StarkWare CEO Eli Ben-Sasson proposed replacing Bitcoin’s fixed supply cap with a capped annual issuance rate. His core argument is that lost private keys permanently shrink the usable supply
The Bitcoin 21 million cap, the single most defended number in crypto, is under public attack from one of the most cited cryptographers in the industry. Eli Ben-Sasson, CEO of StarkWare and a co-founder of the science behind Zcash, argued on X this week that a fixed cap stops making sense once you account for coins that vanish behind lost private keys, and proposed replacing it with a maximum issuance rate of 4% per year. The post landed on July 7 and by the next morning had reopened a debate most Bitcoiners considered settled a decade ago: whether absolute scarcity is Bitcoin’s greatest strength or a slow-motion design flaw.
The logic behind the proposal is mathematical rather than ideological. Every year some fraction of Bitcoin holders dies without passing on their seed phrases, throws away a hardware wallet, or simply forgets a password, and none of those coins ever come back. Ben-Sasson pushes the observation to its limit: “as time goes to infinity, all keys will be lost.” Under a hard cap with no new issuance, the circulating supply can only move in one direction.
The data behind that claim is not marginal. Hardware wallet maker Ledger put the figure at up to 4 million permanently inaccessible coins in a November estimate, while a 2024 study by River Financial counted 3.8 million BTC sitting in addresses dormant for more than a decade. Roughly 20 million of the 21 million total have already been mined, meaning the network has issued more than 95% of everything it will ever produce. Strip out the lost coins and the effective supply may already sit closer to 16-17 million
The estimates back him up more than his critics would like. Hardware wallet maker Ledger calculated in November that up to 4 million BTC are permanently inaccessible, nearly a fifth of everything that will ever exist. Roughly 20 million of the 21 million total have already been mined, meaning the network has issued more than 95% of everything it will ever produce. Strip out the lost coins and the effective supply may already sit closer to 16-17 million.
James Howells gave the problem its most famous face. The IT worker from Newport, Wales, threw away a hard drive holding 8,000 BTC in 2013 and has spent years fighting for permission to excavate a landfill. His coins still sit on the ledger, visible to anyone, and no one will ever move them.
Bitcoin supply metricFigure (mid-2026)Hard cap21,000,000 BTCAlready mined~20,000,000 BTC (over 95%)Left to issue (until ~2140)Under 1,000,000 BTCEstimated lost (Ledger, Nov 2025)Up to 4,000,000 BTCThe Bitcoin side of X answered within hours, and the answer was not subtle. Replies framed the 21 million cap as a non-negotiable social contract and warned that touching it would effectively create a different asset. One recurring line called the number “central to the religion.”
The economic counterargument runs deeper than tribal loyalty. In the Austrian economics framing that underpins the digital gold thesis, lost coins are not a leak in the system but a transfer, because when someone’s keys vanish, the purchasing power of every remaining coin rises proportionally. Michael Saylor, executive chairman of Strategy, has taken this view to its logical extreme, saying he plans to destroy his own private keys upon his death and describing the act as a pro-rata contribution to remaining holders.
On the practical question of whether a shrinking supply could ever run out of units, Bitcoiners point to divisibility. One Bitcoin splits into 100 million satoshis, so the total cap contains 2.1 quadrillion base units, and Layer-2 systems like Lightning can account in even finer increments, which leaves more than enough units for global commerce under almost any loss scenario. Ben-Sasson’s reply to that point is that satoshis behind lost keys are just as inaccessible as whole coins, so divisibility does not touch the underlying liquidity drain.
This debate is not purely hypothetical, because other networks have already made the opposite choice. Monero abandoned a hard cap years ago in favor of a tail emission that mints 0.6 XMR per block indefinitely. Miner revenue never falls off a cliff, fees stay low, and the network avoids the four-year halving shocks that periodically squeeze Bitcoin miners. Critics, including analysts at CoinShares, counter that permanent issuance quietly erodes purchasing power and makes the long-term economics harder to defend to investors.
Zcash founder Zooko Wilcox pointed to a middle route under discussion in his own ecosystem, the Network Sustainability Mechanism. The 21 million cap stays intact, but users can voluntarily burn coins that are then reissued to miners as block rewards over a four-year period, keeping security funded without printing anything new.
ModelSupply ruleMiner funding after subsidiesBitcoin (current)Hard cap at 21M, halvings until ~2140Transaction fees onlyBen-Sasson proposalMax issuance up to 4% per yearPerpetual block rewardsMonero tail emission0.6 XMR per block, foreverFixed perpetual subsidy plus feesZcash NSM proposal21M cap kept, burned coins recycledReissued burns over four yearsStrip away the philosophy and a concrete engineering problem remains. Miners currently earn most of their revenue from newly minted coins, a subsidy that halves every four years and reaches zero around 2140, at which point transaction fees must carry the entire cost of securing the network. If fee revenue proves insufficient, hashrate drops and the cost of attacking Bitcoin falls with it. Researchers on forums such as Delving Bitcoin have been openly debating whether fees alone can sustain enough hashrate to deter a well-funded attacker, and Ben-Sasson’s rate-capped issuance is one proposed answer among several. This is the part of his post that serious Bitcoin developers engage with, even while dismissing the supply-cap change itself.
Changing the issuance schedule would require consensus among developers, miners, node operators, exchanges, and holders whose entire investment thesis rests on the fixed cap. That coalition does not exist and shows no sign of forming. If a faction ever shipped a 4% issuance client, the realistic outcome would be a contentious hard fork producing a minority chain, while the legacy 21 million network kept the ticker, the liquidity, and the brand.
The more likely direction over the next decade runs through efficiency rather than expansion. Proposals like Cluster Mempool and covenant-based upgrades aim to raise the fee yield miners can extract from existing block space, addressing the security budget without touching monetary policy. There is also an irony in the background of this whole episode: Zcash, the network built on Ben-Sasson’s own research, enforces the same 21 million hard cap it copied from Bitcoin, which means the cryptographer now questioning the number helped export it in the first place.
The post Bitcoin’s 21 Million Cap “Doesn’t Make Sense,” Says StarkWare CEO appeared first on ETHNews.
