Conflux is a two-space L1, and CFX is the single balance sheet behind both

Conflux’s core product decision is architectural, not financial. It runs one network with two execution “spaces”: Core Space (native Conflux transactions, sponsorship, storage collateral) and eSpace (EVM-compatible). That dual-space design is explicitly part of the protocol positioning.

The consensus story is also blended. Conflux describes a hybrid PoW + PoS design where PoW miners produce blocks in a Tree-Graph structure, and PoS provides finality.

In that setup, CFX is not a “governance add-on.” It is the native currency used for (1) consensus rewards, (2) transaction fees, and (3) DAO voting. Conflux’s own glossary frames it that way, and also pins the base unit at 1 CFX = 10^18 Drip.

As a mechanism, that means CFX sits in the middle of everything that matters: security spend (issuance to miners and stakers), user demand (fees and collateral), and policy (parameter votes). It also means you cannot evaluate “burn” in isolation. You have to look at net issuance through the full system.

Supply and genesis allocations: the model starts at 5B, then moves into parameterized issuance

Conflux’s economics documentation states that the genesis block included 5,000,000,000 pre-mined CFX and provides a category breakdown with release timelines.

That genesis distribution is only the first half of the token story. The operational phase introduces ongoing issuance (PoW rewards and PoS interest) and burning mechanisms, with key levers controlled by DAO voting.

For a reality check on where supply sits today, reported supplies on CoinGecko show circulating supply 5,159,697,336 and total supply 5,732,659,499 for CFX, and the page shows “Last updated on 05 Mar 2026.”

Two takeaways. One, CFX is no longer “the 5B token.” Two, any burn narrative has to fight a moving target that is minted by design.

Issuance is governance-controlled: PoW block rewards + PoS interest rate are not fixed constants

Conflux’s docs describe two primary issuance streams in the operational phase:

1) PoW issuance. New CFX is minted for miners, with the reward amount controlled by the powBaseReward parameter set via DAO votes. The docs also state the chain is mined every 0.5 seconds (2 blocks per second) in their issuance math example.

2) PoS issuance. Validators earn interest based on staked CFX. Conflux documents an APY model where APY = x × interestRate, with x depending on the relationship between total circulation and total staked amount, and interestRate set by DAO votes.

The important part is not the formula. It is the control surface. These are mutable monetary policy parameters, not a capped emission curve.

Conflux’s forum announcements show how real that mutability is. In Round 14, Conflux announced a parameter increase that moved PoW reward from 1.00 to 1.60 CFX/block, and interest rate from 4.08% to 6.50%, effective April 12, 2025 (as posted).

Then Round 16 reversed course. Conflux announced PoW reward decreased from 1.60 to 0.80 CFX/block and interest rate decreased from 6.50% to 3.26%, taking effect on August 10, 2025 (timing determined by PoW block height).

Using Conflux’s own “2 blocks per second” cadence, 0.80 CFX/block implies PoW issuance of roughly 50.5M CFX/year from base rewards alone (0.8 × 2 × 60 × 60 × 24 × 365), before considering PoS interest and any other minting flows. The inputs for that calculation are explicitly documented (block cadence and the DAO-set base reward), even if the outcome is just arithmetic.

This is the point where “burn” usually gets oversold. You can burn a meaningful amount and still inflate meaningfully if governance decides to pay security spend with fresh issuance.

Fees, collateral, and burns: Conflux burns, but the burn is structurally partial and often indirect

Conflux has three distinct “value capture” channels that people tend to blur together: transaction fees, storage collateral, and explicit burn conversions. They do not behave the same way.

Transaction fees and partial base-fee burn (post v2.4). Conflux introduced a base fee mechanism via CIP-1559 and CIP-137 with the v2.4 hardfork. The v2.4 docs show mainnet CIP activation was targeted for August 13, 2024 (estimated, tied to block and epoch thresholds).

Unlike Ethereum’s “burn the whole base fee,” Conflux explicitly shares base fees with miners and burns only a portion. The base fee model also states the miner share ratio is determined by on-chain DAO vote.

Forum voting results show this moved. Round 16 states the base fee sharing ratio was 63% and unchanged in that vote.

Mechanically, when the “sharing ratio” rises, less base fee is burned. Under a 63% sharing ratio, the implied burned portion is 37% (the remainder after the miner share), assuming the two-part split described in Conflux’s fee model.

Storage collateral (Core Space): mostly lock-and-refund, not burn. Core Space uses a storage collateral model where occupying state requires locking CFX that is later returned when storage is freed or overwritten. The docs specify 1/16 CFX locked per 64B storage entry, which corresponds to 1 CFX per 1KB.

This is not a burn. It is a temporary (sometimes long-lived) liquidity sink. It matters for circulating supply, but it is not “deflation.”

There is also an explicit interest flow on pledged storage collateral. The transaction fee docs state pledged CFX for storage generates 4% annual interest paid to miners, and the pledged CFX is returned when storage is released or modified by others.

From a burn-skeptic lens, this is a subtle inflation pressure point. Even if base-fee burn exists, the system also has explicit interest payments tied to resource usage.

Storage points conversion (post v2.3): burn triggered by sponsorship. Conflux activated CIP-107 in the v2.3.0 hardfork. The forum announced v2.3.0 completion on September 11, 2023 and explicitly lists CIP-107 among activated proposals.

The v2.3 hardfork docs describe CIP-107 as introducing “storage points” via a burning mechanism: when a sponsor adds storage collateral for a contract, a proportion of the CFX is burned and storage points are minted. Storage points are non-transferable and do not generate storage interest.

This is closer to a real burn, but it is conditional. It happens when sponsorship is used for collateral. If most applications do not sponsor storage, or if the sponsored amounts are low, this burn channel is not a reliable “offset” against issuance.

Also note the split-brain nature of the system: the storage collateral mechanism applies only to Core Space, not eSpace.

Governance and parameter control: CFX voting is a direct monetary policy tool

Conflux governance is not limited to “upgrade proposals.” It explicitly controls economic parameters. The economics docs name the internal contract ParamsControl as the manager of the DAO vote for key parameters, and list governed parameters including powBaseReward, interestRate, storagePointProp, and baseFeeShareProp.

Voting power is tied to locking CFX. Conflux’s vote-locking rules specify voting rights scale with lock duration in quarters: quarters × tokens × 0.25. They also spell out examples like 1 CFX yielding 0.25 votes beyond a quarter, 0.5 votes beyond half a year, and 1 vote beyond a year.

The lock is not cosmetic. The same governance page states users cannot withdraw tokens or decrease the locking duration once locked, while still retaining staking interest.

At the contract level, Conflux documents a built-in Staking internal contract. It includes a minimum deposit of 1 CFX for staking balance and describes vote locking via voteLock(amount, unlock_block_number), emphasizing that “once the promise has been made, there is no way to cancel it.”

Conflux also formalized “parameter voting without hardfork” in CIP-94. The CIP-94 forum post states parameters are stored in global state, voting uses on-chain voting power (staking + locking CFX through the Staking contract), and votes start periodically every 2 months with limited options to increase, decrease, or keep parameters unchanged.

From an economic modeling standpoint, this governance design creates structural uncertainty. If reward levers can be tuned every couple of months, then long-run issuance is a political variable. Burns do not become irrelevant, but they become secondary unless they are similarly rule-bound and usage-linked.

Risk analysis: burn optics vs net issuance reality

Conflux is explicit that it “employs several token burning mechanisms” to mitigate inflation, including base fee burning and sponsored storage point conversion, and that these levers are governed by DAO votes.

That is the right direction. The weak point is sustainability. Two of the major burn channels are either (a) partial by design (base fee split with miners), or (b) conditional on sponsorship behavior (CIP-107 storage points).

When Conflux itself perceived elevated inflation pressure after parameter increases, the response included discretionary action. A Conflux Foundation governance post states 76,000,000 CFX from the Ecosystem Fund would be burned, and 500,000,000 CFX would be staked to reduce PoS APR, framing the goal as covering estimated issuance during the higher-reward window from April 12, 2025 to August 10, 2025.

That kind of intervention can help. It does not convert burns into a durable value driver. It converts them into a treasury policy choice. If you’re benchmarking similar tradeoffs across networks, our CHZ tokenomics review is a useful comparison point.

Top 3 risks

  1. Dominant risk: monetary policy drift overwhelms burn. Trigger: recurring DAO rounds change reward and fee-sharing parameters (for example, PoW reward and interest rate moved materially between Round 14 and Round 16). Mechanism: higher powBaseReward and/or higher interestRate increases gross issuance quickly, while base-fee burn remains partial and storage-point burn is conditional; net issuance rises and holders get diluted unless usage-driven fee burn scales faster. Who bears it: passive CFX holders first (supply expansion), then application builders if volatility in incentives distorts security and user acquisition costs. Measurable indicators: powBaseReward and interestRate vote outcomes and effective dates; base fee sharing ratio (burn share); on-chain supply growth vs burned amount; and the frequency of parameter changes (CIP-94 cadence). We often track adjacent metrics in our research reports.
  2. Burn depends on fee throughput and sponsorship patterns. Trigger: low fee pressure (cheap blockspace) or widespread sponsorship that shifts who pays fees and collateral, reducing organic “user-paid” fee burn that would otherwise scale with activity. Mechanism: base fee burning only applies to a portion of the base fee, and storage-point burn is only triggered when sponsors provide storage collateral under CIP-107; if those flows are small, burn becomes mostly narrative. Who bears it: holders, because burn fails to offset issuance; miners/stakers, because incentives may need to be raised to maintain security if fee revenue stays thin. Measurable indicators: burntGasFee totals, baseFeeShareProp outcomes, sponsored collateral volumes, and the ratio of sponsored vs non-sponsored interactions in Core Space receipts (gasCoveredBySponsor / storageCoveredBySponsor fields exist in receipts as shown in sponsorship tooling).
  3. Two-space fragmentation weakens fee-based value capture. Trigger: activity concentrates in eSpace while Core Space mechanisms (like storage collateral economics) remain underused, or vice versa, producing uneven token sinks and inconsistent fee markets. Mechanism: different execution environments create different user behaviors and different “CFX sinks.” Storage collateral does not exist in eSpace, which removes one class of lock-based demand from the EVM side. Who bears it: dApp teams (liquidity and UX fragmentation) and holders (weaker, less predictable fee burn and collateral lock demand). Measurable indicators: share of transactions and fees by space, cross-space transfer volume, and any sustained divergence in fee dynamics between Core Space and eSpace (base fee behavior, priority fees, sponsorship prevalence).

Dominant risk (expanded): policy variability makes CFX harder to underwrite than “burnable fee tokens”

Conflux has all the ingredients people like to market: CIP-1559 branding, burning, staking, governance. The uncomfortable detail is that the same governance system that can raise burn parameters can also raise issuance parameters. And it demonstrably does. Round 14 raised both PoW rewards and the PoS base interest rate, effective April 12, 2025.

When issuance is a knob, burns become a balancing act. You can see Conflux framing it that way in the DAO #14 burn + stake plan, which explicitly ties the burn size (76M CFX) to an estimate of incremental issuance during a higher inflation window.

That is a coherent governance approach. It is not the same thing as “fee burn creates durable value.” It is closer to fiscal policy, where a large holder (the Ecosystem Fund / Foundation) can choose to offset inflation optics when it becomes politically salient.

Even ignoring discretionary burns, Conflux’s protocol-level burn is structurally limited. Base fee is split with miners by design, and the share ratio is itself a governance target. The network started with a 50% miner share in the v2.4 fee model. Governance later pushed that sharing ratio to 63% (per Round 14 and maintained in Round 16), which mechanically reduces the burned share of base fees.

That is rational if the network wants to pay miners more directly via fees. It also undermines a simplistic “burn = scarcity” thesis. The burn exists, but governance is literally empowered to route more of the fee stream away from burn and toward miners.

The practical underwriting stance is straightforward: treat CFX as a token with (1) ongoing security issuance, (2) partial fee burn that is parameterized, and (3) additional conditional burns tied to sponsorship and storage-point conversion. None of those pieces guarantee net deflation, even in high activity regimes. If you want a primer on modeling assumptions and outputs, our tokenomics FAQ covers common questions.

If you are evaluating CFX for treasury policy, or doing tokenomics consulting around a Conflux-native protocol, the main work is scenario analysis: governance parameter shifts, staking participation changes, and fee throughput assumptions. Burns can be included, but only as a function of observable fee generation and sponsor behavior, not as a default constant.



This article is part of our Tokenomics Deep Dive series.