The most interesting number in Solana’s governance conversation right now is not a price. It’s 648.
That’s how many SOL are currently burned per day from base fees. It feels small next to Solana’s multi-billion-dollar ecosystem, and small next to the 8% annual issuance that still enters circulation. But if SGP-0003 reaches its terminal fee model, that daily burn could jump to 7,500–9,000 SOL. A 1,057% to 1,289% increase in deflationary pressure is the kind of number that makes markets pay attention.
And yet, after reading the full technical breakdown and walking through the token economics, I keep coming back to a different conclusion. The burn is not the real story. The real story is who gets paid, who stops getting paid, and whether Solana’s validator community is ready for the quiet redistribution that both proposals would trigger.
Connect first, transact second. Always.
Context: Two Proposals, One Direction
For anyone who hasn’t been following Solana governance in late August 2025: two proposals have crossed the staking threshold to enter the formal voting pipeline. SGP-0002 is an inflation-schedule adjustment. SGP-0003 is a fee-market overhaul. They are separate, but they move in the same direction: less new SOL entering circulation, more SOL being destroyed, and a network that increasingly prices blockspace by the actual resources a transaction consumes.
SGP-0002 is the more conservative of the two. It accelerates the dis-inflation rate from 15% per year to 30% per year. That doesn’t mean the inflation rate is cut in half overnight. It means the speed at which Solana’s issuance decays toward its 1.5% terminal target doubles. Under the current schedule, reaching 1.5% would take roughly 5.7 more years. Under SGP-0002, Solana gets there in about 2.8 years. Over six years, the proposal reduces cumulative issuance by 18.9 million SOL — roughly 3.15 million SOL per year. Nothing about validator rewards changes immediately. The path just gets steeper.
SGP-0003 is the more complex shift. Solana currently charges a fixed fee per signature — 5,000 lamports — plus an optional priority fee that goes to the block leader. SGP-0003 replaces that with a resource-based model: a base fee of 2,500 lamports still goes to the leader, but the network would also charge a per-compute-unit fee, with a stepped schedule of 0.1, 0.25, and 0.5 lamports per CU. The resource fees are burned. The priority fee still goes to the leader.
This is not just a smaller fee. It’s a different philosophy. The current model treats every signature as if it costs the same. SGP-0003 treats a simple transfer as cheap and a complex liquidation as expensive. That is closer to a utility bill than a postage stamp.
Core: The Technical Shift That Nobody Is Stress-Testing Enough
Let me start with what I think is genuinely smart about SGP-0003. It aligns with the industry’s slow move away from one-size-fits-all pricing. Ethereum’s EIP-1559 prices block space dynamically, but it’s still an aggregate fee for a block. Parallel EVM chains have tried execution-fee tiers. SGP-0003 is more granular: it attempts to price the actual compute unit, which is the closest thing Solana has to a universal measure of resource consumption.
That is an elegant idea in theory. Solana’s runtime can process tens of thousands of transactions per second in parallel, and a compute unit is supposed to measure how much of that capacity a transaction uses. In a world where blockspace is abundant but compute is not, charging per CU is more honest than charging per signature.
The problem is that honesty is hard to measure.
Based on my experience working with protocol teams during DeFi Summer, I’ve learned to be suspicious of fee models that assume perfect metering. I ran education workshops for Aave’s Latin American beta launch, and the biggest source of user errors wasn’t smart contract risk. It was a misunderstanding of cost. Users assumed a transaction would cost what it cost last time. When it didn’t, they panicked. That memory stays with me every time I see a proposal that makes fees more variable.
SGP-0003’s per-CU pricing requires the Solana runtime to track resource consumption at a much finer level of detail than it does today. Solana doesn’t execute transactions sequentially like Ethereum’s EVM. It uses multi-threaded scheduling, parallel execution, and global state access patterns that make precise per-instruction metering significantly more complex. The proposal’s stepped fee schedule implicitly assumes that every operation’s CU cost can be accurately measured and fairly billed. That’s not a consensus-layer change, and it doesn’t change Solana’s security assumptions. But it is a high-complexity engineering problem, and I haven’t seen enough independent peer review of that assumption.
Let me be clear: I’m not saying it can’t be built. I’m saying that the market is likely to price the burn narrative before it prices the implementation risk.
And the burn narrative is substantial. At current base-fee levels, Solana burns about 648 SOL per day, or roughly 237,000 SOL per year. That’s about 0.6% of current annual issuance. Under SGP-0003’s initial rates, daily burns would rise to 1,500–1,800 SOL. Under terminal rates, daily burns of 7,500–9,000 SOL imply annual destruction of 2.7 to 3.3 million SOL. Combined with SGP-0002’s roughly 3.15 million SOL per year in reduced issuance, the two proposals create a double deflationary shock. Net supply growth could fall from roughly 7.4% today to near zero — and possibly to net deflation.
That is the headline. But the quieter part of the model is the one that matters more.
The Quiet Redistribution: From Validators and Stakers to Holders
SGP-0002 and SGP-0003 don’t just change the supply schedule. They change the reward curve.
Under a 68% staking participation assumption, the nominal staking yield would fall from about 5.84% today to 4.34% after one year, 3% after two years, and 2.25% after three years. Lower issuance means fewer new SOL for stakers. Higher burns mean the remaining supply grows more slowly. For pure holders, that’s a benefit. For stakers and validators, it’s a slow squeeze.
The model predicts unprofitable validators would rise from 290 to 320 over three years. That’s only a 10% increase, and the authors frame it as manageable. But my concern is the soft damage that doesn’t show up in the count. When validator margins compress, operators don’t always exit loudly. They cut RPC support. They skip updates for a few extra days. They stop responding to governance discussions. The network doesn’t break on day one. It just becomes slightly less resilient, and nobody notices until the next incident.
I’ve mediated enough DAO conflicts after the 2022 collapse to know that governance decisions are never just parameter changes. They are emotional contracts. The proposal says “the reward mechanism remains unchanged,” but the yield curve tells a different story. The market will reprice SOL as an asset that rewards holders more than stakers. That’s a legitimate design choice, but it should be made consciously.
Connect first, transact second. Always.
Contrarian: The Deflation Narrative Might Be the Least Interesting Part
Here’s the angle that most market commentary will miss: the burn is not the point. The point is that Solana is moving from a “high-inflation, high-reward” model to a “low-inflation, value-appreciation” model. That is a transfer of value from the people who secure the network to the people who merely hold the token. It is also a test of whether the validator community will keep doing the hard work when the nominal rewards are thinner.
There’s also a rarely mentioned side effect: the base fee is being cut in half, from 5,000 lamports per signature to 2,500 lamports. Low-value, high-frequency transactions become cheaper. That includes spam. If the per-CU fees aren’t metered aggressively enough, SGP-0003 could actually make certain types of network abuse cheaper. The proposal assumes that high-CU transactions will be penalized, but it doesn’t fully model the MEV ecosystem’s ability to adapt. Arbitrageurs and liquidators will optimize their CU consumption, and if they succeed, the burn will be lower than the model projects.
And then there’s the governance risk. The article didn’t disclose Solana Foundation’s official position. The Foundation holds significant voting power through staked or delegated SPL positions. If it has reservations, even a community-supported proposal can hit a wall. Meanwhile, Helius’s 16 million SOL backing and Jupiter’s 12.47 million SOL backing show that infrastructure and application leaders are aligned. But what about the top 20 validators? If they feel the fee model punishes them, they can engage in soft resistance — inefficient execution, delayed votes, or a quiet PR campaign. I’ve seen it happen.
None of this means the proposals are bad. It means the market’s “FAS” effect — FOMO after success — might push SOL higher before the vote, and then reality will set in during implementation. The market has likely priced 40% to 60% of the outcome already. I’d expect ±15% to 25% volatility around the formal vote, not a straight-line rally.
Takeaway: Watch the Validators, Not Just the Burn
The vote is not the ending. The implementation is the beginning. If SGP-0003’s resource metering turns out to be as precise as the model assumes, Solana will have built something Ethereum hasn’t: a fee market that charges by the compute unit rather than by the block. If it doesn’t, the burn numbers will look great on the dashboard while the real cost lands on validator health.
So the question I keep asking is not “will SOL go up?” It’s this: after the inflation curve gets steeper and the fee model gets smarter, will the people who run the network still feel like they’re part of the project? Because decentralization was never about the token price. It was about who has the power to stay in the game.
Connect first, transact second. Always. And this time, transact with your validators — before they transact with their feet.
