Proof-of-Work Mining Economics and Hashprice

A bitcoin mine looks nothing like a trading desk — it's a warehouse full of specialized computers burning electricity 24 hours a day, racing every ten minutes to guess a number. But strip away the cryptography and it is a plain commodity-production business: buy an input (electricity, converted through machines into computing power), produce an output (a chance at newly minted bitcoin plus transaction fees), and survive on the spread between the two. Whether that spread is positive is decided by one number the industry watches obsessively: hashprice.

Think of a gold mine where every miner on earth is digging in the same shared deposit, and every ten minutes, one lucky miner — chosen with probability exactly proportional to how much digging power they bring — gets to keep the gold found in that window. Bring more shovels (or better shovels) relative to everyone else, and your odds of winning that round rise proportionally. That's proof-of-work mining. The "digging power" is called hashrate, measured in hashes per second — guesses at the puzzle the network poses every block — and your machine's revenue depends entirely on its hashrate as a fraction of the entire network's hashrate.

The mechanics, piece by piece

A miner's expected daily revenue depends on four things: the block reward $B$ (bitcoin paid to whoever finds a block, currently 3.125 BTC after the 2024 halving), the bitcoin price $P$, the miner's own hashrate $h$, and the total network hashrate $H$.

$$\text{daily revenue} \approx \frac{h}{H} \times \frac{144 \times B \times P + F}{1}$$

In words: the network finds roughly 144 blocks a day (one every ten minutes on average). Each block pays out the block reward $B$ in bitcoin, worth $B \times P$ dollars, plus $F$, the transaction fees bundled into that block. A miner's share of that daily total is simply their hashrate $h$ as a fraction of everyone's hashrate $H$ — because the puzzle is memoryless and winning is proportional to computing power contributed, not to cleverness or luck beyond that ratio.

Hashprice collapses this into a single per-unit number:

$$\text{hashprice} = \frac{\text{network daily revenue}}{H}$$

In words: total dollars paid out to all miners today, divided by the total hashrate competing for it — the revenue any single terahash of computing power earned today, regardless of who owns it. A miner's actual daily dollar revenue is then just hashprice multiplied by their own hashrate.

Profitability is what's left after paying for power:

$$\text{daily profit per TH} = \text{hashprice} - (\text{watts per TH} \times \text{dollars per kWh} \times 24 / 1000)$$

In words: hashprice earned per terahash, minus the electricity cost of running that terahash for a day — machine efficiency (watts consumed per terahash) times the price paid per kilowatt-hour times 24 hours, converted to kilowatt-hours.

Worked example: is a mining rig profitable today?

Bitcoin trades at $60,000. The network finds 144 blocks today, each paying a 3.125 BTC reward plus roughly 0.05 BTC of average fees, so total daily miner revenue is:

$$144 \times (3.125 + 0.05) \times 60{,}000$$

Step by step: $3.125 + 0.05 = 3.175$ BTC per block; $144 \times 3.175 = 457.2$ BTC network-wide today; $457.2 \times 60{,}000 = 27{,}432{,}000$ dollars paid out to all miners combined. If total network hashrate is 600 exahash/s (600,000,000 terahash/s), hashprice is:

$$\frac{27{,}432{,}000}{600{,}000{,}000} = 0.0457 \text{ dollars per TH per day}$$

A modern miner runs at about 20 joules per terahash, i.e., 20 watts of power per terahash of hashrate. At $0.06 per kWh, running one terahash for 24 hours costs $20 \times 24 / 1000 \times 0.06 = 0.0288$ dollars. Daily profit per terahash is $0.0457 - 0.0288 = 0.0169$ dollars, so this rig is profitable — for now.

Worked example: what a halving does to that math

Every four years, the block reward $B$ halves — a hard-coded, unchangeable feature of the protocol. Redo the same calculation the day after a halving, holding price and network hashrate fixed for the moment: block reward falls from 3.125 to 1.5625 BTC.

$$144 \times (1.5625 + 0.05) \times 60{,}000 = 144 \times 1.6125 \times 60{,}000 = 13{,}932{,}000$$

Hashprice roughly halves to about $0.0227 per TH per day. Against the same $0.0288 electricity cost, the miner is now losing $0.0061 per terahash per day. Nothing about the machine changed — only the reward per block did. This is exactly why halvings trigger waves of miner shutdowns: the least efficient rigs (older machines, expensive power contracts) go negative first and get switched off, which lowers total network hashrate, which raises everyone else's remaining share of the reward, which nudges hashprice back up. The system self-corrects through exactly the miners who can least afford to keep running.

halving — reward cut in half stable margin recovers as weak miners exit

What this means in practice

Hashprice is the mining industry's equivalent of a spot commodity price, and mining companies hedge it the way an oil producer hedges crude: through futures, options, and long-term power purchase agreements. Public mining companies (Marathon, Riot, CleanSpark, and others) are effectively leveraged bets on bitcoin price and on their own electricity costs relative to the industry average — a miner with a 3 cent per kWh power contract survives halvings that bankrupt a competitor paying 8 cents, even though both hold identical machines. Analysts use hashprice trends to forecast network hashrate itself: falling hashprice predicts capitulation and hashrate decline a few weeks later, which in turn is watched as a rough proxy for miner financial stress and potential forced bitcoin selling to cover costs.

Key terms

• Hashrate — computing power devoted to solving the mining puzzle, in hashes per second.
• Hashprice — dollar revenue earned per unit of hashrate per day; mining's spot commodity price.
• Block reward — newly minted bitcoin paid to the miner who finds each block, halved on a fixed schedule.
• Halving — a protocol-scheduled event roughly every four years that cuts the block reward in half.
• Difficulty adjustment — the protocol's periodic recalibration of puzzle difficulty to keep block times near ten minutes as network hashrate changes.