Galaxy is not a traditional miner in the pure sense; it is a financial services firm that built a data center arm. But its economics rhyme with the dilemma facing every mining company sitting on powered land, substations, and cooling infrastructure that AI developers now covet.
On-Chain Evidence: The Mining Squeeze Is Real
The reason miners are pivoting at all is visible in the network data. According to The Block’s on-chain metrics, daily bitcoin miner revenue on a seven-day average stood at roughly 39.1 million USD in late October 2026, with the block subsidy contributing about 38.9 million USD and transaction fees only about 190,000 USD. Fees covered well under one percent of total miner income.
- Roughly 1.06 billion USD in subsidy revenue was paid to miners in October, versus only about 5.9 million USD in fees, per The Block’s monthly figures.
- About 1.04 USD per TH/s per month was the going rate for hashrate in the latest monthly reading, thin margins for operators with older hardware.
- Bitcoin traded near 80,000 USD this week, pressured by broader market weakness.
That fee drought explains the urgency. When income is almost entirely the fixed subsidy, every miner’s revenue depends on holding share of a shrinking real reward, and the survivors are the ones with the cheapest power or a second income stream.
The Core Conflict: Two Businesses, One Balance Sheet
The tension in Galaxy’s report is the same one facing competitors that have rebranded as AI-plus-mining hybrids. AI leases are long-term contracts with investment-grade tenants, which lenders and equity investors like. But the transition years are awkward: capital gets diverted from buying mining machines to building halls for other people’s chips, and near-term earnings suffer. Galaxy’s small net loss shows the crossover point has not arrived yet, even with 79 million USD of revenue flowing in.
For pure-play miners, the math is even harsher. Hosted AI compute monetizes existing powered capacity without buying new machines, but it competes with hyperscale builders who can outspend anyone in crypto. The miners’ advantage is speed: their sites already have electricity Interconnection, the single scarcest ingredient in AI infrastructure.
Market Implications for Your Portfolio
For investors, the takeaway is to separate narrative from cash flow. Mining stocks often trade on pivot announcements, but Galaxy’s results suggest the revenue is real while the profits lag by quarters or years. If you hold mining-sector positions, watch three numbers in upcoming earnings: lease revenue growth, the date the AI segment turns profitable, and whether the legacy mining business is still self-funding. A pivot that requires selling bitcoin reserves or issuing shares to survive the transition is a warning sign, not a growth story.
The Verdict
Galaxy’s 79 million USD quarter is genuine progress for the diversification thesis, and the loss is the predictable cost of building first and earning later. The mining industry’s future likely belongs to firms that can run both machines and leases under one roof. For now, treat every AI-pivot headline with the same discipline Galaxy’s own numbers demand: revenue is step one, profitability is the race.
The cryptocurrency market remains highly volatile. This article is for informational purposes only and does not constitute financial advice.
Galaxy Digital’s push into artificial intelligence data centers generated 79 million USD in lease revenue in its latest quarterly report, but the segment still posted a small net loss, underscoring how expensive the pivot away from pure crypto mining has become.
By Michael Nguyen | October 29, 2026
If you follow bitcoin mining stocks, this earnings line matters far beyond one company. Miners spent years chained to a single income stream: newly minted bitcoin. As rewards shrink and competition rises, the whole industry is hunting for a second business, and AI compute leases are the leading candidate. Galaxy’s numbers are one of the first honest scorecards on whether that pivot pays off yet.
The Hook: 79 Million USD In, Still Underwater
On October 28, 2026, Galaxy Digital reported that its AI data center business brought in 79 million USD in lease revenue during the quarter. The segment nonetheless recorded a small net loss, a reminder that revenue from long-horizon infrastructure does not instantly translate into profit. Building, equipping, and powering facilities for AI tenants requires heavy upfront capital, and depreciation plus financing costs sit heavily on young projects.
Galaxy is not a traditional miner in the pure sense; it is a financial services firm that built a data center arm. But its economics rhyme with the dilemma facing every mining company sitting on powered land, substations, and cooling infrastructure that AI developers now covet.
On-Chain Evidence: The Mining Squeeze Is Real
The reason miners are pivoting at all is visible in the network data. According to The Block’s on-chain metrics, daily bitcoin miner revenue on a seven-day average stood at roughly 39.1 million USD in late October 2026, with the block subsidy contributing about 38.9 million USD and transaction fees only about 190,000 USD. Fees covered well under one percent of total miner income.
- Roughly 1.06 billion USD in subsidy revenue was paid to miners in October, versus only about 5.9 million USD in fees, per The Block’s monthly figures.
- About 1.04 USD per TH/s per month was the going rate for hashrate in the latest monthly reading, thin margins for operators with older hardware.
- Bitcoin traded near 80,000 USD this week, pressured by broader market weakness.
That fee drought explains the urgency. When income is almost entirely the fixed subsidy, every miner’s revenue depends on holding share of a shrinking real reward, and the survivors are the ones with the cheapest power or a second income stream.
The Core Conflict: Two Businesses, One Balance Sheet
The tension in Galaxy’s report is the same one facing competitors that have rebranded as AI-plus-mining hybrids. AI leases are long-term contracts with investment-grade tenants, which lenders and equity investors like. But the transition years are awkward: capital gets diverted from buying mining machines to building halls for other people’s chips, and near-term earnings suffer. Galaxy’s small net loss shows the crossover point has not arrived yet, even with 79 million USD of revenue flowing in.
For pure-play miners, the math is even harsher. Hosted AI compute monetizes existing powered capacity without buying new machines, but it competes with hyperscale builders who can outspend anyone in crypto. The miners’ advantage is speed: their sites already have electricity Interconnection, the single scarcest ingredient in AI infrastructure.
Market Implications for Your Portfolio
For investors, the takeaway is to separate narrative from cash flow. Mining stocks often trade on pivot announcements, but Galaxy’s results suggest the revenue is real while the profits lag by quarters or years. If you hold mining-sector positions, watch three numbers in upcoming earnings: lease revenue growth, the date the AI segment turns profitable, and whether the legacy mining business is still self-funding. A pivot that requires selling bitcoin reserves or issuing shares to survive the transition is a warning sign, not a growth story.
The Verdict
Galaxy’s 79 million USD quarter is genuine progress for the diversification thesis, and the loss is the predictable cost of building first and earning later. The mining industry’s future likely belongs to firms that can run both machines and leases under one roof. For now, treat every AI-pivot headline with the same discipline Galaxy’s own numbers demand: revenue is step one, profitability is the race.
The cryptocurrency market remains highly volatile. This article is for informational purposes only and does not constitute financial advice.
79M in lease revenue and still underwater means colocation and h100 fitout costs run way heavier than the market models assumed. pivot is real but slow
galaxy has the balance sheet to wait this out. the small miners doing the same pivot dont. watch capex per megawatt next quarter
everyone models revenue per megawatt, nobody models the fitout burn. colocation capex is the silent killer in every one of these pivots
fitout plus power procurement is where every pivot model falls apart. revenue per mw is a slide, all in capex per commissioned mw is the actual number
fully commissioned capacity is the whole ballgame. everyone front runs the construction curve and then quotes the run rate like its already here
79M in lease revenue against a still red quarter. the pivot is directionally right but the market prices it like the margins already exist, depreciation says otherwise
79M in lease revenue and still a net loss tells you everything about the capex on these builds. depreciation eats young facilities alive
79M revenue and still red, and that is before the h100 leases reprice. the pivot math only works if lease rates hold
the h100 repricing cliff is the part nobody prices. those leases were signed in a shortage, the next renewal cycle tells us the real margin
miners pivoting to ai compute keep proving the same thing, the second business costs as much as the first. galaxy at least has the balance sheet to wait it out
second business costing as much as the first is right. difference is ai leases do not halve every four years like the block subsidy
early innings for the pivot. the quarter that matters is when lease revenue stacks on fully commissioned capacity instead of construction stage numbers