DOT Inflation Analysis · August 2026 · Supply growing, projected to keep growing
Polkadot minted 13.99M DOT over the last 90 days and destroyed nothing at all, a net +0.82% of the 1.70B circulating supply, against our own monitor at +0.99% — a gap of 0.17 percentage points, well inside tolerance, so there is no data conflict to flag. Polkadot is now a hard-capped nominated-proof-of-stake network: a Mar 2026 governance vote fixed a permanent ceiling of 2.1B DOT and cut annual issuance 53.6%, to a stepped curve that issues 13.14%of the remaining gap every two years. The same vote switched every burn off, so DOT’s supply is slow, bounded and strictly one-directional.
The verdict, in one paragraph
Over the 90 days to Aug 23 2026 the Polkadot sell ledger totals 13.99M DOT and the buy ledger is zero, for a net of +0.82% against a circulating supply of 1,699,263,079 DOT. Our inflation monitor, measuring the same network from the opposite direction, reads +0.99%; the gap of 0.17 percentage points is inside the half-point tolerance, so no warning chip is shown. Getting that agreement required reading the right chain. Polkadot moved every DOT balance off its relay chain in Nov 2025, and the relay’s own supply counter now holds just 242,486 DOT — a teleport mirror worth 0.014% of the real figure. A build that reads the historic relay key would report a supply four orders of magnitude too small. Read correctly, Polkadot is mildly inflationary on a capped float with no downward lever.
Sell pressure: where new DOT comes from
All of it is one row. Protocol inflation on Polkadot means staking and validator rewards minted under the schedule ratified in Mar 2026, and the combined supply counter — Asset Hub plus the relay-chain mirror — rose from 1,685,525,431 DOT to 1,699,517,092 DOT across the window, an increase of 13.99M DOT, about 3.36% a year. That figure is measured, not quoted. The live runtime encodes the curve as constants — a Mar 2026 starting supply of 1,674,342,153 DOT, a target of 2,100,000,000, and a rate of 26.28% of the remaining gap per two-year step — which works out to 153,132 DOT a day. The chain actually paid 153,192 a day after the mid-window upgrade, a ratio of 1.0004. On a network whose emission is governance-set rather than mined, that agreement between a stored parameter and a realised flow is the check that matters, and Polkadot passes it cleanly.
One nuance shapes the forward column. A runtime upgrade enacted Jun 12 2026 switched Polkadot staking to non-minting and moved emission into an allocation buffer that drips rewards out on a governance-set split. Across the full window the measured pace is 155,463 DOT a day; entirely after the change it is 153,192. The trailing column keeps what was measured, and the forward column carries only the post-change rate, giving 13.79M DOT for the next 90 days. Every other Polkadot sell row is zero. Vesting unlocks are zero because there is no unlock calendar and no cliff: 1,047 accounts hold 30.56M DOT in on-chain vesting schedules, but those coins were minted years ago and already sit inside the counted float, and the locked total actually rose from 29.88M across the window. Unlock trackers do publish a monthly DOT vesting line running to Mar 2028— that is the protocol’s own issuance schedule relabelled as vesting, and counting it would book the same 13.99M twice. Bankruptcy is zero because Polkadot launched from a public sale in 2017 and has never been through an insolvency.
Buy pressure: where new DOT goes
Nowhere. Every Polkadot buy row is zero, and the most important of them is zero for a reason that changed this year. Polkadot used to destroy unspent treasury funds at the end of each 24-dayspend period — a real destruction channel with no burn address to watch, the kind that only ever shows up in the chain’s own accounting. The Mar 2026 vote ended it. In the live runtime the treasury’s burn setting and its burn destination both read empty, validator slashes are routed to the allocation buffer rather than destroyed, and transaction fees and coretime sale revenue are routed there too. Every channel that used to remove DOT is now a transfer. We checked it from the other end as well: the supply counter was sampled at both window edges and at fourteen points in between, and it rose in every single interval. Not one DOT was destroyed in 90 days.
Polkadot runs no buyback of any kind, and no vote has proposed one, so there is no accumulation wallet to track. No Foundation or Labs entity has disclosed a DOT purchase programme, and no entity wallet set is published to sweep — the only identified pools on the chain are protocol accounts, and all of them took DOT in rather than buying it. New long-term locks are zero too, and this is Polkadot’s easiest trap. Staking looks like a lock and is not one: nominators became unslashable in Apr 2026 and their exit was cut to two eras, roughly two days, while the ordinary bond still reads 28. Bonded DOT is already inside the counted float either way, so a larger stake removes nothing. Booking staking growth as buy pressure would invent demand that never existed.
Foundation and overhang
Polkadot has no published Foundation wallet set, so the overhang here is protocol-controlled rather than entity-controlled, and all of it is readable. The on-chain Treasury held 24,310,026 DOT at the end of the window, up from 23,118,982, and spends only by passing a governance vote. The validator reward pots held 33,858 DOT. The one that deserves attention is new: the allocation buffer introduced with the Mar 2026 reform held 0.01 DOT on May 24 2026 and 3,087,189 DOT on Aug 22 2026, because a June vote set its split at 45.2% to stakers, 22.6% to validator self-stake incentives and 32.2% held back as a buffer. That means roughly 22%of the quarter’s entire mint was created and then parked, and a single governance vote can release it. We read all three pools at both window ends; all three finished larger than they started, so no outflow is booked. If any of these balances falls between refreshes, the outflow enters the sell ledger at the next refresh.
How DOT compares to other hard-capped proof-of-stake chains
Polkadot has just crossed from one structural class into another, and the comparison that matters is with where it came from. For years DOT belonged with the uncapped continuous-emission layer ones — an unbounded schedule issuing roughly 120M DOT a year. It now belongs with the hard-capped chains, but its cap works nothing like Bitcoin’s. A halving chain steps its subsidy down by a fixed fraction of the subsidy; Polkadot steps down by a fixed fraction of the distance still left to the ceiling, so issuance decays asymptotically and the 2.1B cap is approached rather than reached. The practical difference is that DOT’s emission never hits zero and never has a dramatic cliff — 13.14% of a shrinking gap every two years is a smooth glide, and the next step is not until Mar 2028.
Against the fee-burning chains the contrast is sharper, and it runs the wrong way for DOT. Networks with a base-fee burn have a counter-cyclical brake: heavy usage destroys coins and can push net supply negative. Polkadot deliberately gave that up. Its fee revenue, coretime sales and slashed stake used to end in a treasury that burned its surplus, and they now end in a buffer that governance re-allocates. That is a defensible design choice — burning surplus is a blunt instrument, and a network that funds itself needs the money — but it means DOT has no mechanism at all that can shrink supply, at any usage level. Compared with a delegated-proof-of-stake chain that pairs staking issuance with a real burn, Polkadot is the cleaner read and the weaker one: nothing hidden, nothing offsetting.
Where Polkadot does score well is predictability, which is the property the Pressure Framework is actually built around. DOT’s issuance is a constant written into the runtime, not a curve that re-tunes itself off the staked ratio and not a discretionary emission a foundation can accelerate. It is knowable two years ahead to four decimal places, and this quarter it behaved exactly as written. Very few uncapped-turned-capped networks can say that.
What to watch in the next 90 days
First, the allocation buffer. It is filling at roughly 43,000 DOT a day and stood at 3.09M on Aug 22 2026; a governance vote that releases it into rewards or treasury spending would put DOT on the market that the ledger currently treats as parked. Second, Phase 2 of the economic reform, which the project has said is coming but has not dated — it is the most likely route by which either the split or the issuance curve changes again. Third, further runtime upgrades: three landed inside this window alone, on Jun 12, Jun 21 and Jul 27 2026, and the June one changed how emission is paid. Fourth, any proposal to restore a burn — the treasury’s burn setting is empty rather than removed, so a vote could switch it back on and give DOT a downward lever for the first time since March. Fifth, the next step in the issuance curve, which is fixed and far off: Mar 2028, when the rate recomputes to roughly 41M DOT a year.
Summary
Polkadot minted 13.99M DOT in 90 days and destroyed none, for a net +0.82% that projects to +0.81% — slow, bounded, and entirely one-directional. The structural mechanism is a stepped issuance curve ratified in Mar 2026 that fixed a permanent 2.1B ceiling and cut emission 53.6%, and the chain paid out within 0.04% of what that curve specifies. The key risk is not the mint but the parking lot beside it: 3.09M DOT of this quarter’s issuance sits in an allocation buffer that did not exist in May and that a single vote can release, on top of a 24.31M DOT treasury. The ceiling is real and permanent, but it is an asymptote — DOT will keep diluting at a shrinking rate for decades, and nothing in the current design can ever remove a coin.
MrNasdog Pressure Framework analysis of DOT, Metric 1 — Inflation. Data + explanation only. Not financial advice. Updated Aug 23 2026.