This Mining Ledger checks the base case I set out three weeks ago, explains why the network has just made mining easier, and works through what the cut is worth to a miner at a $67,000 bitcoin.
The retarget
At 05:02 UTC on Saturday, at block 975,744, bitcoin’s mining difficulty fell 9.13%, from 148.36 trillion to 134.82 trillion, according to mempool.space. It is the third-largest cut of 2026, after 11.16% on 7 Feb and 10.09% on 14 Jun, and the largest since June.
The reason is simple. The 2,016 blocks before the retarget took just over 15 days and 9 hours to mine, an average of 11 minutes each against the ten-minute target, by our calculation from the block timestamps. Fewer machines were hashing, so blocks came slowly, and the protocol has now lowered the bar to bring them back to schedule.
First, a correction to myself
In my 25 Oct Mining Ledger, with the seven-day hashrate above 1,000 exahashes per second, I called my base case “the steeper road”: rising hashrate, a falling price and harder blocks ahead. The 29 Oct retarget duly added 5.54%.
The road did not stay steep for long. The first thing I said would change my mind was block times drifting back above ten minutes after that retarget, which would make the mid-November adjustment “flat or lower”. They did, and it was lower, by a wide margin.
My third trigger, a sharp fall in hashrate after the retarget, also arrived. The second, a price recovery to $84,250, did not: bitcoin was at $67,271 on Kraken at 10:43 UTC today, about 16% below the $80,166 I used three weeks ago. Two out of three is enough to retire the label.
The mechanism, restated
Hashrate, the total computing power securing the network, is an estimate inferred from how fast blocks arrive. Difficulty is the protocol’s delayed response to it, reset every 2,016 blocks. When machines switch off, blocks slow down, and the next retarget makes each hash worth more.
Here is how the last month looks, using mempool.space’s daily estimates and our seven-day averages:
| Date | 7-day average hashrate | Difficulty in force |
|---|---|---|
| 24 Oct | 1,065.6 EH/s | 140.58T |
| 27 Oct (three-month high) | 1,081.0 EH/s | 140.58T |
| 6 Nov | 1,036.5 EH/s | 148.36T |
| 13 Nov | 929.9 EH/s | 148.36T |
| 14 Nov, after 05:02 UTC | — | 134.82T |
Source: mempool.space; CryptoWatchDesk calculation. Daily hashrate estimates are noisy; seven-day averages smooth some of that.
The seven-day average is down about 14% from its late-October peak. We cannot see from the data whose machines went dark, only that the work done fell.
What the cut is worth, worked through
Hashprice is what one petahash per second of mining power can expect to earn in a day. Over the last 144 blocks, about a day, miners earned 3.155 BTC per block on average, of which fees were under 1%, mempool.space’s reward data show.
At today’s difficulty and a $67,271 price, hashprice works out to about $31.67, by our calculation. At the old difficulty and the same price it would be $28.78, so the cut is worth about 10% to every machine still running. Three weeks ago, after the 29 Oct rise and at $80,166, the same sum gave about $34.13.
To make that concrete, take the illustrative operator I used in October. The efficiency and power price are my assumptions, not any company’s figures:
- Fleet size: 100 PH/s
- Revenue today at $31.67 per PH/s per day: about $3,167 a day
- At 20 joules per terahash, the fleet draws 2 megawatts, or 48 megawatt-hours a day; at $50 per megawatt-hour, that costs $2,400
- Gross margin after the cut: about $767 a day, or 24%
- Without the cut, at the same price: about $478 a day
The older fleet at 30 J/TH spends $3,600 a day on power. It still loses about $433 a day after the cut, against about $722 without it. The retarget softens the squeeze; it does not end it for the least efficient machines.
The business news fits the data. MARA sold 13,597 bitcoin after the third quarter to repay $750 million of loans, and Binance Pool closed on 10 Nov. Neither tells us how much hashrate left the network, but both describe an industry under pressure.
The base case
I’ll call my new base case the thinning. Hashrate has come off its October peak while the price has fallen further, and the retarget restores part of the margin for those still running.
More likely than not, the next adjustment, due at block 977,760 and estimated by mempool.space for around 03:00 UTC on 29 Nov, will be small in either direction rather than another large cut. Its current estimate of a 6.3% fall rests on less than 2% of the new period’s blocks, so I give it little weight yet.
What would change my mind. Three things. First, block times staying well above ten minutes in the coming week, which would mean switch-offs are continuing despite the cut. Second, a quick return of the seven-day hashrate above 1,000 EH/s, which would mean the drop was machines being moved or upgraded rather than retired. Third, a further fall in price to $60,000, which on my arithmetic would cut the 20 J/TH operator’s margin by almost half, to about $425 a day, and push more of the network towards the older fleet’s position.
The bottom line
A difficulty cut is the network’s own relief valve. It worked as designed this morning, and it buys miners about a tenth more revenue per machine at an unchanged price. For the price backdrop, see this morning’s First Print.
The bottom line: the steeper road I described on 25 Oct lasted one retarget, and the network has now eased the climb for the miners who stayed.
— Alasdair
Data: mempool.space, Kraken. As of 14 Nov 2026, 10:48 UTC.
This article is for information only and is not investment advice.