SKY Inflation Analysis · September 2026 · Supply was growing, trend cooling
Sky has not created or destroyed a single SKY in three months — the SKY token contract returns 23,462,665,147 at both ends of the quarter, identical to the wei — and yet the Pressure Framework reads +0.67% of net supply reaching the market. The whole figure is one wallet: the Sky protocol treasury paid 224.4M SKY of staking rewards out, while the Smart Burn Engine bought 68.3M SKY back off the open market and put every coin into that same treasury instead of a burn address. Our inflation monitor reads +0.68% for the same window, a gap of 0.02 percentage points, so no data-conflict flag is raised. A governance change executed on Aug 17 2026 cut the SKY reward stream to 96.9M and sped the buyback up to 93.1M, which flattens the next 90 days to about +0.02%.
The verdict, in one paragraph
Over the 90 days to Sep 8 2026 the SKY ledger reads 224.4M of sell pressure against 68.3M of buy pressure — a net of 156.1M SKY, or +0.67% on a circulating base of 23.43B SKY. The inflation monitor reads +0.68%. The two agree to 0.02 percentage points, inside tolerance, so no ⚠ chip ships — and the reason they agree so closely is worth stating plainly: they are measuring the same event from opposite sides. The monitor watches the counted float grow; the framework watches the Sky protocol treasury drain by the same 156.1M SKY. On a 23.4 billion base those are one number. Sky is best labelled a closed-loop redistributor— a governance token whose supply did not change at all, whose measured inflation is entirely a treasury balance passing into stakers’ hands, and whose buyback exists to refill that treasury rather than to shrink the supply.
Sell pressure: where new SKY comes from
None of it is new. Sky minted nothing: the SKY contract on Ethereum reported the same total at the start and the end of the window, and again a year earlier, so every coin counted here already existed. That flat reading is a real measurement rather than a hard-coded constant — the total lives in writable storage, the value does not appear anywhere in the compiled contract, and calling the mint function returns the contract’s own Sky/not-authorized refusal, which is what a live, permissioned mint looks like. Sell #1, protocol inflation, is therefore 224.4M SKY of staking rewards moving out of the Sky protocol treasury into the distributor that pays people who stake SKY — 19 weekly payments, each one traceable, stepping from 18.6M a week in June to 22.2M in late July and then down to 7.5M from Aug 24 2026. Sell #2, vesting unlocks, is zero: every SKY release schedule other than the staking-reward stream has paid out in full and closed, the last of them on Dec 31 2025, and nothing was claimed from either release contract in this window. Sell #3, foundation and unscheduled unlocks, is zero because the treasury releases are already counted in Sell #1 and would otherwise be counted twice; the balance itself is tracked as overhang below. Sell #4, long-term locked or bankruptcy, is zero because SKY has no estate, no trustee and no court-ordered distribution attached to it. A fifth row tracks the old-token upgrade pool, where 116.7M SKY left the MKR-to-SKY converter this quarter; it counts as zero because no SKY is minted to serve a conversion and those coins are already inside the counted float.
Buy pressure: where new SKY goes
Buy #1, the programmatic buyback, is 68.3M SKY across 891 separate open-market purchases — roughly 4.7M dollars of protocol surplus spent through a single Uniswap V2 USDS/SKY pair. The important detail is where those coins land. The Smart Burn Engine’s receiver address is the Sky protocol treasury, which is the same wallet the staking rewards are paid out of, so the buyback is a buy-and-park, not a burn — and on this configuration it now functions as the funding line for the rewards rather than as a separate reduction in supply. Buy #2, protocol fee burn, is a real zero, and it was checked on both surfaces a burn could possibly appear on: the count of SKY in existence was identical to the wei at both ends of the window, and the dead address held the same 4.82 SKY on both dates. Neither moved by any amount, so nothing was destroyed, and third-party write-ups that describe the Smart Burn Engine as destroying SKY are describing a configuration Sky is not running. Buy #3, foundation buy, is zero: every SKY that arrived in the treasury over the window came from that one trading pair, with no second discretionary buyer anywhere in the flow. Buy #4, new long-term lock, is zero as well — staking did grow by 211.4M SKY, but staked SKY can be withdrawn at will and is already counted as tradable, so locking it up removes no supply from this reading.
Foundation and overhang
Sky has exactly one team-controlled overhang and it is unusually easy to watch. The Sky protocol treasury — the DAO pause proxy — is the only SKY holding the market does not already count as tradable, and it fell from 189.5M SKY to 33.4M SKY across the window. It is read from the chain on every rebuild, so a daily refresh is possible and used. The second pot is not team-controlled but belongs in the same paragraph: the MKR-to-SKY upgrade contract still holds 2.08B SKY owed to anyone who has not swapped their old MKR yet, against only 85,286 MKR still outstanding, and it is already counted as tradable — it is pre-funded, so a conversion mints nothing and shifts nothing. The trigger sentence applies to the first pot: if the Sky protocol treasury’s balance falls between refreshes, that outflow enters Sell #1 at the next refresh, which is exactly what produced this quarter’s number. And the size of the pot is now the whole story. At 33.4M SKY the treasury holds about five weeks of rewards at the new rate, which is only survivable because the buyback pays back in almost exactly what the rewards pay out.
How SKY compares to other revenue-funded buyback tokens
Against an exchange token that runs a quarterly auto-burn, Sky sits on the opposite side of one mechanical line: the exchange token’s buyback ends at a burn address and the supply falls, while Sky’s ends at a treasury and the supply does not move at all. Both are revenue-funded and both take coins off the open market, but only one of them is deflationary in the ledger sense. That distinction is why a page can read the SKY buyback as real demand and still show growing supply — the coins come off the market and then go straight back out as rewards.
Against an uncapped continuous-emission layer 1, Sky is the safer shape and the reading understates it. A proof-of-stake chain pays validators with coins that did not exist before, so its float grows whether or not anyone claims anything. Sky pays stakers out of a pot that was minted years ago, so the ceiling holds no matter how generous the reward schedule gets — the only thing that changes is how fast the pot empties. The cost of that shape is the one Sky is now facing: a pot can run out, and a chain’s issuance cannot.
Against a hard-capped chain with a halving schedule, the difference is who decides. A halving is written into consensus code and no vote can move it. Sky’s 23.46B ceiling is a governance-held ceiling: the mint function is live and the DAO pause proxy is still authorised to call it, which we verified by calling it. Nothing has been minted in a year, and the framework reads that as a policy that has held rather than as scarcity that is enforced. It is the single most important structural caveat on this page, and it is why no row here is marked permanent.
What to watch in the next 90 days
First, Nov 15 2026: the staking-reward stream set on Aug 17 2026 releases 96,903,706 SKY over exactly 90 days and expires on that date. Sky governance has renewed it at every rollover since Oct 27 2025, and the size it is renewed at is the single number that decides this page’s next reading. Second, the buyback rate: the Aug 17 change cut the interval between purchases from about four hours to about one and set the split at 55% of protocol surplus to buying SKY — the first week ran hot at 1.14M SKY a day clearing a backlog before settling at 1.03M, so any further parameter change, or a fall in Sky’s own revenue, moves the buy side directly. Third, the treasury balance at 33.4M SKY — if it drops much further while the reward stream continues, the DAO has to choose between cutting rewards again and minting, and the mint path is live. Fourth, the newly initialised bounded-access module carries a 350M USDS per year ceiling on buybacks, roughly thirteen times the realised rate, so the authorised headroom is very large even though the actual spend is not. Fifth, the burn surfaces: nothing has been destroyed in a year, and the first movement on either of them would change the shape of this page rather than just its number.
Summary
The MrNasdog Pressure Framework reads SKY at +0.67% over the trailing 90 days and +0.02% projected forward: supply was growing, trend cooling. The structural mechanism is not issuance but redistribution — Sky minted nothing and burned nothing all quarter, and the entire reading is 224.4M SKY of staking rewards leaving one protocol treasury against 68.3M SKY of open-market buyback flowing back into it. After the Aug 17 2026 change those two flows are matched almost coin for coin, at 96.9M out and 93.1M in over the next 90 days. The key risk is that the balance is now thin rather than comfortable: the treasury is down to 33.4M SKY, so the buyback is no longer supply support sitting on top of the rewards — it is what pays for them, and if protocol revenue falls the DAO must cut rewards again or mint. The ceiling is the qualified comfort: 23.46B SKY has not moved in a year, but it is held by governance rather than by code.
MrNasdog Pressure Framework analysis of SKY, Metric 1 — Inflation. Data + explanation only. Not financial advice. Updated Sep 8 2026.