Solana's Burn Headline Estimates Are a Model, Not a Mechanism

The headline estimates in the burn debate are models, not mechanisms. At the 0.1-lamports-per-cost-unit calibration and current network usage, the SIMD-547 discussion's estimate sits closer to 1,500 to 1,800 SOL per day. Separate coverage puts the daily burn projection at 10,800 to 64,800 SOL. That is still a meaningful jump from today's roughly 648 SOL in daily burns. It is also about 3% of the roughly 60,000 SOL the network issues daily through staking inflation. So at current usage levels the burn, even if it passes, is not flipping Solana into deflation; the high-calibration 64,800 SOL figure would exceed daily issuance only if the network sustains that usage and fee level. It's a patch on a very wide leak.

The stranger part is not that the estimates are being overstated in the headlines. It's that the entire debate is really about who gets to profit when a chain gets busy, and the burn mechanism is just the accounting interface where that fight is happening.

Here is the simplest true story first. Solana charges a flat per-signature base fee of roughly 5,000 lamports (lamports are the smallest subdivision of SOL; 1 SOL equals 1 billion lamports). Half of that fee is burned, the other half goes to the validator who processes the transaction. During busy periods, users also pay "priority fees" on top - a bid to get their transaction included sooner - and those go entirely to validators. The result is that validators capture the upside of network congestion, while SOL holders get a tiny, flat burn that barely registers against the roughly 60,000 SOL the network mints each day through its inflation schedule. The burn is the part that sounds deflationary on a podcast. In practice, it is about as impactful as a rounding error.

A proposal called SIMD-547, submitted in May 2026 by a developer known as cavemanloverboy from the infrastructure company Temporal, tries to fix this by replacing the flat base fee with a resource-based one. Instead of paying the same fee whether you are doing a simple balance transfer or a complex DeFi swap, you'd pay based on the computational resources your transaction consumes - compute units, loaded data, write locks, and so on. And instead of splitting the fee 50/50 between burn and validator, the entire resource fee would be burned.

The proposal's author explicitly ruled out just cranking up the base fee across the board. That would be too blunt and too politically difficult. A resource-based fee is more surgical: light transactions (market maker oracle updates, simple transfers) see only a small increase. Heavy transactions (complex swaps, NFT mints, interactions loading dozens of accounts) see a larger one. The idea is that you tax what is actually expensive for the network to process, not what is expensive for a marketing deck.

The 1,500 to 1,800 SOL per day figure is the SIMD-547 discussion estimate at current activity levels and the 0.1-lamports-per-cost-unit calibration. The larger 10,800 to 64,800 SOL per day projection sits at the upper end of the proposed calibration range - enough to make the token deflationary if the network sustains that throughput and fee level. Those are real numbers in the models. They just require network usage and fee calibration that haven't happened yet.

Anatoly Yakovenko, Solana Labs co-founder, signaled support with a "+1." Raj Gokal, the other co-founder, has also been vocal about wanting larger burns. That gives the proposal institutional weight. But weight is not approval. The earlier attempt to change Solana's monetary policy - SIMD-0228, which would have doubled the rate at which inflation declines - was rejected in March 2025, with only 37.8% of validator stake voting in favor, well short of the 66.67% supermajority threshold. A nearly identical disinflation proposal, SIMD-550, was re-submitted in June 2026 and is also still in limbo. That track record matters because it shows that when SOL holders and validators disagree about supply, validators often protect their revenue.

Here is the incentive conflict in dialogue form.

SOL holder: I want more of the fee burned so the token is less inflationary and my stake isn't diluted as fast.

Validator: I need the fee revenue to pay for servers, bandwidth, and vote transactions. If you redirect more to burn, I need higher staking rewards or higher priority fees to cover the gap.

The system designer: The whole cluster bears the cost of executing a transaction, not just the leader who includes it, so burning the fee benefits everyone holding SOL rather than enriching one validator. The leader already gets priority fees, which are the real money during congestion. Adding the resource fee on top of that as validator income would just give the leader another lever to fill blocks with whatever pays the most.

That last point is actually the strongest argument for 100% burn. It's not just about scarcity. It's about not stacking another incentive on top of the priority fee that already rewards the leader for packing blocks. If the resource fee also went to the validator, you'd have two fee levers pulling in the same direction - more traffic, more reward, more incentive to favor the loudest bidder, regardless of whether the traffic is useful or wasteful. Burning it keeps the mechanism neutral.

The problem is that "neutral" means validators get less, and validators hold the votes. This is basically the Ethereum EIP-1559 story, but reversed. Ethereum's burn works because the base fee is high enough that it matters and because block space is scarce enough that burning it creates real value capture. Solana is built on the opposite promise: cheap fees, high throughput, applications that need thousands of transactions to function. You can't burn your way to deflation when the fee per transaction is a fraction of a cent. The burn needs enormous volume to move the needle, and even then, the needle only moves a little.

There's also a second proposal running on the parallel track. SIMD-550, submitted by an engineer at Helius, would double the disinflation rate from 15% to 30% per year, cutting roughly $1.5 billion in future SOL emissions and reaching the terminal 1.5% inflation floor in 2.8 years instead of 5.7. That's the supply-side fix. SIMD-547 is the demand-side one. Together they would meaningfully alter Solana's monetary policy. Separately, each is a partial lever that one coalition wants to pull while another coalition sits on the brakes.

The modeling from Helius on SIMD-550 shows that only 2 out of 738 validators would become unprofitable in year one under the faster disinflation schedule. That number grows to 30 by year three. Small in absolute terms, but the concern is directional: faster disinflation means lower staking yields, which means less revenue for the operators at the margin, which means more stake concentration over time. Validators remember SIMD-0228. They may not rush to vote themselves into a revenue cut.

The burn framing in the headlines is not wrong about the models. It's wrong about what they mean. At current usage levels, that estimate is roughly a 3% offset against daily issuance - a dent, not a reversal. Reaching the upper end of the range assumes a calibration and sustained usage that would put the burn above daily issuance - exactly the part nobody in the headlines is talking about. And all of it assumes the governance hurdle clears. The real question is whether the validators who control the vote will approve a mechanism that reduces their fee revenue in exchange for a scarcity narrative that benefits token holders.

That's the structural implication. The burn proposal is dressed up as a tokenomics improvement, and it is that. But it is also a transfer question: who gets paid when the chain is busy. Token holders want the fee burned because it reduces dilution. Validators want the fee because it covers operating costs. The proposal's author has a point that burning keeps the system from creating yet another incentive for leaders to game their block contents. But arguments about neutrality don't show up on a validator's cost spreadsheet.

The simplest model is this: Solana issues roughly 60,000 SOL per day through inflation. The current-usage estimate for SIMD-547 burns roughly 1,500 to 1,800 of it - a 3% offset. The 64,800 SOL per day figure is a high-calibration scenario that would exceed daily issuance only if the network sustains that usage and fee level. Combined with a faster disinflation schedule, the net supply pressure eases. But at current usage SOL doesn't become deflationary. It becomes less inflationary. The headline numbers are real inside the models, conditional on calibration and usage that may not sustain. The machine is a fee-redistribution question wearing a deflation costume.

The classification boundary that will decide the outcome is not the code - the mechanism is straightforward. It's the governance vote. And the precedent is that when Solana asks validators to sacrifice revenue for a long-term token holder narrative, the answer so far has been no.