Bitcoin’s seven-day average hashrate slipped to about 915.8 EH/s by September 26, 2026 — its lowest in roughly three weeks — while publicly tracked miner reserves fell by 1,530 BTC over the week to around 1,192,766 BTC, according to data reported by crypto.news and TronWeekly.
Background: what hashrate and reserves measure
Hashrate is the total computing power securing the Bitcoin network, expressed in hashes per second; 1 EH/s equals one quintillion hashes per second, and 1 ZH/s equals 1,000 EH/s, so the readings above 1 ZH/s earlier in September represented a record-high band of security, crypto.news reported. Miner reserves — the Bitcoin held by mining operations, tracked here using CryptoQuant data — reflect whether miners are accumulating or distributing coins. Because block rewards are the main source of newly issued Bitcoin, miners are structurally net sellers over time, funding electricity, hardware and financing costs by selling part of what they mine. The balance between output and sales in a given week offers a rough read on operating pressure across the sector.
How far did the hashrate fall?
The 915.8 EH/s seven-day average marked a three-week low, down from readings above 950 EH/s earlier in the week and above 1 ZH/s on September 9 and 15, crypto.news reported. Hashrate — the total computing power securing the Bitcoin network — routinely fluctuates with block timing, power conditions, difficulty and how many machines are actively running, so short-term dips are common and do not necessarily indicate a lasting change. For scale, 1 EH/s equals one quintillion hashes per second, and 1 ZH/s equals 1,000 EH/s, so the readings above 1 ZH/s earlier in September represent a record-high band of computing power securing the network.
How much Bitcoin did miners sell?
Miner reserves tracked by the cited data fell 1,530 BTC in the week to September 26, 2026, leaving holdings near 1,192,766 BTC, per crypto.news and TronWeekly. For context, crypto.news noted reserves had risen 261 BTC the week of September 5, so the late-September move marked a shift back toward distribution after a brief stretch of accumulation.
Which miners were selling?
Individual company disclosures underscored the trend. CleanSpark mined 593 BTC in August 2026 but sold 821 BTC during the month, finishing with 13,703 BTC and an operating hashrate that reached 38.3 EH/s, according to crypto.news. The report also said Hyperscale Data’s Bitcoin holdings fell roughly 79%, from about 1,006 BTC in July to around 215 BTC, after it shut Michigan mining operations on September 1. The examples illustrate how treasury policies differ across operators, with some selling newly mined coins while others hold.
Why are miners reducing reserves?
Miners typically sell newly minted coins to cover electricity, hardware and financing costs, and selling can pick up when margins tighten. The combination of a softer hashrate and declining reserves suggests some operators trimmed output or liquidated holdings during the period, though the cited data does not attribute the move to a single cause. Seasonal power costs, maintenance and equipment upgrades can all weigh on output and prompt selling. Because block rewards are the primary source of newly issued Bitcoin, miners are structurally net sellers over time, and the balance between what they mine and what they sell in any given week offers a rough read on operating pressure across the sector.
What does this mean for the network?
A short-term dip in hashrate does not by itself indicate a lasting change in network security; difficulty adjusts roughly every 2,016 blocks to keep average block times near 10 minutes, so the network self-corrects as computing power ebbs and flows. Elevated miner selling can add supply to the market, but weekly reserve changes of this size are small relative to total miner holdings above 1.19 million BTC. The figures here are a snapshot from the cited trackers for the week ending September 26, 2026 and are provided for information only, not as trading advice.
Why it matters
The combined signal — a softer hashrate alongside falling reserves — is what makes the week worth noting, though each figure is modest in isolation. crypto.news reported reserves had risen 261 BTC the week of September 5, so the 1,530 BTC decline to about 1,192,766 BTC marked a shift back toward distribution after brief accumulation. Company disclosures put faces to the trend: CleanSpark mined 593 BTC in August but sold 821, while Hyperscale Data’s holdings fell about 79% after it shut Michigan operations on September 1. Still, the weekly reserve change is small against total miner holdings above 1.19 million BTC, and network security stayed robust at well over 900 EH/s even at the three-week low — context that matters for interpreting short-term dips.
What should readers watch next?
Key variables include whether hashrate recovers toward its early-September highs above 1 ZH/s, how winter power costs affect mining margins, and whether large operators continue drawing down reserves in the fourth quarter. Those data points, along with the figures above, come from the cited mining reports and may change as new weekly data is published. Hashrate and reserve trackers can also differ slightly between providers depending on how they average data, so figures are best read as directional indicators rather than precise point values. For most investors, the practical takeaway is that network security remained robust at well over 900 EH/s even at the three-week low, while miner selling stayed modest relative to the sector’s total holdings.
What to watch next
Key near-term markers include the next Bitcoin difficulty adjustment, which the protocol applies roughly every 2,016 blocks — about every two weeks — to keep average block times near 10 minutes, and which will respond to the recent dip in computing power. Weekly hashrate and miner-reserve updates from trackers such as CryptoQuant will show whether output recovers toward early-September highs above 1 ZH/s and whether large operators keep drawing down reserves into the fourth quarter. Mining companies also publish monthly production and sales reports, like the CleanSpark and Hyperscale Data figures cited above. Winter power costs are a recurring seasonal factor flagged in coverage. These are scheduled data points, not forecasts of hashrate or price.
Sources
On Google, tick the box next to stnews.live to see our reporting higher in Top Stories.
Disclaimer: The content on this page is for informational purposes only and does not constitute financial advice. Always do your own research before making investment decisions.
James Nakamoto is a Markets Reporter at STnews, tracking Bitcoin spot and futures, ETF flows, miner economics and macro liquidity. His reporting follows STnews's editorial standards: primary-sourced, cited, and non-advisory.
Conflicts of interest
I disclose any positions held at time of writing within each article. I do not trade Bitcoin futures.
