Ask ten people whether Bitcoin mining is an investment and you'll get ten different answers, most of them from people selling something. The honest answer is that it's not an investment in the way buying a share or an index fund is. It's closer to running a small business: you buy equipment, you pay a running cost, and you hope the output is worth more than both.
That framing changes how you should judge it. Nobody asks whether a bakery is "an investment". They ask what the oven costs, what flour costs, and how many loaves it sells. Mining deserves the same questions.
What a miner actually does
Bitcoin's network needs computers to process transactions and lock them into blocks. Machines called ASICs compete to do that work, and every ten minutes or so one of them wins a reward paid in new Bitcoin plus transaction fees. Your share of that reward is roughly your machine's computing power divided by everyone's computing power combined.
That last part is the whole story. You don't control your income. The network does, through two numbers:
- Hashrate: the total computing power pointed at Bitcoin. It rises when more machines switch on.
- Difficulty: a dial the network turns about every two weeks so that blocks keep arriving every ten minutes no matter how much hashrate there is. More hashrate, higher difficulty, smaller reward per machine.
So a machine that earned a certain amount last month will earn less this month if a lot of new hardware came online, even if the Bitcoin price didn't move.
The cost side: one number dominates
A mining machine draws power around the clock. Electricity is usually the largest ongoing cost by a wide margin; one industry write-up puts it at 60 to 80 percent of operating expenses for a typical operation.
Because of that, the price you pay per kilowatt-hour (kWh) decides almost everything. A machine that is comfortably profitable at four cents per kWh can lose money at eighteen cents, which is roughly what a U.S. household pays. This is why serious miners don't run machines at home. They put them in purpose-built facilities, often in places with cheap hydro or gas power, and pay a hosting rate instead.
To give a sense of the scale: a cost model published by TechBullion in November 2026 worked through the numbers for one current-generation machine (an Antminer S23e, drawing about 8.65 kW). In that model, every one cent change in the power rate moved the margin by roughly $62 a month for that single machine. Over a multi-year contract, that one cent is thousands of dollars. The same piece lists hosting rates it found across providers, from OneMiners' 3.6 cents per kWh on a seven-year prepaid plan up to around eight cents elsewhere. Those are the article's figures, not mine, and rates change; the point is the range, which is wide enough to be the difference between profit and loss.
Why a price rally doesn't rescue a bad setup
Here's the part most newcomers miss. When Bitcoin's price jumps, mining looks more profitable, so operators switch on idle machines and buy new ones. Hashrate rises, difficulty follows, and the extra reward gets spread across more competitors. The StreetInsider piece above gives an illustrative example: a 30 percent price rise with an 18 percent hashrate rise might improve a miner's revenue per unit of computing power by only 8 to 12 percent.
Miners track this with a number called hashprice: how much revenue one unit of computing power earns per day. It fell 17.5 percent in October 2026 to about $33 per petahash per day, according to Hashrate Index figures cited in the TechBullion model. When hashprice falls, high-cost miners lose money first. Low-cost miners keep going. That's the whole competitive game.
There's one more scheduled shock. Roughly every four years, the Bitcoin reward per block halves. The April 2024 halving cut it from 6.25 to 3.125 BTC. The next is expected around 2028. Any mining plan that ignores the halving is a plan for the first half of the contract only.
So how should you think about it?
Treat it like the bakery:
- Machine cost. What does the hardware cost, and how old is its design? Newer machines do more work per watt.
- Power rate. What will you pay per kWh, and is that rate fixed in writing, or can it rise with notice?
- Payback period. At today's hashprice, how many months until the machine has earned back its price? Then ask what happens if hashprice falls by a third.
- Exit. What does it cost to stop? Some hosting contracts charge fees to leave or to store an idle machine.
If you'd rather own Bitcoin without any of this, buying it directly is simpler, and the next post covers how holding and mining compare. If the business side appeals to you, the one habit that separates people who do well from people who don't is treating electricity price as the decision, not an afterthought.
As with any crypto exposure, this is volatile and can lose money. The SEC's investor education site has a plain overview of crypto asset risks worth reading before you spend anything.
Sources
- BTC Is Rallying, But Are Bitcoin Miners Actually Making More Money?, StreetInsider (press release), September 2026
- OneMiners' 3.6¢ Power Nearly Doubles Bitcoin Mining Profit, TechBullion (partner content), November 2026
- Crypto Asset-Related Investments, Investor.gov (U.S. SEC)
Disclosure: the author works with OneMiners, a crypto mining hosting provider. This isn't a paid placement, and nothing here is a recommendation to buy hardware or hosting.