Data insight · Sep 30, 2026 · 6 min read
Bitcoin fell 53% from its October 2025 high and closed at half price or worse on only 14 days. That is the shallowest fall from a cycle high in our record, and it came with a twist for miners: income per machine, counted in coins, went up.
Bitcoin’s highest close in our series is $124,493, on 6 October 2025. Its lowest close since is $59,013, on 30 June 2026, a fall of 52.6%. Bitcoin Magazine put it at roughly 53% off the record on 24 June, as the price slid through $60,000.
That is a deep fall by any ordinary standard. By bitcoin’s own it is a shallow one. Counting from each cycle’s highest close until the price regained it, bitcoin closed at least 50% below that high on 736 days after December 2013, 732 days after December 2017 and 533 days after November 2021. After October 2025 it has done so on 14 days: 5 to 7 June, 10 June, and 24 June to 3 July. Since 3 July every close has been above the half-way line.
The two deeper lines tell the same story. Closes at least 70% below the high numbered 348, 178 and 137 days in the three earlier cycles. This time there have been none: the deepest close was 52.6% down.
Fourteen days at half price or worse. The earlier bear markets spent between 533 and 736.
From the highest close of each cycle to the lowest close that followed, bitcoin fell 84.9% from $1,136.90 on 5 December 2013 to $172.00 on 15 January 2015, 82.5% from $18,911.79 on 19 December 2017 to $3,308.11 on 14 December 2018, and 76.4% from $67,534.17 on 9 November 2021 to $15,929.85 on 22 November 2022.
The fall from October 2025 to June 2026 was 52.6%. Every earlier fall took more than three-quarters of the price. Our record starts in 2013, so this is the shallowest fall from a cycle high in the thirteen years we hold, and the words “so far” belong to it: the price is 33.0% below the high today, and nothing on the chart says June was the last low.
It was also quicker. The lowest close so far came 267 days after the high, against 406, 360 and 378 days for the earlier three lows. And it was slower to get going: bitcoin took 242 days to fall to half its high, where the earlier cycles took 14, 46 and 181.
One visible difference from the last cycle is who held the coins. The US spot ETFs did not exist in 2022. Their combined holdings peaked at 1,348,977 bitcoin on 13 October 2025 and reached a low of 1,214,457 on 8 July 2026, a reduction of 134,520 coins, or 10.0%. Bitcoin fell 52.6% from its high to its low in those months.
By 29 September the funds held 1,281,416, so about half of what they gave up has come back. The ETF holdings chart shows how flat the line stayed for a fall of this size. We cannot prove holders caused the shallow fall, and the data does not try to. It shows that the coins did not leave in a rush.
In all three earlier cycles the network kept growing through the fall. Difficulty, which follows the hashrate competing for blocks, rose 201% between the December 2017 high and the December 2018 low and 71% between the November 2021 high and the November 2022 low. Even at the bottom of a bear market, more machines were arriving.
This time it went the other way. Difficulty was 150.84 trillion on 6 October 2025 and 133.87 trillion on 30 June 2026, down 11.3%. It is the only cycle in our record in which the network was smaller at the low than at the high. The companion piece on difficulty covers where it stands after the rally; here the point is what the retreat did to income.
The pot that miners share is the block subsidy plus fees, divided by the network’s hashrate. Between the high and the low, that pot per petahash per day fell from $401.44 to $55.92 in the 2021 to 2022 cycle, a fall of 86%. In 2025 to 2026 it fell from $52.22 to $27.95, a fall of 46%. Our hashprice series is measured rather than estimated from late September 2021, so these are the two cycles we can compare.
Counted in coins, the gap is wider. In 2022 the same petahash earned 41% fewer bitcoin per day at the low than at the high, because difficulty had risen while the price fell. In 2026 it earned 13% more, because difficulty had fallen. A machine at a fixed power price produced more coins at the bottom than at the top.
The dollar figure still fell, because the coin was worth 53% less. What the bitcoin count shows is that in this bear market the price did the damage and competition did not add to it.
The lesson for a miner is narrower than “the bear market is over”, which nobody can know. The last three lows arrived 360 to 406 days after their highs; today is day 358, and the price is 33% below the record. What the record supports is a statement about how this fall behaved.
A fall of 53% that also shrinks the network is the mildest fall in our record for anyone running machines. The coins a machine produced did not fall at the low, and a miner who kept them holds the same count today in a market 41% above the 30 June close. Power is billed in dollars, so a lower price still squeezes the margin, and efficiency and the power price decide how tight. Cheap power and an efficient machine are what turn a bear market into a slow stretch rather than a threat.
The drawdown chart and the hashprice chart update every day. When the next fall comes, they show how deep it is and how long it has lasted, and whether the network is getting bigger or smaller underneath it.
In the 2022 bear market a machine mined 41% fewer coins. In this one it mined 13% more.
Firsthand Bitcoin sells and hosts mining hardware, including this machine. No manufacturer, distributor or affiliate programme paid for or reviewed this page and we take no commission on the links above. Historical figures are computed from daily bitcoin price and network hashprice, each day valued at its own prices. Nothing here is investment advice.