Overview
A 270 TH/s rig grosses $8.68 a day, down 17.2% from $10.49 in mid-October, and the line that actually matters is the break-even electricity price, which has fallen to 9.73 cents per kWh against a US average industrial rate of 9.77 cents. That means the best-selling air-cooled flagship, running at the national average industrial rate, now loses money.
Two forces pushed it there at once. Difficulty rose 5.92% to 140.58T on October 16 and a further 5.54% to a record 148.36T on October 29, compounding to 11.78%, while bitcoin fell from the $81,573 used in the calculations to $75,463 on November 6, a decline of 7.49%. Multiplying the two gives the 17.24% revenue fall, of which difficulty contributes 10.54 percentage points and price 7.49, making difficulty the larger half.
Key Takeaways
Hashprice printed $32.86 per PH/s per day at 18:00 UTC on November 2, down 8.7% week on week; backing it out from 148.36T difficulty and a $75,463 price gives roughly $32.15 today, 17.6% below the $39.05 in the original piece, 16.1% above the June 2026 low of $27.70 and 49% below the July 2025 peak near $63. Difficulty-implied network hashrate rose from 950 EH/s to 1,062 EH/s. Net profit per rig is $4.13 a day at 5 cents, or $1,509 a year; $3.26 at 6 cents; just $0.63 at 9 cents, or $232 a year; and a loss of $1.99 a day at 12 cents. Break-even bitcoin prices at all four rates moved up by exactly 11.78%, matching the difficulty increase, to $38,795, $46,554, $69,831 and $93,108. Luxor's forward curve sets the November 2026 to April 2027 hashprice average at $33.04, essentially level with spot, meaning the market is pricing no recovery at all.
Why Rig Revenue Fell Below Nine Dollars a Day
The original piece described revenue approaching ten dollars. A month later that line has been crossed, and price is not the main reason.
Take difficulty first.
Bitcoin retargets every 2,016 blocks, roughly every two weeks, and the last two retargets both went up.
October 16 brought a 5.92% rise to 140.58T, the third largest of 2026, and October 29 added 5.54% at block 973,728 to reach a record 148.36T. The two compound to 11.78%. Difficulty-implied hashrate follows directly from multiplying difficulty by 2 to the 32nd power and dividing by 600 seconds, which puts 148.36T at 1,062 EH/s against 950 EH/s for the 132.72T in the original. Measured seven-day averages ran near 1,069 EH/s on November 2 and roughly 1,034 to 1,038 EH/s on November 3, broadly consistent with the implied figure.
Rising difficulty dilutes per-rig output proportionally. The 132.72T in the original corresponded to 0.000474 BTC per PH/s per day, while 148.36T gives 0.000424, a fall of 10.54%. That relationship to the difficulty increase is exact rather than approximate.
Now price. The original calculations used $81,573 against $75,463 today, a decline of 7.49%. Multiplying the two ratios, 0.8946 times 0.9251 equals 0.8278, a revenue decline of 17.22%.
Hashprice Measures What a Unit of Hashrate Buys
Hashprice compresses both variables into one number, defined as dollar revenue per PH/s per day.
Luxor's spot reading was $32.86 at 18:00 UTC on November 2 against $35.99 a week earlier, a fall of 8.7%. Across October it dropped from $40.24 on October 1 to $33.18 on November 1, down 17.54%, with
another tracker measuring 17.27% for the month.
Backing the figure out from current difficulty and price gives $31.98 for the subsidy portion and $32.15 once the roughly 0.56% fee contribution is added. The $32.86 print on November 2 corresponded to a price then of roughly $76,800 to $77,800, so the two readings are consistent; the gap is simply the further price decline since.
The forward market is the part worth noting. Luxor's contracts set the November 2026 to April 2027 average at $33.04, and the October 30 version set the strip through March 2027 at $32.88, with implied hashrate rising from 1,056 EH/s in November to 1,100 EH/s by March 2027. Forward sits 0.55% above spot, which amounts to the market saying the next six months will be neither better nor worse, at odds with the seasonal recovery many operators expect. These forward quotes are indicative rather than executable offers.
Returning Hashrate Keeps Taking Share Back
Network hashrate bottomed at 879 EH/s in July 2026, returned to roughly 1,001 EH/s in early September and reached a seven-day average near 1,069 EH/s in early November, up 8.6% from roughly 984 EH/s in early October and still short of the October 2025 peak near 1,160 EH/s. Total miner revenue for October was $1.215 billion, the best month since November 2025, averaging roughly $39.2 million a day, though the current run rate has fallen back to roughly $35.4 million. Gross revenue per EH/s per day fell from $38,552 in the week ending October 9 to $33,095 in the week ending November 2, down 14.2%.
Hashrate returning while hashprice falls are two faces of one thing: against a fixed output of roughly 450 BTC a day, more hashrate can only take a smaller share each.
Running the Numbers on a Single Machine
The method is unchanged from the original; only the inputs are current.
The inputs are machine hashrate, machine efficiency, electricity price, bitcoin price and difficulty. Take
Bitmain's Antminer S21 XP at 270 TH/s and 13.5 J/TH, giving 270 times 13.5, or 3,645 watts, and daily consumption of 3,645 times 24 divided by 1,000, or 87.48 kWh. These three match the original because the hardware has not changed.
A Full Worked Example
On output, 0.000424 BTC per PH/s per day across 0.270 PH/s, plus the fee contribution, gives 0.0001150 BTC a day against 0.000128 in the original. At $75,463 that is $8.68 of gross revenue a day, with a 2% pool fee of $0.174.
Power costs across four rates come to $4.374 at 5 cents, $5.249 at 6 cents, $7.873 at 9 cents and $10.498 at 12 cents.
Net profit is therefore $4.13 a day at 5 cents, or $1,509 a year; $3.26 at 6 cents, or $1,190 a year; $0.63 at 9 cents, or $232 a year; and a loss of $1.99 a day at 12 cents, or $726 a year. The original figures were $5.96, $5.08, $2.46 and negative $0.16. The 9-cent tier is the one to watch, where annual profit compresses from roughly $900 to $232, so against a secondhand machine price of roughly $2,300 to $2,800 the payback period has stopped being a meaningful number.
Difficulty Math Versus the Hashprice Shortcut
The two routes check each other. The difficulty route divides machine hashrate by difficulty-implied network hashrate, then multiplies by 144 blocks a day and the 3.125 BTC subsidy. The hashprice route multiplies machine hashrate directly by current hashprice. The first gives $32.15 per PH/s per day, matching the second at the same bitcoin price, which establishes that the two cannot disagree; the only difference is whether you handle difficulty yourself.
Drawing the Two Break-Even Lines
This is the most actionable part of the piece, and both lines moved this month, though not in the same way.
Break-Even Electricity Price
Take gross revenue, subtract the pool fee and divide by daily consumption: 8.68 minus 0.174, divided by 87.48, gives $0.0973 per kWh, or 9.73 cents. The original figure was 11.75 cents.
What makes that number matter is the reference line it crossed. The
U.S. Energy Information Administration put the average industrial retail price at 9.77 cents per kWh in July 2026, up 4.7% year on year, and its
most recent Electric Power Monthly covers August 2026. A 13.5 J/TH flagship running at the national average industrial rate is therefore 0.04 cents away from losing money.
Inverting the relationship makes it clearer still: at 9.77 cents, the worst efficiency that still breaks even is 13.44 J/TH. The S21 XP is 13.5 J/TH, short by 0.06. The national average industrial rate now falls exactly on the efficiency of the best air-cooled machine available.
Break-Even Power Scales Inversely With Efficiency, Exactly
This is a relationship that can be verified precisely rather than a rule of thumb.
Efficiency | Machine | kWh per TH per day | Break-even price | Versus 13.5 J/TH |
8.9 J/TH | Antminer S23 XP Hyd 600 TH/s | 0.2136 | 14.75 cents | 1.517x |
9.45 J/TH | Sealminer A4 Pro Hydro 680 TH/s | 0.2268 | 13.89 cents | 1.429x |
13.5 J/TH | Antminer S21 XP 270 TH/s | 0.3240 | 9.73 cents | 1.000x |
14.5 J/TH | WhatsMiner M70 220 TH/s | 0.3480 | 9.05 cents | 0.931x |
18.5 J/TH | Older air-cooled class | 0.4440 | 7.10 cents | 0.730x |
The final column matches the efficiency ratios exactly: 13.5 divided by 8.9 is 1.517, and 13.5 divided by 18.5 is 0.730. Break-even electricity price varies inversely with J/TH with no approximation in it.
What that column means in practice is that a hardware upgrade can change a site's profit status independently of bitcoin's price. At 12 cents the S21 XP loses $0.00737 per TH per day while the S23 XP Hyd earns $0.00587. Same site, same hashprice, profitable purely on the swap.
Bitmain's S23 XP Hyd was announced on August 28 with first batches shipping in November 2026 and deliveries running into January 2027, while
Bitdeer's Sealminer A4 Pro Hydro offers 680 TH/s at 9.45 J/TH.
Break-Even Bitcoin Price
Divide power cost by daily bitcoin output and then by 0.98 to find the price at which revenue meets cost. The four rates give $38,795 at 5 cents, $46,554 at 6 cents, $69,831 at 9 cents and $93,108 at 12 cents.
Set those against the original figures and something worth pausing on appears.
Power rate | Original break-even | Current break-even | Change |
5 cents | $34,705 | $38,795 | Up 11.78% |
6 cents | $41,646 | $46,554 | Up 11.78% |
9 cents | $62,470 | $69,831 | Up 11.78% |
12 cents | $83,293 | $93,108 | Up 11.78% |
All four moved identically, and by exactly the difficulty increase of 11.78%. That is not coincidence but a direct result of the formula: break-even bitcoin price is set by difficulty and power cost alone and is independent of the current bitcoin price. Price fell 7.49% this month and had no effect whatsoever on these four numbers.
This gets inverted in practice constantly. Saying that a falling price makes life harder for miners is true of cash flow and false of the break-even line. The line moved up because of difficulty, and that means even if price returned to $81,573, a 9-cent miner would earn less than a month ago, because the break-even point has relocated from $62,470 to $69,831.
Why Both Lines Move Together
One numerator drives both. When hashprice falls, break-even electricity price falls proportionally while break-even bitcoin price rises proportionally. The difference is that hashprice can fall because of difficulty or because of price, while break-even bitcoin price responds only to the difficulty component. So reading a miner's position starts with splitting the hashprice move into its difficulty and price halves, then asking which line moved. This month the split was 10.54 percentage points of difficulty against 7.49 of price, so both lines moved and the difficulty half was larger.
How Subsidy, Difficulty and Fees Shape the Result
The block subsidy remains 3.125 BTC, unchanged since the April 2024 halving, across roughly 144 blocks a day for roughly 450 BTC of new supply, with the next halving expected in 2028.
Fees remain negligible and are still shrinking. They made up 0.56% of total miner revenue last week and have stayed below 1% every month since July 2025.
The Block's earlier reporting described this as long-term reliance on the block subsidy, and the 0.69% reading on August 11 sat near a ten-year low, with only April 2026 lower at 0.52%. For single-machine math, 0.56% means fees can be treated as zero with an error under a sixth of a dollar a day.
Difficulty is the only one of the three that moves materially in the short run. The next retarget is expected around November 13 at block 975,744, where
the current estimate is a decline of 2.6% to 3.2%. That is an estimate rather than a result, since the actual figure depends on block times across the remaining blocks. A 3% decline would lift gross revenue to $8.95 a day and return break-even power to 10.03 cents, back above the industrial average, but that magnitude only reverses a quarter of the two October increases.
What the Listed Miners' Cost Curve Shows
The original cited the
CoinShares Q2 2026 report and its weighted average cash cost of roughly $75,500 per bitcoin. Several companies have since reported actual third-quarter figures on bases that differ from the research houses, which is worth separating out.
Riot Platforms disclosed the most. The company mined 1,637 bitcoin against 1,406 a year earlier, with self-mining power costs of $74.8 million, or roughly $45,700 per coin, and power curtailment credits of $24.7 million, or roughly $15,100 per coin, offsetting 33% of the power bill. The company reported an all-in cost of $78,805 per bitcoin including miner depreciation, and $39,265 excluding depreciation and net of curtailment credits. Third-quarter revenue was $181.4 million, split between $116.0 million of mining at 63.9%, $47.0 million of engineering at 25.9% and $18.4 million of data center at 10.1%, with net income of $4.0 million.
That $78,805 belongs next to today's $75,463. On the company's own all-in basis, the current price sits $3,342, or 4.2%, below its cost. That is why the gap between all-in cost including depreciation and cash cost excluding it is a question of survival in this environment rather than an accounting detail.
Curtailment credits are the item single-machine math cannot produce. A home rig has no grid dispatch contract and earns nothing from curtailment, while a company able to power down when the grid is tight can offset a third of its power bill. That is one of the most fundamental differences between corporate cost structure and a single rig's electricity math, and it matters more than purchasing scale.
IREN reported a different story this month. The company
published first-quarter fiscal 2027 results on November 5, with mining revenue falling from $67 million the prior quarter to $28 million, down 58.2%, AI cloud revenue rising from $71 million to $200 million, up 181.7%, total revenue rising from $137 million to $228 million, up 66.4%, net loss narrowing from $684 million to $263 million, and adjusted EBITDA rising from $19 million to $66 million at a 29% margin. Mining fell from 48.9% of revenue to 12.3%. The company pushed its target of more than $4 billion of ARR from year-end 2026 to January 2027, possibly extending to the end of February, citing a customer-requested design change and supplier timing.
Those figures show the transition the original described actually happening, but happening through absolute contraction in mining rather than growth in both lines. A single source has also reported that the company plans to shut down roughly 23.2 EH/s of mining capacity before December 31, a figure not confirmed in the earnings release. The revenue-per-MW comparison cited in the original, roughly $1.5 million a year for AI and HPC against roughly $500,000 for mining, remains the core basis for these decisions.
Risks, Scenarios and What to Watch
What the Model Leaves Out
Single-machine math excludes machine purchase cost, rack and cooling capex, hosting fees, network and staffing costs, depreciation and residual value, and downtime losses, along with how and when a pool actually settles. It also excludes the curtailment credits above, which for a company can be a third of the power bill. So 9.73 cents is a boundary at the electricity layer and not a return on investment.
Three Scenarios
Difficulty eases while price holds: the November 13 retarget delivers the estimated 3% decline, gross revenue returns to $8.95 a day and break-even power returns to 10.03 cents. This is closest to a base case, but it reverses only a quarter of October's increases.
Price recovers while difficulty keeps climbing: if price returned to $81,573 with difficulty at 148.36T, gross revenue would reach $9.38 a day while break-even bitcoin price stayed at $69,831, meaning a 9-cent miner still earns less than a month ago. The forward curve's implied 1,100 EH/s for March 2027 points to exactly this combination.
Next-generation hardware deploys at scale: an 8.9 J/TH machine lifts break-even power to 14.75 cents, which rescues a large number of high-cost sites while pushing difficulty higher and further compressing break-even power for older models. First S23 XP Hyd batches ship in November with deliveries into January 2027, so this should start appearing in difficulty during the first quarter of 2027.
The Watchlist
October CPI publishes on November 10, the difficulty retarget falls around November 13, and the Federal Reserve meets December 8 and 9 with a new dot plot. On hardware, watch actual delivery progress for the S23 XP Hyd and the Sealminer A4 Pro Hydro; on companies, watch whether IREN confirms the 23.2 EH/s shutdown and whether its ARR target lands in January 2027.
Readers wanting to set retarget dates against price reactions can mark October 16, October 29 and November 13 on the chart on the
BTC price and market data page. Spot and futures markets on
MEXC typically see short-lived shifts in liquidity and spreads around retargets and dense macro windows.
Exclusive View from the MEXC Editorial Team
The most important change this month is not that revenue fell 17% but that break-even electricity price crossed the US average industrial rate. At 9.73 cents against 9.77, a gap of 0.04 cents, a line long treated as the floor of the safe zone has been breached. Inverting it says more: the worst efficiency that breaks even at the national average industrial rate is 13.44 J/TH, and the best-selling air-cooled flagship, the S21 XP, is 13.5. The entire air-cooled generation now sits below the line, with only the new hydro models above it.
What will most likely be misread is attributing this revenue decline to price. Split out, difficulty contributed 10.54 percentage points against price's 7.49, making difficulty the larger half and the only half that affects break-even bitcoin price. Break-even prices at all four power rates moved up by exactly 11.78%, matching the difficulty increase to the decimal, while the 7.49% price decline had no effect on any of the four. That distinction is substantive for decisions: a price recovery improves cash flow but cannot move the break-even point, and only a difficulty decline relocates it. Meanwhile the forward curve sets the next six months at $33.04, level with spot, with implied hashrate rising from 1,056 EH/s to 1,100, which amounts to the market saying the point will not be relocated.
Riot's third-quarter disclosure supplies a variable single-machine math can never see. The company's $74.8 million power bill was offset 33% by $24.7 million of curtailment credits, a difference of roughly $15,100 per bitcoin mined. That is the most fundamental gap between an operator that can respond to grid dispatch and a machine plugged into a wall, and it matters more than procurement discounts or scale. At the same time an all-in cost of $78,805 per coin sits 4.2% above today's price, which explains why mining accounts for only 63.9% of its revenue while engineering and data center together reach 36%.
IREN's quarter carries that logic to its conclusion. Mining revenue fell 58.2% in a single quarter to $28 million while AI cloud rose 181.7% to $200 million, mining's share of revenue dropped from 48.9% to 12.3%, and adjusted EBITDA went from $19 million to $66 million. Notably this is not two lines growing together but one contracting absolutely while the other expands. For anyone using this calculator to assess an investment, that conclusion may matter more than the net profit at any power rate: when revenue per MW differs threefold between $1.5 million and $500,000, no amount of electricity optimisation closes that gap.
FAQ
How much does it cost to mine one bitcoin?
It depends whose basis you use. On single-machine electricity, a 13.5 J/TH rig at 9 cents produces 0.0001150 BTC a day for $7.873 of power, implying roughly $68,500 of electricity cost per coin. On a corporate basis, Riot's third quarter reported an all-in cost of $78,805 per coin including miner depreciation and $39,265 excluding depreciation and net of curtailment credits, while CoinShares put the Q2 2026 weighted average cash cost for listed miners at roughly $75,500. Those three differ enormously because they include different cost items, so any citation has to name the basis.
What is hashprice and why does it matter?
Hashprice is dollar revenue per PH/s per day, compressing difficulty and bitcoin price into one number. It printed $32.86 at 18:00 UTC on November 2, down 8.7% week on week, and backs out to roughly $32.15 at current difficulty and price, 16.1% above the June 2026 low of $27.70 and 49% below the July 2025 peak near $63. It matters because you can multiply it directly by machine hashrate to get gross revenue without handling the difficulty formula.
How do you calculate break-even electricity price?
Take gross revenue, subtract the pool fee and divide by daily consumption. A 270 TH/s machine at 13.5 J/TH draws 87.48 kWh a day, grosses $8.68 and pays $0.174 in pool fees, giving 9.73 cents per kWh. The figure varies inversely with machine efficiency with no approximation: an 8.9 J/TH machine breaks even at 14.75 cents and an 18.5 J/TH machine at 7.10 cents, and those ratios match the efficiency ratios exactly.
At what bitcoin price do miners shut down?
Across the four power rates, break-even prices are $38,795 at 5 cents, $46,554 at 6 cents, $69,831 at 9 cents and $93,108 at 12 cents. The key point is that these four are set by difficulty and power cost alone and are independent of the current bitcoin price. Price fell 7.49% this month, and every bit of the change in these four numbers came from the 11.78% difficulty increase instead.
How much does a difficulty increase reduce earnings?
Strictly proportionally. Difficulty rose from 132.72T to 148.36T, up 11.78%, and output per PH/s per day fell from 0.000474 BTC to 0.000424, down 10.54%, which are reciprocal. Break-even bitcoin prices at every power rate simultaneously rose 11.78%, matching the difficulty increase exactly. The next retarget is expected around November 13 with a current estimate of a 2.6% to 3.2% decline, which is an estimate rather than a result.
How much do transaction fees contribute to mining revenue?
They were 0.56% of total miner revenue last week and have stayed below 1% every month since July 2025, with the 0.69% reading on August 11 near a ten-year low and only April 2026 lower at 0.52%. For single-machine math they can be treated as zero with an error under a sixth of a dollar a day. Miner revenue currently comes almost entirely from the 3.125 BTC block subsidy.
Why are corporate mining costs so much higher than my electricity math?
Because the cost items are entirely different. A corporate basis includes miner depreciation, hosting, staffing, networking and site costs, while single-machine math includes only electricity. In the other direction, companies have offsets a single rig cannot access: Riot's $24.7 million of curtailment credits in the third quarter offset 33% of its power bill, or roughly $15,100 per bitcoin. So a company's all-in cost exceeds single-machine electricity while its cash cost can fall below it.
Can new hardware change profit status?
Yes, and independently of bitcoin's price. At 12 cents the S21 XP loses $0.00737 per TH per day while the 8.9 J/TH S23 XP Hyd earns $0.00587, so the same site at the same hashprice turns profitable purely on the swap. That logic does not hold for the network as a whole, however, since deploying new machines at scale raises difficulty and further compresses break-even power for older models. First S23 XP Hyd batches ship in November 2026 with deliveries into January 2027.
What does the forward market expect for the next six months?
Luxor's contracts set the November 2026 to April 2027 hashprice average at $33.04, essentially level with the $32.86 spot print on November 2, with implied hashrate rising from 1,056 EH/s in November to 1,100 EH/s by March 2027. That means the market is pricing neither a recovery nor further deterioration. These forward quotes are indicative rather than executable offers.
Disclaimer
This article is for general information only and does not constitute investment advice, financial advice, legal advice, tax advice or any trading or purchasing recommendation. Every mining calculation here rests on point-in-time assumptions for difficulty, bitcoin price and electricity cost, where difficulty retargets roughly every two weeks and the estimate for the next retarget is not its result, price movements change every dollar-denominated conclusion directly, and single-machine math excludes machine purchase, hosting, operations, depreciation and downtime, so none of it constitutes a return on investment. Corporate cost figures cited here come from company disclosures and third-party research on bases that differ from one another and are not directly comparable. Readers should conduct their own research and reach independent judgments based on their own power contracts, hardware and risk tolerance, consulting a qualified professional where appropriate, and the MEXC Crypto Pulse team accepts no liability for any direct or indirect losses arising from use of this information.
About the Author
MEXC Editorial Team provides the latest platform updates, product launches, trading feature guides, and industry insights. With a focus on developments in the global crypto market, it delivers transparent information, professional data, and forward-looking analysis. Both beginners and experienced traders can find firsthand updates, in-depth observations, and official announcements related to the MEXC ecosystem.
Research References