Mining profitability is a short calculation with inputs that will not sit still. Revenue comes from the network, not from anyone's advertised return: a machine earns its share of the block rewards the network pays out each day, and that share is set by its hashrate against everyone else's. Costs are electricity and the pool fee. Everything else, including the price of the machine, only matters once you know whether the machine earns more than it burns.
This guide works through the calculation by hand with stated assumptions, so that the figures the profitability calculator produces stop being a black box. It then covers the three summary figures that matter, break-even electricity price, payback period and hashprice, and finishes with the reason every forecast expires: difficulty growth.
What decides how much bitcoin a machine earns?
A machine's daily bitcoin revenue is its fraction of the network's hashrate multiplied by the bitcoin the network pays out in a day. The network pays about 144 blocks a day, each carrying a 3.125 BTC subsidy plus transaction fees, so the subsidy alone is around 450 BTC a day. A machine with one millionth of the network's hashrate expects one millionth of that, or 0.00045 BTC, before fees and before the pool's cut.
The same result can be reached from difficulty rather than hashrate, which is how the calculator does it because difficulty is a fixed number for two weeks while hashrate is an estimate. The formula MiningPool uses, set out in full on the methodology page, is:
coins per day = (hashrate × 86,400 × block reward) ÷ (difficulty × 2^32)Hashrate is the machine's, in hashes per second; 86,400 is the number of seconds in a day; block reward is the subsidy plus average fees; and difficulty × 2^32 is the expected number of hashes needed to find one block. Either route gives the same answer to within the noise in the hashrate estimate.
A worked example with stated assumptions
The following example uses fixed assumptions so the arithmetic can be checked. Live values differ and are shown on the mining hub; substitute them and the method is unchanged.
What is break-even electricity price?
Break-even electricity price is the highest tariff at which a machine's daily revenue still covers its daily power cost, and it is the single most useful figure in mining because it compares a machine against your electricity bill without needing to know the future. Divide daily revenue by daily kilowatt-hours: in the example, $11.70 ÷ 84.24 kWh = $0.139 per kWh before pool fees, or $0.136 after a 2% fee.
Read it against your own tariff. A miner paying $0.08 has a margin of about $0.056 per kWh, a little over 40% of revenue. A miner paying $0.15 loses money on every kilowatt-hour and no change in hardware cost can fix that. Break-even is also how the miner database ranks machines, because it depends only on efficiency and the network, not on a purchase price that MiningPool cannot verify.
| Machine | Hashrate | Power | Efficiency | Revenue/day | Break-even |
|---|---|---|---|---|---|
| Antminer S21 XP Hyd | 473 TH/s | 5,676 W | 12.0 J/TH | $23.65 | $0.174/kWh |
| Antminer S21 Pro | 234 TH/s | 3,510 W | 15.0 J/TH | $11.70 | $0.139/kWh |
| Whatsminer M60S | 186 TH/s | 3,441 W | 18.5 J/TH | $9.30 | $0.113/kWh |
Efficiency in joules per terahash is what separates the rows. Revenue scales with hashrate, cost scales with power, so break-even is proportional to one over J/TH. A machine at 12 J/TH tolerates a tariff 25% higher than one at 15 J/TH, whatever the price of bitcoin does.
How is payback period calculated, and when is it meaningless?
Payback period is the machine's purchase and setup cost divided by its daily operating profit, giving the number of days until cumulative profit repays the outlay. A machine costing $4,000 and earning $4.73 a day after electricity and fees pays back in about 846 days, or 2.3 years, if every input stays where it is. None of them will.
Payback is undefined when the machine runs at a loss. Dividing $4,000 by a negative daily profit produces a negative number of days, which is not a shorter payback but no payback at all. MiningPool's calculators hide the figure in that case rather than display it. Payback also ignores the machine's resale value, its declining efficiency relative to newer models, and downtime, all of which lengthen the real answer.
Why does difficulty growth make every forecast expire?
Difficulty adjusts every 2,016 blocks so that blocks keep arriving about ten minutes apart, which means that when more hashrate joins the network, each existing machine's share of rewards shrinks at the next adjustment. Since 2009 the adjustment has moved up far more often than down, because more efficient machines keep being built and plugged in. A profitability figure calculated at today's difficulty is therefore the best the machine will do at today's price; the following adjustments will lower it.
The compounding is what catches people out. A 3% rise per adjustment sounds small. There are about 26 adjustments a year, and 1.03 to the power of 26 is roughly 2.2, so difficulty a little more than doubles and revenue per terahash roughly halves at a constant price. The example machine's $11.70 a day becomes about $5.40, below its $6.74 electricity cost, and the 2.3-year payback never completes. Whether that happens depends on how much new hashrate actually arrives, and on whether price rises offset it; the point is that the forecast has a shelf life of one adjustment.
Three inputs move in the miner's favor and should be modeled too. Transaction fees add to the subsidy and spike when the mempool fills. Hashrate falls when older machines switch off after a price drop, which raises the survivors' share. And a machine's own efficiency is fixed, so a rise in the bitcoin price flows through to profit in full while electricity cost stays flat.
How should a miner use these figures?
Use break-even electricity price to decide whether a machine can run at all, use daily operating profit to decide whether it is worth the effort, and treat payback as a scenario rather than a promise. The calculator takes a machine, a tariff, a pool fee and an optional purchase cost, applies live difficulty and price, and returns all three, with the source and age of every figure shown on the page.
Every figure on this site is an estimate of present conditions, not a forecast, and nothing here is investment advice. The arithmetic is the part that does not change.