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

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

TON, the coin of The Open Network — renamed Gram on Jun 15 2026 in a name-only change — is uncapped, and it pays for its speed in new coins. A fixed block reward that was never cut when blocks got faster minted 50.67M TON over the last 90 days. A readable early-supporter lock let out another 16.14M TON, and the Telegram-linked treasury wallets released 10.56M TON into their payout rail, against a fee burn of only 0.14M TON. The MrNasdog Pressure Framework reads +2.79% net supply growth on a circulating base of 2.76B TON, and our supply monitor reads +2.38% — a gap of 0.41 percentage points, inside tolerance, so the two readings agree and no flag is raised.

The verdict, in one paragraph

Over the 90 days to Aug 25 2026, the framework reads TON at +2.79% net supply growth — 77.37M TON of new, unlocked and released supply against 0.14M TON removed — and projects +2.83% for the next 90 days, because a network upgrade on Aug 17 2026 lifted block frequency again and the forward view is re-based on the faster post-upgrade rate. Our supply monitor reads the same window at +2.38%, a gap of 0.41 percentage points, which sits inside the framework's tolerance, so no monitor-gap flag is raised. The number worth staring at is a different one: TON's total supply grew only 0.97% over the window — from 5,186.70M to 5,237.23M TON, measured directly from the chain's own state at both ends — while the tradable float grew faster, because coins kept stepping out of a lock and out of a treasury and into circulation. TON is structurally inflationary by block subsidy: an uncapped chain that mints a fixed amount per block, and then made blocks far more frequent without touching the amount.

Sell pressure: where new TON comes from

Sell #1, protocol inflation, is the bulk of the story at 50.67M TON over 90 days. The Open Network has no maximum supply and no proof-of-work subsidy left — the original giver contracts mined out in 2022 — so every new TON in existence today is a block creation fee minted to validators. That fee lives in chain config parameter 14, read three separate times for this build (at the key block opening the window, at the key block closing it, and live): 1.7 TON per masterchain block and 1 TON per basechain block, unchanged at all three, and confirmed a fourth time from the minting field inside real blocks. Counting blocks directly, the masterchain produced 18,946,254 blocks across the window and the basechain about 18.46M block-slots. Annualised, TON is minting roughly 205M TON a year — about 3.9% of total supply, up from the 0.6% the chain used to run at before its 2026 speed upgrades.

Two traps sit inside that count, and this build checked both rather than assuming. First, The Open Network is a sharded chain, and its basechain splits into multiple shards under load — but a split halves the per-block reward, verified directly: during the split period each child block minted 0.5 TON while the merged chain mints 1 TON. Total emission per slot is therefore invariant to sharding, and a build that counted split blocks at the full reward would have overstated this row. Second, TON publishes no supply endpoint at all, so total supply had to be measured rather than looked up: it is the masterchain's own state balance plus the state balance of every live basechain shard, summed at the same block height. Both ends of the window happened to be single-shard, so the reading is complete, and it lands within 0.0003% of the independently published total. Subtracting that measured net growth from the gross mint is what isolates the burn.

Sell #2, vesting unlocks, contributes 16.14M TON — far less than the calendar suggests. The TON Believers Fund is an escrow that early holders voluntarily locked for two years from October 2023, releasing monthly across 36 installments through October 2028. The published entitlement is about 37M TON a month, implying roughly 111M TON over this window. The contract's own balance says otherwise: it fell from 1,277.23M TON to 1,261.09M TON across the 90 days, so only 16.14M TON actually left. Vested-but-unclaimed coins simply stay inside the lock, where they are not tradable. The framework counts the coins that left, not the entitlement — booking the schedule instead would have overstated this row roughly sevenfold.

Sell #3, foundation and unscheduled unlocks, is 10.56M TON, and it is the row that changed most since the last build. The Telegram-linked treasury is not one wallet but a cluster, and this build enumerated it: two treasury wallets holding 248.2M and 122.4M TON, a multisig that routes between them and the network's staking pools, and two Fragment payout wallets holding 79.6M and 36.7M TON. Across the window the main treasury wallet sent five dated tranches into that payout rail — 5M on May 28 2026, 5M on Jun 15 2026, 5M on Jun 22 2026, 10M on Jul 3 2026 and 10M on Aug 8 2026, 35.2M TON in all. But the payout wallets kept most of it: their combined balance rose 24.25M TON over the same period. The same rule that governs the vesting escrow governs the payout rail — coins that moved but never left the cluster are not sell pressure yet — so the row books the cluster's own net decline, 10.56M TON, and a boundary reconstruction of every flow in and out lands at 10.95M, inside 4% of it. Sell #4, long-term locked or bankruptcy, is 0: TON has no bankruptcy estate, and the one enormous frozen block on this chain is a governance freeze that does not thaw until Feb 21 2027.

Buy pressure: where new TON goes

Buy #2, the protocol fee burn, is the only live buy-side mechanism on The Open Network, and it removes 0.14M TON over 90 days. Half of every transaction and storage fee is destroyed rather than paid to the validator, which sounds meaningful until it meets the arithmetic. This build read the burn as its own meter, twice and independently: subtracting measured net supply growth from the gross mint gives 145,989 TON, and sampling the burn field inside 440 masterchain blocks spread across the window gives 138,830 TON — agreement inside 5%. Either way the burn offsets about a quarter of one percent of new supply. It shrank because the same wave of upgrades that multiplied issuance also cut transaction fees roughly sixfold in 2026, and Telegram has said it wants to push the network toward feeless transactions entirely. Reading only the mint leg, or only the burn leg, would give a wildly wrong picture of this chain; read together, the burn is real and immaterial.

The other three buy rows are empty. Buy #1, programmatic buyback, is 0: TON runs no buyback contract and routes no network revenue into purchasing the coin. Buy #3, foundation buy, is 0: neither the project nor Telegram has disclosed discretionary open-market accumulation, and the largest listed corporate holder — which ended its most recent reported quarter with 230.5M TON, about 4.4% of total supply — grew that position through staking rewards rather than purchases, and those rewards are already counted inside Sell #1. It is also a third party rather than the project itself, so its buying would never sit in the Foundation row in any case. Buy #4, new long-term lock, is 0: staking on TON does not remove coins from supply, because validator stake is returned each election round and keeps counting as circulating.

Foundation and overhang

Three team-controlled or governance-controlled overhangs sit behind the float, and all three were read on-chain for this build. The largest is the frozen early-miner reserve: 1,081.39M TON across 171 dormant early-mining accounts, frozen by a validator vote until Feb 21 2027. The single biggest of those accounts holds 112.4M TON and moved by about three coins in 90 days, which is exactly what a frozen account should look like. Second is the undrawn remainder of the early-supporter lock, 1,261.09M TON, draining at the realised rate booked in Sell #2 and monitored on every rebuild. Third is the Telegram-linked treasury cluster, 487.0M TON across five identified wallets, which is the only one of the three that was observed releasing coins in the window, and therefore the only one carrying a value.

Three things deliberately stay out of that count. Validator stake — 1,314.7M TON sitting in the network's election contract — belongs to the validators who posted it, not to the project. Exchange wallets belong to depositors. And the listed treasury company's 230.5M TON is a third-party corporate balance sheet, not team-controlled supply, however large it is. Counting any of them as overhang would inflate the number with coins nobody on the project side can decide to sell. The trigger is the same for all three real overhangs: if any of these balances falls between refreshes, the outflow enters Sell #3 at the next refresh — which is precisely what happened to the treasury cluster this build, moving that row from zero to 10.56M TON the moment the wallets were identified and read.

How TON compares to other uncapped layer-1 chains

TON belongs to the uncapped continuous-emission class of layer-1 chains, and inside that class its mechanism is unusually blunt. Most proof-of-stake chains size issuance as a percentage of supply — a target inflation rate, or a curve that decays as more of the supply is staked — so the number of new coins per year is set by policy and moves slowly. TON instead pays a fixed number of coins per block. That makes emission a function of block frequency rather than of supply, and it means a pure performance upgrade can multiply inflation as a side effect, which is exactly what happened here: the network went from roughly 0.6% to roughly 3.9% annual issuance without anyone editing a tokenomics parameter.

The comparison to fee-burn chains is the second structural difference. Chains that burn a base fee proportional to demand get an automatic counterweight — the busier they are, the more they destroy, and heavy use can flip them net-deflationary. TON burns half of its fees, so the mechanism is the same in kind, but the fee itself was cut roughly sixfold and is heading toward zero by design, because the product goal is feeless payments inside a messaging app. That decouples the burn from activity: more transactions no longer means proportionally more burn. Against hard-capped halving chains the contrast is sharper still — there the forward supply schedule is knowable years out and only falls, while TON's forward schedule is knowable only as long as nobody ships another speed upgrade or passes the reward cut that has been proposed and not enacted.

What TON does have, and most uncapped chains do not, is a very large slice of supply sitting outside the float under identifiable control: a governance freeze, a voluntary lock, and a treasury cluster together account for roughly 2.83B TON. That is a double-edged property. It suppresses today's float, and it means the framework's job on this coin is less about forecasting the mint — which is arithmetic — and more about watching whether those three buckets stay put.

What to watch in the next 90 days

First, chain config parameter 14. A proposal to cut the block reward from 1.7 to 0.35 TON on the masterchain and from 1 to 0.2 TON on the basechain has been discussed since mid-2026 and is still not enacted on-chain; if it lands, Sell #1 falls by roughly 80% and this page's verdict changes in one step. Second, block frequency after the Aug 17 2026 collator activation — the forward projection here is re-based on a seven-day post-upgrade sample, and a longer settled rate could move it either way. Third, the early-supporter lock's monthly release, which drips on the 12th of each month through Oct 12 2028; the realised share has been running near a seventh of the entitlement, and a rise in claim rate would lift Sell #2 sharply. Fourth, the treasury cluster's tranche rhythm — the last two transfers into the payout rail were 10M TON each, on Jul 3 2026 and Aug 8 2026, roughly five weeks apart and growing. Fifth, the listed treasury company's next quarterly report, the first that could show open-market buying rather than staking-reward growth. And beyond the window, the single largest scheduled event on this chain: the Feb 21 2027 thaw of the frozen 1,081.39M TON early-miner reserve.

Summary

The MrNasdog Pressure Framework reads TON at +2.79% net supply growth over the last 90 days and +2.83% projected for the next 90, on a circulating base of 2.76B TON and a total supply of 5.24B. The mechanism is a fixed block subsidy on an uncapped chain that keeps getting faster, so issuance rises with performance rather than with policy — 50.67M TON minted in 90 days against a fee burn of just 0.14M, with another 16.14M stepping out of a voluntary lock and 10.56M out of the Telegram-linked treasury cluster. The key risk is that none of this is capped or scheduled down: the only brake is a proposed reward cut that has not been enacted, and the largest single event ahead is the Feb 21 2027 unfreezing of 1,081.39M TON of dormant early-mining supply, more than a third of today's float.

---

MrNasdog Pressure Framework analysis of TON, Metric 1 — Inflation. Data + explanation only. Not financial advice. Updated Aug 25 2026.