stops, so a STONK/SPYx quote exists at 05:00 UTC on a Sunday marked against an asset that cannot be arbitraged back until equities reopen. That gap is the deepest risk in the architecture — a feature of the differentiator, not a bug. Third, LP economics: impermanent loss is amplified by a volatile equity quote on both sides, and the market-hours gap lets informed traders pick off a stale weekend price, an adverse-selection tax paid out of LP inventory. LP-ing these pairs is structurally unattractive, so depth stays thin — and thin depth is exactly what makes STONK a market a few wallets can move.
A strong engine, no hostile supply — and a policy rather than a right
Strip the model to its bones and STONK is what PUMP would be without its 33% insider cliff: a revenue-funded burn against a fixed, deflating float. That is a materially cleaner supply setup than almost any “revenue token” this desk has covered, and it is the strongest single argument for STONK. The engine is not stronger than PUMP's — the fees behind it are far smaller — what is stronger is the float it runs into, with nothing pushing the other way. The catch is that a fixed, deflating supply is worthless if demand is purely reflexive, and STONK's is: fees rise when the token is hot, funding more buybacks, lifting the price, drawing more trading. It is a beautiful flywheel upward and an equally efficient one in reverse, because when activity cools the buyback shrinks exactly when the token needs it most.
Governance is where the report gets blunt. The chain confirms the two things that cannot be changed — no mint authority, no freeze authority — and everything else is off-chain discretion: the ~60% rate, the fee split, the use of the retained ~40%, and continuation of the programme itself. There is no DAO, no token vote, and no named operator to hold to account. The surrounding record is thin in the same way: multiple impostor tokens trade as “STONK”/“STONKS” so the contract address is the only safe identifier; the revenue dashboard renders in-browser and returns nothing to a direct query, with reported revenue and reported buybacks failing to reconcile (a ~4× gap in late August); and the quote layer belongs to Backed Finance, whose xStocks are barred to U.S. persons, so a regulatory stop there would impair the differentiator overnight. The verdict the report reaches: a speculative-sleeve position, not a core holding — real mechanism, clean supply, genuine narrative, entered after a +250% day already pulled the price to the front of the story. Watch the supply curve and the equity-quoted share, because the chain answers both questions for free and long before the price does.
Key findings
- The accrual is real, and you can check it without trusting anyone. ~60% of platform revenue auto-routes into open-market STONK purchases on Jupiter and burns them. A hard on-chain read of 897,727,001.74 at slot 442,591,115 (29 Aug, 11:06 UTC) implies ~102.3M burned (~10.2%); feeds put it near ~130M (~13%) by 8 Sep. With no mint authority and no freeze authority, supply can only fall.
- The supply is clean — the mirror image of PUMP. No priced venture round, no investor list, no vesting cliff we can find; ~88–90% circulating and deflating, so market cap ≈ FDV. PUMP's burn had to out-race a 33% insider tranche unlocking through 2029 just to hold the float flat. Here every dollar of buyback is net supply destruction, because nothing is being released against it.
- But holders own no right — only a policy. STONK carries no fee-share, no yield, no vote and no utility requirement. The ~60% is described on a webpage as “approximately 60%”, not encoded the way pump.fun eventually locked its 50% into an immutable contract, and the retained ~40% has no disclosed treasury policy. A pseudonymous operator (@LaunchOnSF) can dial it down or switch it off without breaching any enforceable promise.
- LaunchLab changed the size of the business — for two days. The pre-integration base was ~$12.25M annualised and ~$1.21M cumulative revenue against $705K burned (a ~58% ratio, right on the stated ~60%). On 6 Sep StonkFun went live as Raydium LaunchLab's first custom-quote-token partner, cutting deploy cost ~90% (0.29 → 0.03 SOL): ~$1.5M of revenue in a day, ~$905K straight into buyback-and-burn, and >$1M on each of the last two days. On 7 Sep xStocks quoted ~42% of new launches against plain SOL's ~10.5% — at the launch layer the wedge became the default.
- And you are paying a full-hype price for a run-rate days old. STONK ran ~250% on 6 Sep to an intraday record ~$0.212 (peak cap ~$140M at the ~$0.16 close), then handed back about a third within a day to ~$0.13 / ~$115M. That is ~9× the pre-integration revenue base and under 1× the hot one — a spread that is itself the warning. Feeds disagree ($0.12–0.15 on 8 Sep) and liquidity is thin: ~$1.8M pooled against ~$3.2M daily volume across 20 pools on 29 Aug, a market a few wallets can move.