Originally published at mrnasdog.com/research/pump/inflation by MrNasdog.
PUMP cannot be minted — the Pump.fun mint on Solana reads mintAuthority: null, so the 1 trillion PUMP created at launch is a permanent ceiling — and PUMP supply on the market still grew +10.16% in the 90 days to Sep 5 2026. The cause is a vesting cliff, not an emission: 65.21B PUMP reached team and early-investor wallets as Pump.fun's first insider year ended, against 25.02B PUMP bought on the open market and destroyed under Pump.fun's 50%-of-revenue buyback contract. Our supply monitor reads the same window at +12.83%, a gap of 2.66 percentage points. The cliff itself is behind PUMP, but the forward 90 days still read +1.86%: three dated unlock tranches of 6.875B PUMP each are larger than what the same dollars of buyback can buy at today's PUMP price.
The verdict, in one paragraph
For the 90-day window ending Sep 5 2026, the MrNasdog Pressure Framework reads PUMP at +10.16% net supply growth and projects +1.86% forward. Total sell pressure was 65.21B PUMP of realised vesting release; total buy pressure was 25.02B PUMP of buyback-and-burn, and every other row on both ledgers is zero. Our supply monitor reads +12.83% for the same window — a gap of 2.66 percentage points, well outside the half-point tolerance, so a monitor-gap chip ships on the PUMP overview. Two named mechanisms more than account for it: 6.99B PUMP released from the Pump.fun vault but still sitting undelivered in two paying wallets, which the monitor counts as trading supply and this framework does not, and a base convention in which the monitor divides the same rise by the float as it stood 90 days ago while this page divides by the float as it stands now. The cite-able label for PUMP is a hard-capped, un-mintable launchpad token whose float is set by an insider vesting calendar on one side and a dollar-denominated revenue burn on the other — inflationary while the vesting calendar is larger than the burn it can fund.
Sell pressure: where new PUMP comes from
Sell #1, protocol inflation, is zero for PUMP, and it is one of the few zeros in this framework that is genuinely permanent. The PUMP mint on Solana is an SPL Token-2022 mint, and reading its account directly this week returns mintAuthority: null and freezeAuthority: null. A null authority is not an authority that has never been used — it is the absence of one. No key exists anywhere that can create another PUMP, and none exists that can freeze a PUMP balance. Pump.fun's 1 trillion PUMP genesis is a ceiling that can only be approached from below, and the PUMP mint's live supply reads 835.58B PUMP — 164.42B PUMP already destroyed and unrecoverable.
Sell #2, vesting unlocks, is the whole of PUMP's sell side and the load-bearing number on this page. The Pump.fun allocation puts 200B PUMP with the team and 130B PUMP with existing investors behind a twelve-month cliff dated from the July 2025 ICO: 25% of each — 82.5B PUMP together — released on Jul 12 2026, with the remaining 247.5B PUMP vesting linearly across 36 months at 6.875B PUMP a month. On paper that made the trailing window worth 89.38B PUMP. The chain says otherwise, and the chain governs: PUMP that vests on a calendar but never leaves the vault is not sell pressure yet.
So this build read the Pump.fun launch vault itself rather than the calendar, walking every one of its transactions across its whole life. That vault — the Solana token account that received the entire 1 trillion PUMP genesis in July 2025 — released 62.2B PUMP in a single transaction pair on Jun 25 2026, weeks ahead of the cliff distribution, and a dormant side vault emptied a further 10B PUMP on Aug 10 2026. Both flows went to two paying wallets, which then fanned PUMP out to individual recipients. Measured across the custody boundary, the four accounts held 375.46B PUMP at the window open and 310.25B PUMP at the close, so 65.21B PUMP actually reached wallets — 24.17B PUMP less than the published Pump.fun calendar implied. The transfer legs reproduce that figure to within 949 PUMP, and one paying wallet's delivered total independently matches dated reporting of 57.28B PUMP distributed to 121 wallets on Jul 15 2026 to within a quarter of one percent.
Sell #3, foundation and unscheduled unlocks, carries no value this window: no dated, sized discretionary release of PUMP was observed beyond the vesting calendar already booked above. Sell #4, long-term locked or bankruptcy, is zero because Pump.fun is a running business with no estate, no trustee and no court-ordered distribution of PUMP to track.
Buy pressure: where new PUMP goes
Buy #1, the programmatic buyback, is the entire buy side. Since Apr 28 2026 Pump.fun has committed 50% of net platform revenue — from the bonding curve, from PumpSwap and from its trading products — to buying PUMP on the open market and destroying it, under a smart contract locked for twelve months and running to roughly Apr 2027. Across the trailing window Pump.fun spent $56.62M and bought and burned 25.02B PUMP against $110.70M of platform revenue, a payout ratio of almost exactly the promised half.
The destination test matters more than the headline here, because a buyback that buys and parks is not a burn. Pump.fun publishes two burn wallets, and both read 0 PUMP — they are pass-through hoppers that fill through the day and are swept empty. Reading only those balances would have reported no burn at all and shipped PUMP's entire buy side at zero. The surface that actually moved is the PUMP mint's own supply: 164.42B PUMP below the 1 trillion genesis, destroyed by the burn instruction rather than parked anywhere. Pump.fun's own published purchase record sums, day by day since July 2025, to 164.15B PUMP — closing against the chain with a residual of 0.16%. The burn is real, the tokens are gone, and nothing accumulates that could later leak back onto the market.
Buy #2, protocol fee burn, is zero — and deliberately so. Pump.fun's fees are what fund the buyback above, and burning is where those purchases end; booking a fee burn as well would count one flow twice on a page whose buy side is already its most attractive feature. Buy #3, foundation buy, is zero because the Pump.fun corporate treasury is held in dollars and SOL, with no dated open-market PUMP purchase outside the buyback contract. Buy #4, new long-term lock, is zero because PUMP has no protocol staking and no lock: outside products that pay a yield on PUMP hand back a freely tradable receipt, and a receipt is not a lock.
Foundation and overhang
Three team-controlled PUMP balances are tracked, all of them readable on Solana. The Pump.fun launch vault still holds 303.26B PUMP; the future team and investor tranches inside it sit on a published calendar and are booked into Sell #2 as each fires, but the 24% community allocation has no published dated schedule at all, which makes the vault the largest unscheduled overhang on this page. The two paying wallets hold 4.06B PUMP and 2.93B PUMP — both were empty when the window opened and neither existed before Jun 13 2026. That 6.99B PUMP is already released from the vault and simply has not been delivered; it is the single largest reason this page and our monitor disagree.
There is no fourth overhang to enumerate. The buyback has no accumulation wallet, because bought-back PUMP is destroyed on arrival. Pump.fun has no DAO, no vote portal and no on-chain governance that could move supply. And there is no bankruptcy estate. All three overhangs above are refreshed straight from Solana every rebuild, and the rule is simple: if any of these balances falls between refreshes, the outflow enters Sell #3 at the next refresh.
How PUMP compares to other application-revenue tokens
PUMP belongs to the class of application tokens that buy themselves back out of real revenue — the same structural family as the large perpetual-DEX and exchange tokens. Mechanically PUMP is at the strong end of that family. Its supply is hard-capped at 1 trillion with the mint key destroyed, so unlike an uncapped continuous-emission Layer 1 there is no issuance curve competing with the burn, and unlike a quarterly discretionary buyback the Pump.fun commitment is contractual and dated to Apr 2027. Compared with a fee-burn chain, where the burn scales with block space demand and the issuance scales with security spend, PUMP's burn scales with one thing only: how much Pump.fun earns.
Where PUMP differs sharply from the mature members of that class is float maturity. An exchange token running a buyback typically has most of its supply already circulating, so the buyback works against a nearly static denominator. PUMP does not: 440.13B PUMP of the 1 trillion cap still sits outside the float, and the insider calendar that governs a large share of it only started paying out in Jul 2026. That is the structural difference a reader should carry away — PUMP's burn is genuinely one of the strongest in crypto by revenue share, and it is still running against a vesting calendar young enough to outrun it.
The second structural difference is denomination. The Pump.fun buyback is written in dollars, not in tokens: it removes a fixed share of revenue, so the number of PUMP it destroys falls as the PUMP price rises. The trailing window bought PUMP at an average of about $0.0023 while PUMP trades near $0.0043 today, which means the same revenue now buys roughly half as many PUMP. A comparison that ignores this reads any revenue-funded buyback as more deflationary than it is.
What to watch in the next 90 days
Sep 12 2026, Oct 12 2026 and Nov 12 2026 each release a fixed 6.875B PUMP insider tranche — 20.63B PUMP in total, and the only dated sell pressure inside the forward window. The two paying wallets holding 6.99B PUMP are the second watch line: if that backlog drains, it becomes realised sell pressure the framework has not yet booked; if it grows, more of the calendar is stalling before it reaches the market. Pump.fun platform revenue is the third — the burn is a straight function of it, and a sustained move above or below the trailing $110.70M per quarter changes the forward reading directly. The fourth is the PUMP price itself, which sets how many PUMP each buyback dollar retires. The fifth is the Apr 2027 expiry of the locked buyback contract, still outside this window but the single largest structural risk on the page.
Summary
The MrNasdog Pressure Framework reads Pump.fun's PUMP at +10.16% net supply growth over the 90 days to Sep 5 2026 and projects +1.86% forward. PUMP can never be minted again — the Solana mint authority is null and 164.42B PUMP of the 1 trillion cap has already been destroyed — so every unit of sell pressure on this page is a release of PUMP that already existed, driven entirely by the insider vesting cliff that cleared in Jul 2026 and delivered 65.21B PUMP to team and early-investor wallets. Against that, Pump.fun's contractual 50%-of-revenue buyback bought and burned 25.02B PUMP. The key risk is that the buyback is denominated in dollars while the unlocks are denominated in tokens: at current PUMP prices the same revenue retires roughly half as many PUMP as it did through the window, so three monthly tranches of 6.875B PUMP still outweigh the burn. The ceiling is absolute at 1 trillion PUMP and falling, but 440.13B PUMP remains outside the float and the calendar that governs it runs to 2029.
MrNasdog Pressure Framework analysis of PUMP, Metric 1 — Inflation. Data + explanation only. Not financial advice. Updated Sep 5 2026.
Top comments (0)