Skip to content
Crypto.world

The Ledger

Researchers Model How Capital Costs Delay Bitcoin Mining Gaps

A paper models how miners’ hardware costs affect idle periods that may arise when transaction fees are too low to justify mining. Its Bitcoin scenarios estimate that, with an operating-to-capital-cost ratio of 1.5, a gap begins in 2056 under baseline assumptions and lasts 2.4 minutes by 2070.

Researchers Model How Capital Costs Delay Bitcoin Mining Gaps

“Capital cost caps the equilibrium hashpower and prevents the mining gap from occupying the entire round.”

Hasan Selçuk Kılıç and Nail Akar, authors of the paper

A paper posted to arXiv on October 15 models how the cost of owning mining hardware can shape Bitcoin’s future mining gaps. The gaps are periods after a block when miners may leave machines idle until enough high-fee transactions accumulate to make mining worthwhile. Authors Hasan Selçuk Kılıç and Nail Akar argue that hardware costs can limit the hashpower entering the network, keeping those idle periods shorter than they would be if miners faced only operating costs.

Why miners may wait after a block

Bitcoin miners earn block subsidies and transaction fees, while paying operating costs such as electricity and capital costs associated with owning equipment. After a block clears the highest-fee transactions from the mempool, the next block may initially offer too little fee revenue to cover miners’ operating costs. In the paper’s model, miners wait until the available reward rate reaches that threshold.

The authors track how high-fee transactions carry over from one block round to the next, rather than treating each round’s fee backlog as starting at zero. They represent the market with two fee tiers and use a Lindley-type recursion to model the backlog and derive a stationary distribution for the gap. Under Bitcoin’s difficulty adjustment, their analysis finds that expected gap length rises with aggregate hashpower: as hashpower grows, the puzzle becomes harder and miners need more fees to justify starting.

Free entry changes the long-run result

Kılıç and Akar distinguish between a fixed-hashpower case and a free-entry case, where miners can join or leave as costs and revenue change. With free entry, they estimate an operating-to-capital-cost ratio of 1.5 and find that a gap appears in 2056 in their baseline scenario, where the bitcoin price, fees and costs stay at their 2026 levels while the block subsidy continues to halve. In that scenario, the modeled gap reaches 2.4 minutes of a ten-minute round by 2070.

The timing changes with the assumed fee path. If fees rise by 10% each year in both U.S. dollars and bitcoin, while the bitcoin price stays fixed, the model puts the gap’s start in 2044 at the same estimated cost ratio. If fees fall in bitcoin terms as the price rises, it predicts no gap through 2070 under that ratio. The authors stress that these paths are illustrations, not price, fee or mining-technology forecasts.

Cost can shorten gaps but reduce hashpower

The model’s trade-off is not that higher capital costs improve every measure of security. Higher costs of owning equipment discourage entry, which can prevent a gap or make one shorter; they also lower equilibrium hashpower and the amount of honest mining work per block. By contrast, a higher block subsidy both shortens the gap and supports higher equilibrium hashpower in the paper’s analysis.

That distinction matters because a gap is not simply a period of lower electricity use. The paper says an attacker who keeps mining while honest miners wait can gain ground; in the free-entry baseline, its modeled threshold for eventually outpacing the honest chain is 43% of total hashpower by 2070. The paper does not claim that such an attack is imminent or that a mining gap currently exists: its Bitcoin illustration predicts no gap in 2026.

Bitcoin estimates carry modeling limits

For its input estimates, the study uses six Bitcoin difficulty-adjustment periods spanning July 11 to October 3, 2026. Those blocks had a 3.125 BTC subsidy, average fees of 0.0211 BTC per block and 98% average block fullness. The authors say their two-tier fee model approximates a mempool with many fee levels, so estimates inherit uncertainty from that simplification.

The long-run calculations also assume common operating and capital costs across miners, while actual operators face different equipment, energy and financing costs. The authors leave heterogeneous costs for future research and note that free entry describes a long-run equilibrium: miners with already-purchased machines may adjust more slowly after a subsidy cut. Under the paper’s baseline assumptions, its specific result is a 2056 gap onset and a 2.4-minute expected gap by 2070—not a forecast for Bitcoin’s actual mining market.

Sources

  1. Original source · Preprint Capital Costs Protect the Network: Mining Gap Dynamics in Proof-of-Work Blockchains Under a Stochastic Mempool arXiv:2610.22451

Related stories

Explore

More articles