Bitcoin hashrate continues to decline as miners are unable to turn their machines back on due to price spike

Bitcoin miners entered early 2026 with a familiar but increasingly unforgiving setup. Network hashrate has fallen from its highs in late 2025, difficulty has adjusted with latency, and power costs remain the hard constraint that determines which fleets stay online and which fleets go dark.

As a result, while the market appears resilient on the surface, especially when Bitcoin rallies, it remains vulnerable at the margins, where a single difficulty increase or regional power surge can quickly turn “operation” into “shrinkage.”

Hashrate is on a downward trend after the high in late 2025

Bitcoin’s network hashrate has declined from its peak pace in late 2025 and has not consistently returned to that level, even during periods of spot strength.

JP Morgan reported that Bitcoin’s average monthly network hash rate increased by 5% in October. 1,082EH/secrecord monthly average in that series. Estimated following November 1,074EH/secrather than a continuous continuation, it is a gradual month-on-month decline.

Since late December, daily forecast values ​​have been unstable, with statistical values ​​swinging between high and low values. 1,000EH/sec This threshold is consistent with miners cycling their uptime rather than expanding smoothly.

YCharts’ network series, sourced from Blockchain.com, showed both measurements below 1,000 EH/s and rebounds above that level before and after the mid-January rebound.

The hash price, not just the Bitcoin price, influences the decision to shut down.

A miner’s actions depend on more than just the spot of Bitcoin. hash pricethe expected daily revenue earned per unit of hashrate. This is a metric that determines whether the least efficient rigs can operate without draining capital.

In Luxor’s weekly update dated January 12th, the USD hash price fell weekly. $40.23 to $39.53 per PH/sec/dayThis is a level described as “near or close to break-even for many miners.”

In other words, the network may remain unstable even during spot rebounds as miners’ profitability may remain compressed.

Luxor also reports Bitcoin decline 2.9% Approximately since last week $91,132 Hash prices have tightened, increasing the stress on miners whose cost base is stuck in spot BTC.

In the same update, Luxor’s 7-day simple moving average Because the hashrate has decreased 2.8% from 1,054EH/sec to 1,024EH/sec.

The situation in the second half of 2025 is important. Luxor’s research department previously recorded that the difficulty level reached its highest level after a positive adjustment of 6.31% on October 29th. 155.97T.

Hashprice then fell in November as fees and prices failed to offset the higher difficulty level, and hashrate index data showed that Hashprice fell to near-all-time lows. $36 per day per PH.

The market moved above that bottom towards early 2026, but not by much. This is why the recovery in hashrate since October has been uneven. Many carriers hover around a point where “on” and “off” are separated by a small difference in power costs.

A simple reality check at the machine level

The sensitivity becomes clearer when converting the hash price into revenue per rig and comparing it to the cost of electricity.

Bitmain lists the Antminer S19j Pro as 92 TH/s and 2,714 watts, while the S21 listing says 200 TH/s and 3,500 watts.

The table below uses the following values ​​as hash price input: $38.2 per PH/sec/dayroughly in line with Luxor’s cited six-month forward average.

For electricity, we use the U.S. Energy Information Administration’s September 2025 industry average electricity price of 9.02 cents/kWh as the delivered price benchmark. Wholesale prices may be lower (or higher) than this, but the miner’s total cost will vary depending on contracts, congestion, fees, and reduction terms.

This does not mean that all miners are unprofitable, but rather that many miners have much better power prices, demand response revenues, and operational efficiencies.

The point is marginal Miners cause churn, and at such hash price levels, a maxed-out fleet behaves more and more like a flexible load rather than an “always on” infrastructure.

The problem is delay levers that can blindside miners

Difficulty adjustment only every time 2,016 blocks (approximately every two weeks), meaning it does not react instantly to spot BTC or hashrate fluctuations.

This delay could force miners to absorb an entire epoch of weak hash price conditions before the protocol recalibrates, compressing margins during drawdowns and delaying the return to profitability that some operators expect to arrive soon.

This timing risk is why miners are blindsided by difficulties. Fleet looks viable during a BTC rally, but will only be squeezed if the difficulty increases in the next window and the expected returns per hash do not continue.

Difficulty data for early January also reportedly decreased by 1.20%. 146.4T Forecasts indicate that the January 22 correction could rise into 2026. ～148.20T.

Forward pricing suggests limited relief unless something changes.

Luxor said the futures market is pricing in the average hash price. $38.19 over the next six months. When using spot hash price $39.53this curve means that short-term relief will be limited unless one of the key factors changes, such as BTC appreciation, higher fees, less hardship, or cheaper electricity.

This new pattern is a kind of network whiplash. As hash prices are compressed, hashrate softens, difficulties lag behind changes, and miners have to feed on the economically weak until protocol-level remedies are applied.

Spot rallies like the recent rally to $97,000 can temporarily mask stress, but if the next difficulty window ends up being higher than what operators have modeled, the squeeze could quickly return.

Pressure is concentrated on electricity costs

If hash price tells miners how much the network is paying them, power determines how much real-world operators can keep.

In our Luxor summary, we converted compute revenue to implied revenue per MWh across fleet efficiency tiers.

This ladder is important because electricity prices are not set equally by region or contract type.
The International Energy Agency estimates that the average wholesale electricity price in the United States is approximately $48/MWh In the first half of 2025, the European Union average will be approximately $90/MWh.

The IEA also mentioned the EU’s 2026 electricity futures. $80/MWh.

Wholesale benchmarks do not have a one-to-one correspondence with delivered industrial rates, but they can help frame regional direction and volatility.

For miners operating in Luxor 25–38 J/TH layer, the implicit computing revenue is close $51/MWh This means that many sites could be forced to make cuts quickly if the cost of energy supplied rises, if hedging becomes unfavorable, or if local congestion and fees widen all-in prices.

Negative pricing adds another layer. It could reward flexible loads and penalize rigid sourcing.

The IEA reports that negative prices are becoming more common in Europe, and the percentage of time periods with negative prices is increasing. 8-9% It will launch in the first half of 2025 in countries such as Germany, the Netherlands and Spain.

This environment favors miners who can rapidly ramp up and down, capture demand response payments, and perform behind-the-meter power generation.

Carriers without that flexibility may face higher effective costs in tough times, even if headline wholesale prices are soft.

Texas remains an important mining jurisdiction and policy wildcard

Texas continues to be one of the most important jurisdictions to watch as grid policy and interconnection competition shape the economics of large mining loads.

texas state law senate bill 6 This would allow ERCOT to order certain large electricity users to shut down or use backup power generation in an emergency.

A report on the bill says it would apply to new bulk transportation. 75MW More connections than later December 31, 2025existing facilities are exempt.

Meanwhile, ERCOT’s load request pipeline has exceeded 230GW In 2025, 70% Reports on queues are associated with data centers.

The International Energy Agency has also warned that data centers will be the main driver of electricity demand growth through 2026.

For Bitcoin miners, this combination increases the value of existing interconnections and stable contracts, potentially making expansion significantly more difficult unless abatement terms and grid access are negotiated early.

What to watch next

• Next 1-2 difficulty epochs: Difficulty lag can either ease the squeeze (if it eases) or strengthen the squeeze (if the hash price remains flat and rises).
• Hash price stability: Luxor’s $39-$40 PH/sec/day zone is close to break-even for many miners, and the forward curve around $38 suggests little margin for error.
• Power variability: Fleets in the 25-38 J/TH tier are particularly at risk when the cost of provision approaches or exceeds the implied compute revenue per MWh, or when local-based risks stretch the all-in price.
• ERCOT Mitigation Risk: Emergency powers under SB 6 could lead to sudden, event-driven hashrate declines, independent of Bitcoin price.
• Data center competition: Continued growth in grid demand could limit miners’ access to lowest-cost capacity, further reinforcing regional disparities in profitability.

For now, the measurable baseline is Luxor’s spot hash price. $39.53 per PH/sec/dayin parallel with Bitcoin’s weekly decline to approx. $91,132 The 7-day average hashrate is 1,024EH/sec.

This combination sets a reference point as the network approaches the next difficulty window. The miner then decides again whether to run, reduce, or wait for a rebalance that arrives only after the protocol’s built-in delay.

And for JP Morgan 1,082EH/sec While October’s monthly benchmark still stands as a recent record for that series, the next important question is simple.

Can miner economics support enough sustained uptime to return to that pace, or will difficult delays and power constraints keep the network in stop-start mode, even if BTC continues to do well?