On-chain Data 2026: How Miner Behavior Has Changed After the Bitcoin Halving
Conclusion: After the halving, Bitcoin miners are undergoing "the largest structural transformation in history". On one hand, average daily miner revenue has grown from about $26,000 in the first halving cycle to over $40.2 million, and total network hashrate has surged more than 8,000% since 2016 – network security has not been shaken. On the other hand, in Q1 2026 miners sold more than 32,000 BTC, a new quarterly record, and several leading mining companies are exiting mining at scale to shift toward AI infrastructure. These two seemingly contradictory phenomena coexist, and that is the real picture of miner behavior right now.
Step 1: Look at the total dollar amount – Is miner revenue rising or falling?
To judge changes in miner behavior, first look at their total revenue. This is the foundation that drives all subsequent actions.
What to do:
Open CryptoQuant or Glassnode and search for the "Miner Revenue" chart.
Check the current average daily miner revenue level.
Current status:
Average daily miner revenue has climbed from about $26,000 in the first halving cycle to over $40.2 million today, an increase of more than 1,500×.
Although the block reward has dropped to 3.125 BTC, the rise in Bitcoin's price has fully offset the decline in issuance.
Since 2020, total network hashrate has grown by 394%, despite two halvings in that period.
What counts as "done": You can state the current order of magnitude of average daily miner revenue (tens of millions of dollars), rather than just saying "the block reward was cut in half".
Step 2: Watch the hashrate trend – Are miners expanding or retreating?
Hashrate is the most direct behavior indicator. Whether it rises or falls directly reflects whether the miner group is investing more or pulling back.
What to do:
Check the "7-day average hashrate" trend on CryptoQuant or Glassnode.
Focus on the direction of hashrate changes over the past six months.
Current status:
Since October 2025, Bitcoin's total network hashrate has fallen by more than 25%, one of the longest sustained drawdowns on record.
In Q1 2026, hashrate recorded its first year-over-year decline in six years, dropping about 4% and snapping a five-year streak of double-digit growth.
However, the network absorbed the shock through the difficulty adjustment mechanism, hashrate recovered to new highs in some periods, and no blocks were missed.
What counts as "done": You can differentiate between a "short-term hashrate pullback" and a "structural hashrate collapse" – the current situation is the former, because hashrate has already partially recovered.
Step 3: Monitor miner selling – Are they selling coins or holding?
This is the most direct indicator of miner confidence. The more they sell, the greater the financial pressure.
What to do:
Look at BTC outflows from miner addresses on Glassnode or CryptoQuant.
Pay attention to quarterly-level selling data.
Current status:
In Q1 2026, publicly listed mining companies (MARA, CleanSpark, Riot Platforms, Core Scientific, among others) sold more than 32,000 BTC, exceeding their total net selling for all of 2025 and setting a new all-time quarterly record.
By comparison, during the Terra-Luna crisis in 2022, public miners sold only about 20,000 BTC in Q2.
During the same period, miners signed roughly $70 billion worth of AI computing contracts, with a portion of capital being reallocated into AI infrastructure.
What counts as "done": You know that the current selling scale is the largest in history, and you understand that miners are choosing "preserving cash flow" over "holding coins waiting for price increases".
Step 4: Watch mining difficulty adjustments – How the network heals itself
If a large number of miners exit after a halving, difficulty automatically declines, making it more profitable for the survivors. This mechanism is now at work.
What to do:
Check "Network Difficulty" data on Bitcoin.com or CryptoQuant.
Focus on the percentage drawdown from the peak difficulty.
Current status:
Mining difficulty has fallen by more than 20% from its all-time high, the largest drawdown since China's crackdown on mining in 2021.
In some difficulty adjustment cycles, the retreat reached 10%, helping the hashprice recover to above $30 per PH/s per day.
Galaxy Research notes that miners are entering a "capitulation phase", marking the most significant decline in difficulty from a peak in nearly five years.
What counts as "done": You understand that the decline in difficulty is the network's automatic adjustment mechanism, not a "collapse signal". It is actually compensation for the miners who remain.
Step 5: Inspect miner stress indicators – Have bottom signals been triggered?
"Miner stress" itself is not good news, but historically, peak stress often coincides with price bottoms.
What to do:
Watch the Puell Multiple: currently around 0.74, well below the historical average, meaning miner revenue is below the one-year average.
Watch the Miner Cycle Stress Composite Index: it has fallen to a new low for 2026 and is in a zone that historically signals undervaluation. Similar readings in 2018, 2020, 2022 and 2024 all appeared during periods of severe miner stress and market bottoms.
What these indicators mean:
| Indicator | Current state | Historical implication |
|---|---|---|
| Puell Multiple | ~0.74 | Miner revenue is low; historically near bottom territory |
| Miner Stress Composite Index | New low for 2026 | Coincided with market bottoms in 2018, 2020, 2022 and 2024 |
| Hashprice | ~$33/PH/s/day | Below $30, miners with older gear begin shutting down at a loss |
What counts as "done": You can give the approximate range of the Puell Multiple, and you know it sits in the "stress zone", not the "boom zone".
Step 6: Follow the miner pivot to AI – The most fundamental structural change
This is probably the most distinctive feature of the current halving cycle – miners are not simply "shutting down", they are "switching tracks".
What to do:
Track public announcements and earnings reports of leading mining companies.
Pay attention to the trend of miners "branding energy".
Current status:
Core Scientific announced it will keep only 1–2 Bitcoin mining sites by end-2026, converting the rest into data centers.
VanEck estimates that for listed miners to fully transition to AI, the short-term capital gap is about $50 billion and the long-term gap reaches $221 billion.
The industry is shifting from "miner" to "energy operator": MinerMag has already rebranded to Energy Mag, and mining firm MARA removed direct mentions of Bitcoin from its website two years ago.
The giant challenge of pivoting to AI:
A Bitcoin mining farm and an AI data center are two different things:
A mining farm needs only basic buildings and quickly adjustable ASIC machines.
An AI data center demands higher system uptime, more advanced cooling, power redundancy backup, and enterprise-level customer support.
Being able to run a mining farm does not mean you can operate an AI data center.
What counts as "done": You understand that the change in miner behavior is not only "exit", but more of a "transformation", and that transformation carries significant execution risk.
Risk Warning
Miner selling does not necessarily cause a price drop – after selling 32,000 BTC in Q1 2026, the Bitcoin network continued operating normally and the price did not crash.
The AI pivot requires massive capital; VanEck estimates that so far miners have actually delivered only 25% of the contracted computing power.
Changes in miner behavior after a halving are a cyclical phenomenon. Do not overreact to short-term data swings.
Next step: Open CryptoQuant or Glassnode and first check the Puell Multiple and BTC outflows from miner addresses. If the Puell Multiple is below 0.8 and outflows are not spiking sharply, miner stress is within manageable limits. Add these two indicators to your weekly review checklist – it is far more effective than hastily checking data only when prices drop.
