Originally published at https://mrnasdog.com/research/tezos/inflation

XTZ Inflation Analysis · August 2026 · Supply growing, projected to keep growing

Tezos minted 8,218,269 XTZ of staking rewards over the last 90 days under adaptive issuance — a mechanism that recomputes the chain's own yearly issuance rate every cycle from how much XTZ is staked — and cut that rate from 3.195% to 2.970% a year while doing it. Against that mint, only 59,329 XTZ was destroyed, and the largest part of the burn was not fees at all but 30,000 XTZ of forfeited Etherlink rollup dispute bonds. The MrNasdog Pressure Framework reads Tezos at +0.75% net over the trailing quarter and +0.74% forward, against a supply-monitor reading of +0.76%. XTZ has no maximum supply and no buyback of any kind.

The verdict, in one paragraph

Over the last 90 days the MrNasdog Pressure Framework reads Tezos at +0.75% net: 8,222,857 XTZ of new supply against 59,329 XTZ destroyed, on a circulating base of 1,094,741,523 XTZ. The supply monitor reads the same window at +0.76%, a gap of 0.02 percentage points — comfortably inside the half-point tolerance, so no monitor-gap flag is raised. That agreement is structural rather than corroborating: both readings ultimately measure the same circulating counter, and the framework's contribution here is the decomposition, not the total. Two things only the decomposition surfaces are the 30,000 XTZ burned by lost rollup bonds and the 16,391,567 XTZ that flowed into staking and removed nothing at all. Tezos is structurally inflationary with a self-damping rate: the more XTZ that gets staked, the less the protocol issues, and the staked share is still climbing.

Sell pressure: where new XTZ comes from

Sell #1 — protocol inflation — is the entire new-supply story for Tezos, because the chain has exactly one issuance path. New XTZ exists only as baking, attesting and data-availability rewards, minted under adaptive issuance, and there is no maximum supply to run into. Adaptive issuance is unusual enough to be worth stating plainly: the protocol looks at the fraction of supply that is staked, compares it to a roughly 50% target, and moves its own yearly issuance rate inside a fixed band accordingly. This build read that rate straight off the node at both ends of the window — 3.195% a year on May 31 2026 and 2.970% on Aug 29 2026 — and it fell because the staked share of Tezos rose to 30.8%. What actually shipped in the ledger is not either of those rates but the chain's own mint counter, which moved 8,218,269 XTZ across the 90 days. Annualised, that realised figure is 3.01%, sitting inside the band the two rate reads bracket — a clean cross-check between the published parameter and the measured flow.

A protocol upgrade landed inside this window, so the framework tested it rather than assumed it. Ushuaia, the twenty-first Tezos upgrade, activated on Jun 30 2026 at block 13,857,889. Comparing protocol constants either side of the switch shows every monetary field unchanged — the same six-second block target, the same 14,400-block cycle — and the upgrade's actual content was a fifteenfold rise in data-availability bandwidth, dynamic attestation for that layer, rollup governance, and liquid staking shipped to testnet only. No issuance parameter moved, so the trailing quarter is not a blend of two mechanisms and needs no re-basing. The measured block interval was 6.027 seconds against a six-second target, meaning a nominal calculation would over-state the mint by 0.45%; because the shipped figure is the realised counter, that measurement stays a cross-check and is never applied twice.

Sell #2 — vesting unlocks — is zero and permanently so: the four-year vesting schedule that released the Tezos Foundation and early-team allocations finished on Sep 17 2022, the contracts it ran through are empty, and the allocation is physically spent. Sell #3 — foundation and unscheduled unlocks — is also zero, proven on the transfer log rather than on a balance comparison: this build swept every transaction touching all fourteen identified Tezos Foundation wallets across the window, and thirteen of them had no transactions at all while the fourteenth had a single 8.54 XTZ payment coming in. Sell #4 — long-term locked or bankruptcy — is zero structurally, since Tezos has no bankruptcy estate and no trustee distributing XTZ on a court schedule. The one extra sell row, Sell #5, is a genuine oddity of this chain: 19,981,067 XTZ of 2017 fundraiser allocations have still never been claimed, they can be claimed at any time with no expiry, and one claim of 4,588 XTZ landed on Jul 28 2026.

Buy pressure: where new XTZ goes

Buy #1 — programmatic buyback — is zero, and it is zero by design rather than by neglect. Transaction fees on Tezos are baker income, not funding for a repurchase contract, and no governance proposal exists to change that. Buy #3 — foundation buy — is zero on the same wallet sweep that cleared Sell #3: no identified Foundation address bought XTZ on the open market during the quarter, though the four Foundation bakers did grow about 1,010,350 XTZ from reward accrual, which is not buying. Buy #4 — new long-term lock — is zero despite staking absorbing 16,391,567 XTZ across the window, because unstaking on Tezos takes days rather than years and staked XTZ is still counted as circulating supply. That is custody, not a lock, and booking it as buy pressure would invent an offset roughly twice the size of the entire quarterly mint.

The burn is where Tezos gets interesting, and it had to be checked on both surfaces. The dead-address route is nearly empty: the public burn address held 283 XTZ at the open and 458 XTZ at the close, a rise that rounds to nothing and belongs to NFT burns rather than to supply policy. The protocol counters are where the real destruction happens, and they moved 59,329 XTZ in total. Splitting that by time rather than by differencing surfaces gives three distinct mechanisms. Buy #2 — protocol fee burn — is 17,992 XTZ, the fixed price-per-byte charge that storing data on Tezos destroys, running at roughly 200 XTZ a day. Buy #5 is 30,000 XTZ of forfeited rollup bonds: Etherlink operators post a 10,000 XTZ bond to publish, lose it if a dispute proves them wrong, and half of each lost bond is burned — six disputes were lost inside this window. Buy #6 is 11,337 XTZ sent to an unspendable address, every single transfer of it originating from the protocol's own liquidity-baking exchange contract. The ranking is the finding: lost dispute bonds out-burned the entire fee mechanism by roughly five to three, on a channel that appears in no tokenomics page and on no calendar.

Foundation and overhang

Three overhangs are tracked on Tezos. The Tezos Foundation runs four bakers holding 32,800,555 XTZ between them, and ten labelled delegator wallets holding a further 50,312,181 XTZ — about 83,112,736 XTZ in total, close to 7.6% of the float, on no published release plan. Those balances are read straight off the chain and re-checked on every rebuild. Alongside them sits the unclaimed 2017 fundraiser pool of 19,981,067 XTZ, which is not team-controlled but is real supply sitting outside the tradable float, claimable at any time by the original participants with no expiry and no schedule. Over the last twelve months that pool released eleven claims totalling 82,913 XTZ — regular enough in occurrence to project forward, erratic enough in size that the trailing-year average is the honest number. There is no buyback accumulation wallet to watch, because there is no buyback. If any of these balances falls between refreshes, the outflow enters Sell #3 — or Sell #5 for the fundraiser pool — at the next refresh.

How XTZ compares to other liquid proof-of-stake chains

The obvious comparison class is uncapped continuous-emission layer ones, and Tezos sits at the quiet end of it. Chains with a flat per-block subsidy mint the same amount whatever the network does, which is why several of them run materially above 4% a year with nothing dampening them. Tezos is one of the few production chains whose issuance rate is a live function of participation rather than a constant: the rate genuinely fell from 3.195% to 2.970% inside a single quarter, without a vote, purely because more XTZ was staked. That is a structurally different shape from a halving-model chain with a hard cap, where the schedule is fixed years in advance and participation changes nothing, and a different shape again from an exchange token running quarterly buybacks, where the buy side is a discretionary corporate decision that can be paused.

Where Tezos looks weaker than its peers is the offset. Layer ones that added a base-fee burn destroy a meaningful fraction of what they mint, and in strong quarters some of them tip net-negative outright. Tezos returns roughly one XTZ for every 457 it creates through ordinary chain use, so the burn is not a supply lever in any practical sense — it is a rounding line. The absence of any buyback means there is no discretionary counterweight either: the only thing that moves the net figure meaningfully is the mint, and the only thing that moves the mint is the staked ratio. That makes Tezos unusually predictable to model and unusually dependent on one variable. The other structural note worth carrying across the comparison class is that the biggest single burn of the quarter came from a rollup dispute mechanism, which is a channel that will exist on every chain hosting bonded rollup operators and is currently classified by nobody.

What to watch in the next 90 days

The first watch line is the staked ratio, currently 30.8%. Adaptive issuance targets roughly 50%, so every further point of staking pulls the yearly rate down and pulls the framework's forward reading with it — this is the single variable that moves the number. The second is the governance vote on protocol-native liquid staking, shipped behind a flag in the Ushuaia upgrade of Jun 30 2026 and switched off pending approval: it would remove no supply directly, but it would likely raise the staked share and therefore lower issuance. The third is the Etherlink rollup dispute channel, which burned 30,000 XTZ across six lost bonds in this window and carries forward at zero because nothing schedules it — a repeat would be a genuine surprise on the buy side. The fourth is the unclaimed fundraiser pool of 19,981,067 XTZ, where a single large claim can arrive on any day with no warning. The fifth is the Tezos Foundation's 83,112,736 XTZ, which has not moved a coin outward this quarter and would change the reading materially if it started to.

Summary

The MrNasdog Pressure Framework reads Tezos at +0.75% net over the trailing 90 days and +0.74% forward, matching the supply monitor to within 0.02 percentage points. The structural mechanism is adaptive issuance: staking rewards are the only way a new XTZ can exist, and the protocol lowers its own yearly rate as the staked share rises — from 3.195% to 2.970% in this quarter alone. The key risk is that there is no offset worth the name: no buyback exists, the fee burn returns about one XTZ in every 457 minted, and the largest destruction of the quarter came from six forfeited rollup dispute bonds that nothing schedules and no tracker classifies. And the ceiling is that there is no ceiling — Tezos has no maximum supply, so the only brake on issuance is the staked ratio itself.

---

MrNasdog Pressure Framework analysis of XTZ, Metric 1 — Inflation. Data + explanation only. Not financial advice. Updated Aug 30 2026.