Galaxy Digital: A Blowout Quarter, a Broken Promise, and a New Word for the Risk List
Galaxy Digital: A Blowout Quarter, a Broken Promise, and a New Word for the Risk List
Galaxy Digital reported the best quarter in its history on October 27 and the market answered by taking its shares down 44% over ninety trading days. The stock sits 60% below the $42.86 peak it printed on October 21, 2025. Net income of $323 million, adjusted gross profit of $569 million, adjusted EBITDA of $455 million — and a share price that behaves like a company in retreat. That divergence is the story, and it resolves cleanly once you look at what the quarter was actually made of.
The earnings were real. They just weren't operating.
Strip the segments apart and the shape changes. The operating businesses — trading, lending, asset management, staking, infrastructure — generated roughly $150 million of adjusted gross profit, up from $86 million, and $60 million of adjusted EBITDA versus about $1 million in Q2. Good, incremental progress. The treasury and corporate segment generated $419 million of adjusted gross profit against a $42 million loss in Q2, and the CFO was explicit about why: “driven primarily by significant appreciation across our digital asset holdings and private investments during the quarter” — Anthony Philip Paquette, CFO · 2026-10-27. In other words, roughly three quarters of the headline gross profit was a mark on a balance sheet that had just been marked up by a crypto rally in which Bitcoin rose more than 40%.
That is not a criticism of the position — Galaxy's $1.7 billion of net digital assets and investments, up nearly 50% in the quarter, is a genuine asset, and the venture book (>$80 million across 11 companies, marked at 3.4x) is real. But it means the quarter's earnings power is levered to crypto market beta, not to a newly de-cyclicalized franchise. And the filed numbers emphasize the point: the last 10-Q, covering the quarter ended August 7, shows total revenue of $8.6 billion against an operating loss of $128 million and net income of negative $85 million. The business prints losses in flat markets and spectacular profits in up markets. Investors know which one they own.
Helios II: the deal that was always next quarter
The second reason for the drawdown is more specific and, frankly, more damaging. For three consecutive calls, management dangled a Helios II lease. In February, Mike Novogratz said of the uncontracted capacity that “hopefully, by the time we're on next quarter, we're talking about that with much more detail” — Michael Novogratz, Founder and CEO · 2026-04-28. In April, the framing was a multi-tenant strategy with active customer conversations. By August, the CFO had named the real gating factor: “the #1 factor for us in that today is Helios II energization schedule is currently slated for late 2028” — Anthony Paquette, CFO · 2026-08-05, plus a wave of new structures that would let lower-credit tenants be wrapped by stronger counterparties.
Then, this quarter, they walked. Novogratz was direct that it was a choice: “These are large long term agreements, and we are not going to compromise on terms just to meet an announcement date” — Michael Edward Novogratz, CEO · 2026-10-27. Chris Ferraro described the mechanics of the reversal with unusual candor about the investment grade wrap structure that had been the whole thesis a quarter earlier: “an investment grade wrap that does not actually achieve investment grade protection for both our debt and our equity, is not really investment grade wrap” — Christopher C. Ferraro, COO · 2026-10-27. So they paused, broadened the process, and are now in direct conversations with investment-grade hyperscalers who, conveniently, have begun planning 2028 power.
Read cynically, this is a company that missed its own deadline and rebranded the miss as discipline. Read generously — and the tape suggests the market is not yet generous — they discovered that forward power is genuinely scarcer than they thought and that a signed deal at summer terms would have capped the upside of an asset with 800MW of secured power and a 2028 delivery slot. Both readings are defensible; the stock is pricing the first.
We had been pursuing opportunities throughout the year... Those conversations got pretty serious going into the summer. And then a lot of stuff happened. We saw actual projects around us getting delayed. We saw regulatory headwinds not just in Texas, but across the country tightening up and therefore an expectation that available power on the forward was going to be tighter than I think we or the market really understood.
What got buried under the leasing narrative is the regulatory reality. Five Texas projects received conditional Batch Zero classifications covering ~4.2GW, but three expansion projects — Caspian, Selene, Helios III — were pushed into study load, and Caspian's baseload application was reclassified with its reconsideration dispute denied. Clarity on allocation now slips to April 2027, with December audit results as the nearer checkpoint. That is the ERCOT process risk in one paragraph, and it explains why tenants want to contract but also want the process to clear first.
The newest word in Galaxy's vocabulary: indigestion
Buried in the Q&A — and ranked as Galaxy's single most prominent new keyword this quarter — is credit market. It never appeared at this altitude before. Mike's answer to a question about financing conditions turned into a six-minute macro riff: a war in Iran holding oil above $100, a 10-year yield pushing credit yields higher, and a market repricing the entire AI infrastructure build at a structurally higher rate base.
We have indigestion in credit markets. It is probably close to peak nervousness. There is not an easy path out but I think time is gonna heal this thing.
This connects directly to the capital plan. Ferraro noted lenders will be less forgiving than equity about funding construction ahead of energization certainty — and that Galaxy does not need to tap project debt until 2027. Phase II is fully funded, Phase III is equity-funded, and the $445 million of exchangeable notes maturing in December can be repaid from the $4.9 billion of cash, stablecoins and T-bills. The $757 million of quarterly capex is why the balance sheet matters, and why the free cash flow figure stays negative — negative $43 million less SBC in the last filed quarter, with the real spending still ahead in the first half of 2027.
The global tape is already voting on this. The AI data center basket is down 5.6% over 30 days with 48 decliners against 6 advancers, and the credit market-adjacent complex — long term agreement, data center capacity, high performance compute — sits firmly in the decliners column. Galaxy is not idiosyncratically weak; it is exposed to the one theme the market has decided to discount.
What would actually change the story
Two things. First, a signed Helios II lease with a genuine investment-grade counterparty — the company says it faces CoreWeave's parent on Phase I and expects the underlying tenant mix to change over time, so investors have been taught to look through to credit quality and will demand it on Phase II. Second, evidence that the digital assets franchise can earn money without a mark. Trading volumes grew 12% against flat industry volumes, the loan book grew to ~$1.8 billion, AUM rose 30% to $5.9 billion, and assets under stake hit $4.6 billion — but the segment still printed negative EBITDA, and management attributed part of that to a one-time accounting change rather than durable operating leverage.
Meanwhile the tokenization and financial infrastructure build-out — the BNY partnership, SWEEP approaching $300 million, the Galaxy-Sharplink fund — remains a 24-to-36-month story with no revenue line attached yet. And the buyback ($200 million completed, another $200 million authorized) is a signal less about value than about the absence of a better use for cash at the margin. Mike's closing line was the most honest thing on the call: “nobody likes when their stock goes down or dislikes it as much as I do” — Michael Edward Novogratz, CEO · 2026-10-27. The stock is down because the market has stopped paying for optionality it cannot see, and the only cure is a signature on Helios II.