Originally published at mrnasdog.com/research/avax/inflation by MrNasdog.
Avalanche is hard-capped at 720,000,000 AVAX and destroys every transaction fee it collects, and the Pressure Framework still reads AVAX at +0.81% over the trailing 90 days and +0.81% over the next 90. Two mechanisms do it: the network minted 1,864,765 AVAX to pay stakers, read straight off Avalanche's own supply counter at both ends of the window, and the Avalanche Foundation released a 1,667,000 AVAX vesting cliff on Jul 26 2026. Against 3.532M AVAX of sell pressure the fee burn removed 37,020 AVAX — one coin for every 95 added. The cap is real; it is also 245.83M AVAX away.
The verdict, in one paragraph
Against a circulating base of 431.772M AVAX, the framework books 3.532M AVAX of sell pressure and 0.037M of buy pressure over the trailing 90 days — a net of +0.81% — and projects +0.81% again for the next 90 days, because the same staking mint runs continuously and another 1,667,000 AVAX Foundation cliff lands on Oct 25 2026. The inflation monitor reads -0.05% for the same window, a gap of 0.86 percentage points, which is over the framework's 0.5pp tolerance and therefore ships with a monitor-gap warning on the overview page. That gap has a single cause and it is not a disagreement about Avalanche: the tradable figure the monitor divides is total supply minus a flat 31,669,100, a locked constant that has not been refreshed since Feb 2026 and is now two Foundation cliffs stale, so the monitor cannot see either the mint or the unlock. The label for AVAX is a hard-capped chain still deep inside its issuance phase: the ceiling constrains the end state, not the next quarter.
Sell pressure: where new AVAX comes from
It comes from staking, and Avalanche counts it in one place. The Avalanche network spans three chains — the P-Chain for staking and coordination, the X-Chain for asset transfers, the C-Chain for smart contracts — but there is a single network-wide supply counter, carried on every P-Chain block, so the number of AVAX ever created is readable at any past moment rather than estimated. It read 472,300,998 at the start of this window and 474,165,763 at the end. That is Sell #1, protocol inflation, at 1,864,765 AVAX — brand-new coins, minted out of the half of the 720M ceiling that was reserved at launch to pay validators and delegators. Avalanche's reward rule is a yearly percentage on staked AVAX scaled by how much of the ceiling is still unminted, so it decays as the reserve drains: staked AVAX fell from 223.48M to 213.65M across the window while the unminted reserve fell from 247.70M to 245.83M. There is no per-block subsidy anywhere in that rule, which is why Avalanche's dynamic block times change issuance by exactly nothing.
Sell #2, vesting unlocks, is 1,667,000 AVAX. Every allocation from the 2020 Avalanche launch has finished releasing except the Avalanche Foundation's 9.26%, which hands over 1,667,000 AVAX once a quarter on an unbroken calendar running to 2030. One cliff landed inside this window, on Jul 26 2026, and the next lands on Oct 25 2026. No AVAX is minted by a vesting cliff — the coins already exist and are already inside the 720M ceiling — but they stop being time-locked and become tradable, which is the thing the Pressure Framework measures. Avalanche holds these under a protocol-level locktime rather than in a claim contract, so nothing can sit vested-but-undrawn: what is scheduled and what actually reaches the market are the same number. 38,341,000 AVAX has been released this way and 28,339,000 is still to come.
Sell #3, Foundation and unscheduled unlocks, is 0 — nothing left the Avalanche Foundation beyond the quarterly cliff above, and no dated release was found anywhere in the window. Sell #4, long-term locked or bankruptcy, is 0 as well: AVAX has no bankruptcy estate, no trustee and no court-ordered distribution attached to it.
Buy pressure: where new AVAX goes
Buy #1, programmatic buyback, is 0. Avalanche runs no buyback of any kind: the reward rule only creates AVAX, no share of network revenue is routed into repurchasing it, and nothing was announced or executed inside the window.
Buy #2, protocol fee burn, is 37,020 AVAX, and it is worth being precise about how that is proved, because Avalanche gives you nothing convenient to read. There is no burn address. Avalanche destroys the entire transaction fee — base fee and tip, on all three chains — in place, and sampled C-Chain blocks name a blackhole as the fee recipient, so no balance anywhere rises when AVAX is burned. The proof is an identity instead: Avalanche's supply counter says 474,165,763 AVAX have been created, the classified supply figure says 463,441,061 exist, and the missing 10,724,702 went somewhere no balance can show. Had the fees merely moved to a recipient, those two numbers would agree. A second independent classifier says the same thing from the other end, publishing a burn-adjusted ceiling of 715,748,719 against the 720,000,000 cap. For this window the burn is summed off Avalanche's own fee record at 411 AVAX a day, and the reason it is small is that an earlier Avalanche upgrade cut the C-Chain base fee by roughly 96%: a fee burn that removes one coin for every 95 the chain creates is a mechanism that exists rather than a mechanism that matters.
Buy #3, Foundation buy, is 0. The Avalanche Foundation bought no AVAX on the open market; its arrangement with the listed AVAX treasury company runs the other way, selling to it at a discount, which is a sell-side channel rather than demand. Buy #4, new long-term lock, is 0, and staking moved the wrong way for it anyway — staked AVAX fell from 223.48M to 213.65M. Staking would not count here in any case, because staked AVAX is already treated as tradable float and because staking is precisely what produces the new coins in Sell #1. A fifth, tracked row covers corporate treasury accumulation and also reads 0: the company built its roughly 13.8M AVAX position in late 2025 and has disclosed no purchase since.
Foundation and overhang
The AVAX overhang is real, and it is honestly opaque rather than enumerated. The Avalanche Foundation publishes no wallet address, and its remaining allocation sits across many separate time-locked outputs instead of one readable escrow, so this build does not claim an enumeration it cannot make. The first item is that allocation itself: 28,339,000 AVAX still to release in 1,667,000 quarterly steps to 2030, tracked by schedule and re-walked at every rebuild. The second is the wider gap between AVAX that exists and AVAX counted as tradable, 31,669,100 — a figure that has not been refreshed in two quarters and is therefore the same overhang measured with a stale ruler. The third is the listed corporate treasury, roughly 13.8M AVAX, of which about 7.8M is pledged as loan collateral and which disclosed a going-concern doubt alongside a large quarterly loss.
The trigger sentence applies to all three: if any of these balances falls between refreshes beyond what the published schedule accounts for, that outflow enters Sell #3 at the next refresh. The third one deserves the most attention, because it is the only pot here with a spender rather than a calendar — a forced sale against pledged AVAX would arrive with no notice at all, while the Foundation's 1,667,000 is on a clock everyone can read.
How AVAX compares to other hard-capped proof-of-stake chains
AVAX belongs to the small group of proof-of-stake networks with a genuine hard cap written into the protocol, and comparing it to a halving-model chain is the clearest way to see what that does and does not buy. Bitcoin also has a fixed ceiling, but it is roughly 95% issued, so its remaining emission is a rounding error and its inflation reading sits near zero. Avalanche is 65.9% issued: 245.83M AVAX of the 720,000,000 has never been minted, and that reserve is the fuel for every future staking reward. The cap is not a scarcity claim about the next quarter — it is a promise about a state that arrives after the reserve empties, and the reward rule is deliberately shaped so the reserve never quite does.
Against uncapped continuous-emission layer ones, AVAX looks strict. A typical uncapped proof-of-stake chain runs a staking-linked emission of 4% to 8% a year with no ceiling at all, and its reward rate does not decay as supply grows. Avalanche's does, by construction: the reward is scaled by the unminted remainder, so every coin minted slightly slows the next one. The measured rate here — 1,864,765 AVAX in 90 days, near 1.7% annualised on circulating — is already well below the class average and drifting down.
The comparison that matters most is against exchange tokens with quarterly buybacks, because those are the coins whose inflation readings go genuinely negative. They pair issuance with a demand-linked removal that scales with revenue, and they spend real money doing it. Avalanche has the burn half of that shape and none of the buyback half, and its burn is starved by its own success at making transactions cheap. For the fee burn alone to hold AVAX flat, Avalanche would need to burn 3.532M AVAX a quarter instead of 37,020 — roughly 95x the current volume. That is the whole distance between a chain that destroys every fee and a chain that is deflationary.
What to watch in the next 90 days
First, the Foundation cliff on Oct 25 2026: 1,667,000 AVAX, dated, unavoidable, and roughly half of the entire quarter's sell pressure. Second, ACP-285, the Avalanche proposal that cuts the minimum consumption rate from 10% to 7.5% and would pull the network's overall reward rate down by roughly a third; it activated on the Avalanche test network on Jul 28 2026 and is bundled into the Helicon upgrade, but it carries no mainnet date, so this build books it at zero and treats a mainnet activation as the single most likely thing to lower this page's number. Third, the staked total, at 213.65M AVAX and falling — the mint is proportional to it, so a continued slide cuts Sell #1 without any governance action at all. Fourth, C-Chain activity, because the burn is measured from real fees and would need a genuine order-of-magnitude change to register. Fifth, the listed AVAX treasury company, whose pledged 7.8M AVAX and going-concern flag make it the one place a large, unscheduled sell could originate.
Summary
The MrNasdog Pressure Framework reads AVAX at +0.81% over the trailing 90 days and +0.81% projected forward: supply growing, projected to keep growing. The structural mechanism is ordinary staking issuance plus a quarterly Foundation vest — 1,864,765 AVAX minted and 1,667,000 unlocked — against a fee burn that removed only 37,020 because Avalanche made its own transactions almost free. The key risk is that neither leg is discretionary: the mint runs every second and the cliff lands on Oct 25 2026 regardless of price, while the offsetting burn depends on activity Avalanche has deliberately made cheap. The ceiling is the genuine comfort and the genuine caveat at once — AVAX can never exceed 720,000,000, but 245.83M of that has never been minted, so the cap is a destination rather than a floor under today's float.
MrNasdog Pressure Framework analysis of AVAX, Metric 1 — Inflation. Data + explanation only. Not financial advice. Updated Sep 8 2026.
Top comments (0)