The claim
A common argument says that because it costs money to produce a bitcoin, the price cannot stay below that cost for long. If it did, miners would lose money, switch off, and the price would recover. This is sometimes called a production-cost floor. It sounds like ordinary economics, and in some commodity markets cost does influence prices over long stretches. In cryptocurrency, though, the argument has several weaknesses, which becomes clear once the pieces of mining cost are laid out.
What a miner pays for
Electricity. Usually the largest running cost. It depends on the price per unit of energy, which varies a great deal between locations and contracts, and on how efficient the machines are.
Hardware. Specialised machines, called ASICs, are bought up front and lose value as newer, more efficient models arrive. Accounting for this means spreading the purchase price over the machine’s useful life, which is itself uncertain.
Hosting and facilities. Buildings, cooling, networking and maintenance. Some miners pay a hosting company a fee instead of running their own site.
Staff, financing and other overheads. Interest on loans, insurance, management and, for listed companies, the cost of being public.
Pool fees. Many miners pay a small percentage to a pool for smoothing their payouts, as described in the mining pools article.
Taxes and regulatory costs also vary by location. All these add up to a cost per coin that differs from miner to miner.
How break-even is estimated
Analysts often compute a rough break-even price. The basic idea is simple, and an illustration with made-up numbers shows it:
- Suppose a machine produces, on average, 0.0001 coin per day at the current network conditions.
- Suppose it costs 5 units of currency per day to run and to write off its purchase price.
- Break-even is then 5 divided by 0.0001, which is 50,000 units per coin.
These numbers are only illustrative. The point is that the result depends entirely on the machine’s share of total hashing, the electricity price and the write-off period. A miner with cheap power could have a break-even far below that of one paying more, so there is no single production cost, only a spread of costs across the industry.
Published “cost to mine one coin” figures usually take an average or an assumed electricity price. They are estimates, and the different methods covered in the energy use article apply here too.
Why the floor idea is weak
Cost is a range, not a number. The marginal miner, the one just breaking even, is not identifiable from outside. Efficient miners can stay profitable at prices where inefficient ones shut down.
Price drives cost, as much as the reverse. When the price rises, more machines are switched on and competition pushes up the network’s total hashing. Because of the way the difficulty adjustment works, higher hashing makes each coin harder to produce, so cost per coin tends to follow price upwards. The causality is loose in both directions.
Miners can keep running below break-even. A miner that has already paid for its hardware may keep operating as long as revenue covers electricity, since the equipment cost is already sunk. Others may hold coins hoping prices recover or finance themselves with debt. So prices can sit under average cost for long periods.
Coins are not consumed in the same way as commodities. Demand for a commodity is partly driven by its use. Demand for a cryptocurrency depends on beliefs, adoption and money flows, as the supply and demand article discusses. Production cost has no direct hold on any of that.
Existing supply dwarfs new supply. Miners add a small fraction to the total each year, set by the issuance schedule. Price in the short term is largely set by trades in the existing coins, not by what it costs to make new ones.
Cost changes with technology and halvings. Each halving cuts the reward in half in coin terms, which raises the cost per coin overnight if everything else is unchanged. More efficient machines push in the opposite direction. A floor that moves this much is not much of a floor. The halving mechanics article covers the schedule itself.
What cost does tell you
Cost is still useful context. It shows how the industry’s economics are shaped, which miners are under stress when revenue drops, and how sensitive the network’s hash rate may be to changes in price. When prices fall, the least efficient operators tend to be squeezed first, leading to shutdowns, consolidation and sometimes coin sales, a topic examined in miner selling. Falling margins can also coincide with changes in hash rate, though reading the hash rate as a health indicator has its own caveats.
It is not a prediction tool. History includes periods where the market price sat above and below the usual cost estimates, and no one can say reliably which will hold in future.
A sensible way to use these figures
- Treat a stated production cost as an estimate with wide error bars.
- Check the assumptions about electricity price, hardware and date.
- Be wary of any claim that a price level is “guaranteed” by cost.
- Remember that these arguments are narratives; they do not remove the price volatility of the asset.
Cryptocurrency prices can fall below any level people consider a floor, and people can lose everything they invest. This article is educational and is not financial, tax or legal advice.