CFX Inflation Analysis · September 2026 · Mixed flows, supply roughly steady

Conflux mints CFX twice over — a proof-of-work block subsidy for miners and proof-of-stake interest for stakers, both inside one uncapped supply — and the staking half is more than twice the mining half. Over the trailing 90 days that produced 22.58M CFX of new supply against a burn of just 7.9K CFX, for a net of +0.43%; the Pressure Framework projects +0.40% for the next 90 days, because an on-chain governance vote that settled on Aug 5 2026 cuts the Conflux block subsidy by 35.4% on Oct 3 2026. There is no vesting cliff left, no Foundation release, no buyback, and no supply cap.

The verdict, in one paragraph

Against a circulating base of 5,237.7M CFX, the Pressure Framework books 22.58M CFX of sell pressure and 7.9K CFX of buy pressure over the trailing 90 days — a net of +0.43% — and projects +0.40% for the next 90 days once the settled Conflux block-subsidy cut lands. The inflation monitor reads +0.48% for the same window, a gap of 0.05 percentage points, comfortably inside the framework's 0.5pp tolerance, so no monitor-gap warning ships on the CFX overview page. The label for CFX is a quietly and predictably inflating dual-consensus chain: Conflux has no supply ceiling and no meaningful burn, but its issuance is small, fully visible on chain, and adjustable by a vote — which is exactly what the network just used it for.

Sell pressure: where new CFX comes from

Almost all of it comes from one row. Sell #1, protocol inflation, is 22.58M CFX, and it has two legs because Conflux runs proof-of-work and proof-of-stake inside a single protocol and a single supply. The proof-of-work leg pays miners a flat 0.4013 CFX per block, and the Conflux Tree-Graph produced 15,556,477 blocks across the window at a realised interval of 0.4999 seconds — a whisker under the chain's 0.5-second target — for 6.22M CFX minted. The proof-of-stake leg is the larger one: 2,085 consecutive staking-reward distributions, every one of them read from the chain rather than sampled, paid out 16.35M CFX. Stakers therefore out-mint miners by more than two to one on Conflux, and their share is still climbing, because CFX committed to proof-of-stake rose from 847.4M to 904.9M across the same 90 days. Both legs are genuinely new coins: the chain's own issued-supply counter rose from 5,788.12M to 5,810.70M CFX, which is the independent state read the framework closes this row against, to a residual of 0.07%.

Sell #2, vesting unlocks, is 0. The Conflux genesis release schedule ran out in 2024, and the chain publishes the proof directly: its two-year and four-year locked buckets both read zero. No cliff falls inside this window or the next one. Sell #3, Foundation and unscheduled unlocks, is also 0, and this one is worth being precise about, because it was not zero at the last rebuild. The one wallet that can be proven to belong to the Conflux Foundation — the address that sent the project's announced 2025 burn — drained itself in early Jun 2026, days before this window opens. Read at both ends this session it held 0.25 CFX on Jun 9 2026 and 0.69 CFX on Sep 7 2026, with nothing staked at either end. The wallet is bare; the window slid past the event. Sell #4, long-term locked or bankruptcy, is 0 as well — CFX has no bankruptcy estate, no trustee and no court-ordered distribution attached to it.

Buy pressure: where new CFX goes

Almost nowhere. Buy #1, programmatic buyback, is 0. Conflux operates no programme that spends treasury money repurchasing CFX on the open market, and none was announced or executed inside the window. The closest thing in the project's history runs the other way: a 2025 community vote destroyed 76M CFX the Conflux Ecosystem Fund already held, which is a burn of inventory rather than a purchase of float.

Buy #2, protocol fee burn, is 7.9K CFX — real, measurable, and almost invisible next to the mint. Conflux burns in two places. A share of every transaction base fee is destroyed, which removed 2,436 CFX over the 90 days. The larger leg is storage: when a sponsored storage deposit is released, most of it is converted into storage points rather than refunded, and the CFX behind it is destroyed — 5,467 CFX this window. Against 22.58M CFX created, that is roughly one coin burned for every 2,900 minted. The framework verified this on both surfaces rather than trusting either: the Conflux null address, the one named in the project's own burn announcement, held 572,964,410.5 CFX on Jun 9 2026 and the identical figure on Sep 7 2026, so nothing was burned by transfer, while the supply counter carries the two fee burns above. Those two surfaces are independent — a transfer to the null address would not touch the supply counter, and a base-fee burn would not touch the null address — so the row is booked once, from the counter that moved.

Buy #3, Foundation buy, is 0. Conflux bought no CFX on the market; the one provable project wallet went from 0.25 to 0.69 CFX, which is dust arriving rather than demand. Buy #4, new long-term lock, is 0, and this is a deliberate judgment rather than an absence. Staking on Conflux genuinely rose across the window — proof-of-stake commitments from 847.4M to 904.9M CFX, and the wider locked pool from 887.5M to 970.8M CFX. None of it counts as buy pressure, because Conflux staking unwinds in roughly two weeks and the circulating-supply classification the framework divides by already treats staked CFX as tradable float. Booking an 83.4M CFX staking increase as a lock would have manufactured about 1.6% of deflation that no CFX holder ever experienced.

Foundation and overhang

The identified overhang on CFX is unusually small, and the reason is that Conflux has essentially no non-circulating bucket left. The chain's issued supply minus the null-address balance comes to 5,237.73M CFX, against a classified circulating supply of 5,237.66M — a spread of 0.0015%. Practically every CFX in existence is already counted as float. The first tracked item is the Conflux Foundation's executing wallet, at 0.69 CFX and readable on chain at every rebuild; it is bare, and its 2.27M CFX release in early Jun 2026 sits outside this window. The second is the Conflux Ecosystem Fund allocation behind the 2025 authorisation to place CFX into publicly listed companies' digital-asset treasuries under a lock of no less than four years — no wallet address has been published, no counterparty named, and no deal signed, so it is tracked through the project's own disclosures on a fortnightly walk rather than on chain.

One large balance is deliberately excluded. The Conflux null address holds 572,964,410.5 CFX — more than a tenth of the supply — but burned coins are gone, not held, so they are not overhang. The trigger sentence applies to both tracked items: if the Foundation executing wallet's balance falls between refreshes, or if the Ecosystem Fund allocation is disclosed and moves, that outflow enters Sell #3 at the next refresh. It is also worth naming the blind spot honestly: because the classified float already contains almost everything, a Conflux Foundation release would barely register in the circulating-supply series at all, which is precisely why the framework watches the wallets instead of the series.

How CFX compares to other uncapped, dual-consensus chains

Conflux sits in a class of roughly one. Most chains pick a consensus and mint through it: a proof-of-work chain like Bitcoin pays a block subsidy on a fixed halving clock and burns nothing, while a proof-of-stake Layer-1 pays staking rewards on an emission curve and, if it runs a base-fee burn, can swing negative in busy quarters. Conflux does both at once — a Tree-Graph proof-of-work subsidy and proof-of-stake interest, drawn from the same uncapped supply — which is why its inflation reading has two independent drivers instead of one. The mix is the interesting part: the proof-of-stake leg is 16.35M CFX against the proof-of-work leg's 6.22M, so the chain looks like a mining chain and inflates like a staking chain.

Compared with a hard-capped halving chain, CFX gives up the ceiling and gains a dial. Bitcoin's issuance cannot be changed by anyone; Conflux's can be changed by CFX holders every 60 days, and the record shows they use it — the block subsidy has come down from 2 CFX in the early rounds to 0.8, then 0.4013 in Apr 2026, and a settled vote takes it to 0.2593 on Oct 3 2026. That is a faster de-issuance schedule than any halving, but it is a policy rather than a guarantee, and the same mechanism that halves the subsidy can double it: an earlier round raised it from 1.00 to 1.60 CFX. Compared with an exchange token running quarterly buybacks and burns, CFX has the shape of a burn but not the volume — 7.9K CFX a quarter against 22.58M minted means Conflux would need its burn to grow roughly 2,900-fold before it offset issuance. On the pure numbers, though, CFX is milder than most uncapped Layer-1s: +0.43% per 90 days annualises to under 1.8%, where a staking-linked Cosmos-style emission of 5% to 15% a year is normal.

What to watch in the next 90 days

First and largest, the Oct 3 2026 activation of the round-23 Conflux parameter vote, which cuts the proof-of-work block subsidy from 0.4013 to 0.2593 CFX and takes roughly 1.58M CFX out of the forward 90-day mint on its own — that single change is the whole difference between the +0.43% trailing reading and the +0.40% forward one. Second, the round-24 vote now running on chain, which will settle around Oct 3 2026 and activate around Dec 2 2026; every topic in it currently sits at "unchanged", but the proof-of-stake interest rate is one of the four parameters on the ballot and it governs the larger half of Conflux inflation. Third, the amount of CFX committed to proof-of-stake, at 904.9M and rising, because staking rewards scale with it and are already the dominant mint. Fourth, the Conflux Ecosystem Fund treasury authorisation, still unexecuted, which would place CFX into a listed company's balance sheet under a four-year lock and would land in the ledger twice — as a Foundation release and as a new long-term lock. Fifth, the null address at 572,964,410.5 CFX: any step in that balance is a discretionary burn, and the last one was 76M CFX.

Summary

The MrNasdog Pressure Framework reads CFX at +0.43% over the trailing 90 days and +0.40% projected forward: mixed flows, supply roughly steady. The structural mechanism is dual issuance into an uncapped supply — Conflux mints 6.22M CFX for miners and 16.35M CFX for stakers per 90 days, while its base-fee and storage burns remove only 7.9K CFX, so the chain has no meaningful offset and never claims one. The key risk is that nothing here is fixed by code: Conflux has no supply cap, and the same 60-day governance dial that is about to cut the block subsidy 35.4% on Oct 3 2026 has previously been used to raise issuance. The comfort is that the numbers are small and completely legible — every coin minted, every coin burned, and every parameter vote is readable on the Conflux chain before it takes effect, which is how this reading was derived.

MrNasdog Pressure Framework analysis of CFX, Metric 1 — Inflation. Data + explanation only. Not financial advice. Updated Sep 7 2026.