Over the past six quarters, Galaxy Digital's share price has decoupled from its trading revenue in a way that should concern anyone who understands how public market valuations actually work. The TSX-listed stock now trades at a multiple that implies the market is pricing Galaxy as an AI infrastructure play, not a digital asset financial services firm. But the gap between the narrative and the physical reality is where the money is made — and lost.
Tomorrow morning, Galaxy releases its Q2 2026 financial results. The market will not be reading the income statement line by line. It will be scanning for one thing: evidence that the data center expansion — the Helios campus in Texas, the capital commitments, the power purchase agreements — is transforming from PowerPoint architecture into revenue-generating physical infrastructure. Every institutional investor I know who holds GLXY is watching for the same three data points: segmented data center revenue, CapEx guidance for fiscal 2026 and 2027, and any mention of an AI customer contract. If none of those appear, expect a repricing.
This is not a column about whether Galaxy's earnings will beat or miss consensus. It is an analysis of what the data center pivot actually means from a financial engineering perspective — the math that most commentary around this stock completely ignores. Because after 29 years watching this industry evolve from cypherpunk mailing lists to institutional balance sheets, I have learned that the most dangerous moment in any transition story is when the market stops asking how the money flows and starts believing in the magic.
CONTEXT: FROM FINANCIAL SERVICES TO PHYSICAL CAPITAL
Galaxy Digital is not a crypto company in the traditional sense. It is a publicly traded financial institution — listed on the Toronto Stock Exchange under ticker GLXY — founded by Mike Novogratz, a former Goldman Sachs partner and Fortress Investment executive, that happens to specialize exclusively in digital assets. Its revenue streams have historically been divided across trading, asset management, principal investments, and investment banking. In the crypto market structure, Galaxy operates at the institutional layer: it is a counterparty, a custodian, a fund sponsor, and an underwriter. It does not issue tokens, does not run a proof-of-stake validator network, and does not maintain a blockchain protocol. Its competitive moat has always been regulatory licenses, balance sheet strength, and Novogratz's relationship network.
That moat is now being reconstructed. In 2023, Galaxy acquired the Helios bitcoin mining facility from Argo Blockchain during the latter's distress — a massive physical campus in Dickens County, Texas, with substantial power capacity. At the time, the acquisition was framed as a measured hedge: owning mining infrastructure would allow Galaxy to participate in the Bitcoin network's security market while reducing reliance on third-party hosting providers. It was a defensive move, rational and unremarkable.
What has changed between 2023 and 2026 is not Galaxy's business model. It is the market's willingness to assign speculative premiums to anything connected to the AI infrastructure buildout. Every publicly traded company with access to power and a roof — from distressed bitcoin miners to abandoned industrial facilities — has been repriced as an AI data center candidate. This repricing reached Galaxy several quarters ago, and the stock now carries a hybrid identity: half digital asset financial services, half high-performance computing infrastructure play. The Q2 earnings report is the first serious test of whether that hybrid identity is backed by capital deployment reality or merely by narrative inertia.
To understand what is at stake, you have to understand how the market currently values Galaxy. The financial services business — trading, asset management, principal investment — is valued on a traditional bank or fund manager multiple. The infrastructure business, assuming it matures, is valued on a data center REIT or infrastructure fund multiple. Those two multiples are fundamentally different. A bank that manages other people's money trades at perhaps 8-12x earnings, with the market discounting for cyclicality in trading revenue. A data center company with contracted power and committed customers trades at 20-30x forward EBITDA, because those cash flows have visibility. Galaxy's current valuation sits somewhere between these two anchors, which means the eventual direction of the stock will be determined by which anchor wins.
The data center anchor wins only if Galaxy can demonstrate it has done something more than buy a building in Texas and call it AI infrastructure. That is the core insight the market will be testing tomorrow.
CORE: THE FINANCIAL ENGINEERING OF THE PIVOT
Let me be direct about the accounting mechanics, because this is the substance that gets lost in the coverage. A data center is a capital-intensive asset with a specific cost structure: land, building shell, power distribution, cooling systems, security, and networking infrastructure. For a company like Galaxy, which historically operated as a lightweight financial intermediary with minimal fixed assets, the shift to owning physical infrastructure changes several critical financial metrics simultaneously. It inflates capital expenditures, introduces depreciation and amortization as a significant P&L line item, adds fixed operating costs that exist regardless of revenue, and converts what was previously a variable-cost business into one with substantial operating leverage in both directions.
When I model the financial trajectory of this pivot — and I have run this model dozens of times since Galaxy announced its Helios acquisition — the pattern is consistent. In the first two to three years, free cash flow turns negative as capital is deployed into the physical plant. Depreciation, which is a non-cash charge but still reduces reported operating income, begins to depress net income even during periods of healthy revenue. The balance sheet shifts from a predominantly liquid asset profile to one where a significant percentage of total assets are tied up in property, plant, and equipment. The company's equity beta — its correlation to broader market movements — increases, because physical infrastructure companies trade differently than financial firms.
None of this is inherently bad. It is simply a different animal. And the market rewards different animals with different valuations. The question for Q2 2026 is whether Galaxy's board and management team understood the full implications of this transition when they accelerated the data center expansion — or whether they are simply responding to the same AI narrative pressure that has driven every other infrastructure-adjacent company into a capital spending race.
The CapEx Signal
The single most important number in tomorrow's earnings release will not be revenue. It will be capital expenditure. Specifically, the market needs to see the trajectory of Galaxy's cash outflow into property, plant, and equipment — and the corresponding management guidance for the next two fiscal years. If CapEx is flat and management maintains its prior range, then the data center expansion is proceeding at a measured pace, and the infrastructure narrative remains under construction. If CapEx jumps materially and management raises forward guidance, then Galaxy has committed to the infrastructure path in a way that will fundamentally reshape its financial profile for the next five years.
From my own work modeling infrastructure transitions in the crypto mining sector — I did this extensively during the 2022-2023 bear market, when I advised institutional clients on the survivability of various public miners — the CapEx signal differentiates the serious infrastructure players from the narrative players. Core Scientific, before its emergence from bankruptcy, was deploying hundreds of millions into its Texas sites. Riot Platforms was doing the same. Those were real capital commitments. The companies that talked about AI and data centers without meaningful CapEx increases were the ones whose stocks eventually reverted to their crypto trading beta.
According to the limited information available in the pre-earnings coverage, Galaxy's data center ambitions have accelerated. That word — accelerated — carries a specific meaning in the context of a public company's capital allocation. It signals that the board has authorized expenditures beyond the original Helios acquisition plan. It signals that the company is preparing to bring additional capacity online. And it signals that the market's AI-infrastructure repricing of Galaxy is not entirely unearned — management is actively feeding that narrative. But feeding a narrative and realizing its economic value are two different things. The gap between them will be visible in the cash flow statement.
The Free Cash Flow Trap
Here is the mathematical trap that infrastructure transitions create, and it is the one most analysts gloss over. During a heavy CapEx phase, free cash flow — defined as operating cash flow minus capital expenditures — turns negative or collapses to near zero. This is not, by itself, a sign of distress. It is the normal trajectory of a growth infrastructure company. But it interacts with the rest of the income statement in ways that matter for investor psychology.
Galaxy's trading business generates volatile, quarter-to-quarter revenue that depends on market conditions. Its asset management business generates recurring fee income, but that income scales with assets under management, which in turn scale with crypto market prices. Its principal investment book is marked-to-market and produces unpredictable gains and losses. None of these businesses are natural providers of the stable, contracted cash flows that infrastructure investors expect from capital-intensive assets. Therefore, when the data center consumes cash during construction, the company's overall free cash flow profile becomes doubly volatile: the volatile financial services revenue may or may not cover the steady capital drain of the infrastructure buildout.
This creates what I call the dual-engine stall scenario. If crypto markets decline, trading revenue compresses. If data center construction costs overrun or power prices rise, the infrastructure side bleeds cash faster than planned. Both engines fail simultaneously, and the company is left with negative free cash flow, rising leverage, and a deteriorating equity story.
Hedging is not fear; it is mathematical discipline. The disciplined approach to this transition would be to secure downside protection on the power purchase side, lock in multi-year contracts with anchor tenants for a portion of the data center capacity, and maintain enough balance sheet liquidity to survive a 12-18 month downturn in both crypto trading volumes and AI infrastructure demand. I will be looking at the balance sheet footnotes tomorrow for evidence of that hedging discipline — particularly any disclosed power purchase agreements with fixed pricing, any committed tenant contracts, and the maturity profile of the company's debt.
The Revenue Mix Question
Let me be specific about what I will be looking for in the segment disclosure. Galaxy's historical segments — Trading, Asset Management, Principal Investments, and Mining — will need to show something new. Whether that new segment is labeled Data Center, Infrastructure, HPC, or AI Services, its revenue contribution matters far less than what management says about its ramp trajectory.
Here is my analysis framework, derived from having audited infrastructure companies during the crypto mining consolidation wave: A data center's revenue quality is a function of three variables — committed capacity, contracted duration, and customer credit quality. A facility running at 70% utilization with a mix of short-term and long-term tenants generates lower-quality revenue than a facility running at 50% utilization with three anchor tenants locked in five-year contracts. This is because the first is dependent on spot demand, which fluctuates with bitcoin economics and AI capacity cycles, while the second has contractual visibility that supports debt financing and higher valuation multiples.
If Galaxy's Q2 disclosure includes something like "committed occupancy of our data center campus," I will read that as a strong signal. If it includes revenue from bitcoin mining operations only — meaning the facility is still primarily serving Galaxy's own mining equipment — then the AI infrastructure narrative has not yet reached its revenue stage. This is the single most important qualitative distinction in tomorrow's report.
The market sentiment right now is close to its peak for anything touching AI. The implied volatility embedded in GLXY options, if one can derive it from the listed derivatives, is likely elevated relative to the company's trading history. This tells me the market is positioning for a binary outcome: either Galaxy delivers on the infrastructure narrative with concrete data, or it disappoints and experiences a sharp repricing. The Bloomberg terminal numbers will confirm this, but you don't need a terminal to predict the pattern. When expectations have been built over four quarters without substantive delivery, the earnings resolution — whatever the headline numbers — tends to be violent.
Comparative Valuation: The Core Scientific and Riot Benchmark
To assess how Galaxy's infrastructure transition is progressing, I benchmark against its public comparables. Core Scientific emerged from bankruptcy with a focused strategy: bitcoin mining as the base load, AI high-performance computing as the growth engine. Its data center sites — primarily in Texas — run at significant scale, and it has signed multi-hundred-megawatt contracts with AI tenants. Riot Platforms operates similarly with an emphasis on its Texas expansion. IREN (formerly Iris Energy) has transformed itself into an AI compute play with meaningful contracted revenue.
These companies have traded at substantial premiums to their mining revenue multiples because the market is underwriting their AI infrastructure potential. But there is a difference between them and Galaxy that matters. Core Scientific, Riot, and IREN are pure infrastructure companies with a focused mission. Their capital allocation is singular: build, power, contract, repeat. Their management teams are infrastructure operators. Their compensation metrics are tied to power capacity and megawatt utilization.
Galaxy is not that. Galaxy has a diversified financial services business that requires ongoing management attention, regulatory compliance, and capital — a completely different set of disciplines than running a power-hungry physical campus. The question of whether Galaxy's management can successfully run both businesses simultaneously — and whether the financial services business begins to starve of management time and capital resources as infrastructure consumes more of both — is a genuine operational risk that the market underweights.
The market has a term for companies that attempt this dual-track transition. It calls them conglomerates. And conglomerates typically trade at a discount to the sum of their parts because the market struggles to apply a single valuation framework to dissimilar businesses. If Galaxy's financial services business were spun separately from its infrastructure business, the sum of the two independent entities would likely be higher than the current combined entity's market capitalization. This is a known structural discount, and it is a persistent drag on the stock.
I have been through this before in my career, analyzing companies that tried to bridge two worlds — most memorably during the 2017 ICO boom, when I spent six weeks reverse-engineering the Solidity codebase of an ICO that claimed to be a technology company but was actually a marketing operation. Code does not lie, only the architecture of intent. The same principle applies here: Galaxy's corporate structure and capital allocation tell you what it truly plans to be, regardless of what its press releases say. If the data center expansion is funded solely through the balance sheet and cash flows from the financial services business, then the company is underwriting its own infrastructure transition. If it brings in external partners, sells stakes, or signs long-term contracts with anchor tenants, then the transition is being validated by outsiders — which is a far stronger signal.
The next twelve months will reveal Galaxy's true architecture. The Q2 earnings call is where the first evidence arrives.
THE ERCOT PROBLEM: POWER IS THE UNSEEN CONSTRAINT
The most underappreciated risk in any Texas data center story is not construction cost or customer acquisition. It is the Electric Reliability Council of Texas — ERCOT — the entity that manages approximately 90% of the state's electric load. Texas has the most market-driven electricity system in the United States. This is both a blessing and a curse for data center operators. It means power is cheap relative to other jurisdictions — there is no regulated generation recovery mechanism that forces ratepayers to subsidize infrastructure. But it also means prices are volatile, particularly during summer heat waves when demand spikes.
A data center in Texas with a fixed-price power purchase agreement from a contracted supplier faces one set of economics. A data center participating in ERCOT's spot market for a portion of its electricity faces a completely different risk profile. When I was modeling the survival of various mining companies during the 2022 bear market, the ones that suffered most were those exposed to unhedged spot power. The ones that survived were those with either fixed-price contracts or demand-response programs that allowed them to curtail operations when prices spiked — effectively selling their non-consumption back to the grid.
Galaxy's Helios facility is in Dickens County, which is in the ERCOT service area. Its economic viability depends heavily on the structure of its power agreements. If Galaxy has locked in favorable rates and has the ability to curtail operations during peak demand — earning revenue from grid stabilization programs — its infrastructure economics are far more robust than spot-market exposure would suggest. If the company has not disclosed these arrangements, the risk is unquantified, and tomorrow's quarterly report may or may not clarify it.
The state of the electricity market is a factor in this quarter's results. Summer 2026 brought an early heat wave across Texas. ERCOT issued multiple conservation notices in June. Real-time electricity prices spiked during several periods. Any company with unhedged power exposure in Texas will show margin compression in Q2. For Galaxy, this presents as higher mining cost per bitcoin or higher data center operating costs, depending on the utilization profile. The market will interpret those costs through the lens of the infrastructure narrative: if they are contained, it signals good hedging discipline; if they break through, it signals that the infrastructure strategy is consuming more than it produces.
Truth is found in the gas, not the press release. The numbers that matter will be in the cash flow statement, in the notes to the financial statements regarding power purchase agreements, and in management's commentary on the earnings call. Anything less specific than a direct acknowledgment of power price conditions should be treated as a warning sign.
THE DATA CENTER'S DUAL ROLE
There is a reasonable probability — I estimate it as moderately high, based on comparable operating models across the public crypto mining sector — that Galaxy's Helios facility is currently serving dual purposes. Part of the capacity is likely dedicated to Galaxy's own bitcoin mining operations, which provide a direct hedge against the cost of acquiring bitcoin on the open market. The remainder may host customer equipment, either for bitcoin miners contracting for hosting services or, potentially, early AI compute workloads. This dual structure is a common pattern among publicly listed miners. It is logical and economic. But the market must not confuse hosting Bitcoin ASICs with operating an AI data center.
The engineering requirements are completely different. Bitcoin mining requires power-dense, high-heat-tolerant environments optimized for maximum hash rate per megawatt. AI inference and training workloads require sophisticated cooling, high-speed interconnect, low-latency networking, and redundancy standards that Bitcoin mining facilities simply do not meet. Converting a bitcoin mining facility into an AI data center is not a trivial checkbox exercise. It requires capital investment in cooling infrastructure, power conditioning, and network architecture. This is not a minor detail. It directly determines whether Galaxy's data center can actually earn an AI infrastructure premium or merely an industrial power hosting premium.
The market's pricing of Galaxy will depend on which category the market places the company in. An industrial power hosting company trades at a fraction of an AI infrastructure company. Should tomorrow's earnings reveal that the Helios facility is still primarily running ASICs, the AI premium embedded in Galaxy's valuation becomes fragile.
THE CONTRARIAN: WHY THE AI NARRATIVE MAY BE A TRAP FOR GALAXY
The current emotional landscape of crypto markets and public equities is characterized by a peculiar phenomenon: extreme enthusiasm for anything labeled AI combined with indifferent tolerance for actual cash-flow generation. The FOMO-to-fundamentals ratio across publicly traded infrastructure-adjacent stocks is somewhere between three-to-one and four-to-one. That is elevated. It has been elevated before. Every cycle, the market confuses an external narrative with an internal strategy and prices companies according to the former until evidence forces a correction.
Here is the contrarian case, and it deserves to be stated clearly: the AI data center buildout is a massive capital cycle, but not every participant in that cycle is built to survive it. The economics of the AI data center market are already showing signs of a bifurcation. Top-tier cloud providers — the hyperscalers — are writing enormous checks for dedicated AI data center capacity. Second-tier AI companies are signing contracts with providers like CoreWeave and specialized infrastructure firms. The second and third tier of demand is thinner than the hype suggests. A meaningful portion of AI infrastructure investment is directed at a small number of anchor tenants, and that concentration risk is building.
Galaxy entering this market as a late-to-scale player without the demonstrated engineering expertise of Core Scientific, the concentrated focus of IREN, or the established customer relationships of traditional data center REITs faces a genuine strategic disadvantage. The company's competitive moat is in financial services, not physical infrastructure operations. Competing in the AI data center market against entrenched players with deeper operating experience and stronger balance sheets — or against pure-play miners that have been managing power infrastructure since 2018 — is a different game entirely. This is not to say the pivot cannot succeed. It is simply to identify that the equity market now assumes success before Galaxy has demonstrated any operating track record in this domain.
The other angle rarely discussed is what happens to Galaxy's financial services business if management attention and capital become consumed by the infrastructure build. Trading businesses require constant management vigilance over risk limits, counterparty relationships, and market-making positioning. Asset management requires continuous investor relations and performance delivery. These are full-time disciplines. If Novogratz and his management team become preoccupied with megawatt contracts and cooling systems — as the narrative inevitably draws attention — the core financial services business may underperform its historical standards. The market will not price that erosion in immediately because management's stated focus is the infrastructure growth story. Until it appears in the numbers, usually eighteen months later, the underperformance is invisible.
To frame this differently: the market's demand for a new story is overpowering its memory of the business that generates the cash to fund the new story. That is a displacement that rarely ends well for shareholders.
History is a dataset we have already optimized. Every cycle, companies under pressure to deliver growth chase a new narrative and underserve the existing business that funds the chase. In the 2017 ICO audits, I saw projects with functioning products simply pivot to blockchain because the market demanded it and the cost of the pivot was destroying the original product value. In the 2022 bear market, I saw diversified companies cut their core revenue-generating units to preserve the infrastructure expansion narrative. The pattern is not random. It is the optimization of a dataset that insists on extrapolating the current narrative indefinitely.
Simplicity is the final form of security. A company that runs one business well — financial services or infrastructure — has a clearer equity story, a more predictable earnings profile, and a more auditable operating model than a company attempting to run both simultaneously. Galaxy is becoming compound. And compound complexity, in financial markets, is always priced as a discount, not a premium.
The market, however, is currently pricing Galaxy as if compound complexity were the point. That mismatch is the opportunity — and the risk.
REGULATORY AND ACCOUNTING CROSSWINDS
There is a less-discussed but equally consequential layer to this analysis: the accounting and regulatory treatment of Galaxy's digital asset holdings within a data center-heavy balance sheet. Under current U.S. GAAP — and the Financial Accounting Standards Board's fair value measurement updates for crypto assets — digital assets are measured at fair value, with gains and losses flowing through net income. This means Galaxy's quarterly earnings will always carry volatility from its digital asset portfolio, regardless of how the underlying business performs. The market is aware of this. But when a company simultaneously holds volatile crypto assets and invests heavily in fixed physical infrastructure, the balance sheet becomes a study in contrast: liquid crypto price swings against illiquid depreciation charges. Analysts must be careful not to confuse accounting volatility with business performance.
Galaxy's regulatory footprint is equally distinct. As a public company with operations in the United States, it is subject to SEC disclosure obligations, financial reporting standards, and the associated scrutiny. It has navigated this complex compliance environment since its inception. But the data center business introduces an entirely different regulatory category: energy consumption, environmental permitting, and infrastructure construction approval. A crypto-financial-services firm does not typically deal with county-level permitting departments or ERCOT interconnection queues. A data center operator does. This new regulatory friction is a real operational risk that is rarely included in the market's earnings expectations.
Additionally, there is the broader policy environment. The current U.S. regulatory posture toward digital assets has stabilized in recent years. There is no longer existential legal uncertainty about whether bitcoin and ether are commodities or currencies — that debate has been resolved in the practical regulatory framework. But the accounting treatment of digital asset holdings is still evolving, and any FASB changes affect the volatility of Galaxy's reported earnings. If the company's digital asset holdings produce large fair value gains in Q2, headline EPS will look strong. If the market is in a pullback, digital asset losses will offset infrastructure-related operating costs on the same income statement. The result is an earnings picture that is extremely difficult to read at first glance — which is precisely why the market relies on management guidance and non-GAAP metrics during earnings season.
I will be watching the non-GAAP measures Galaxy reports. Core operating profit, adjusted EBITDA, normalized trading revenue — these are the metrics that separate the digital asset valuation swings from the underlying business performance. If management emphasizes these metrics on the earnings call, they are signaling that they understand their shareholder base. If they emphasize the GAAP numbers or the story around them, the picture becomes murkier.
THE EARNINGS CALL AS A SIGNAL EVENT
The Q2 2026 earnings call will be Galaxy's first structured opportunity to directly address the market's infrastructure thesis. Novogratz is a charismatic, market-focused CEO. He will not be shy about discussing the data center vision. But the substance that matters — capital allocation commitments, contract disclosures, power purchase strategy — will reveal itself in his tone and specificity. A CEO discussing a data center expansion in general terms is different from a CEO discussing megawatts, committed occupancy, and depreciation schedules. The latter shows operational grip. The former is narrative extension.
Key questions the call must answer:
First, what is the current utilization of the Helios facility, and how is that utilization split between Galaxy's own mining operations, third-party hosting, and any AI workloads? A simple capacity metric would be the most transparent answer.
Second, what is management's CapEx guidance for the next four to eight quarters, and how is this being financed? Rising debt levels, equity issuance, or asset sales to fund expansion will each produce different market reactions.
Third, has Galaxy signed any long-term contracts with external tenants for data center capacity? A single high-profile tenant would transform the market's view of the infrastructure business.
Fourth, what is the company's power purchase structure? Fixed-price, market-indexed, or contracted with demand-response flexibility?
The answers to these questions will determine whether Galaxy's stock maintains its infrastructure premium or reverts to traditional crypto financial services valuation. The market has been lenient during the pre-earnings phase. That leniency ends tomorrow morning.
THE COMPARATIVE FRAME: WHAT CORE SCIENTIFIC'S TRANSITION CAN TEACH
Around this same point in the market cycle, Core Scientific provides a useful comparative template. After emerging from bankruptcy in January 2024, Core Scientific restructured around a dual thesis: bitcoin mining for base load plus AI HPC for growth. The company signed multi-year contracts with an AI tenant — a contract that initially was derided as speculative but was later realized. In the first two quarters after announcing its AI contracts, Core Scientific's stock re-rated significantly, and its analysts' coverage universe shifted from crypto miners to infrastructure.
The lesson for Galaxy is not that the pivot works — it is that the pivot requires contracted anchor tenants, not just capital expenditures. Core Scientific's stock only re-rated after the tenant signings were publicly disclosed. Without contracted external demand, the infrastructure thesis remains untested. Galaxy may have yet to reach that stage. If tomorrow's earnings call includes a signed tenant announcement, the market reaction will be immediate and positive. If it does not, the price action will remain tethered to the broader crypto beta.
A second lesson comes from Riot Platforms, which has been aggressively expanding its Texas mining facilities. Riot's valuation has remained more tethered to bitcoin prices than to infrastructure multiples because the market reads its expansion as primarily mining-focused. The market will read Galaxy's data center expansion the same way unless its customer mix demonstrates otherwise. There is no reason to expect a valuation premium for infrastructure capacity that serves only the company's own mining fleet.
THE BROADER CYCLE AND POSITIONING
We are currently in a market phase where capital is rotating toward physical infrastructure as a hedge against the volatility of digital asset prices. There is some historical precedent for this. During the 2020 DeFi summer, I conducted a deep-dive audit of Compound's governance token mechanism and identified a critical edge case in the interest rate model that could trigger liquidation cascades during high volatility. That analysis taught me something that has stayed with me: when markets become euphoric about one narrative, they systematically undervalue the risk infrastructure underneath the narrative. The same applies today. The AI data center narrative is emotional. The physical infrastructure requirements — power, cooling, construction timelines, permitting — are cold, hard logistics. The gap between the two is where repricing events originate.
The market has been through this before with crypto mining companies. In 2021, publicly listed miners traded at extreme premiums because the market extrapolated bitcoin's rally into every participant's financial statements. When the 2022 bear market arrived, miners with the most aggressive expansion plans suffered the most severe equity drawdowns, because their cash flows could not cover the debt servicing on capital investments sized for an optimistic cycle. The survivors were those with conservative balance sheets, flexible power arrangements, and the ability to slow CapEx quickly. Galaxy entered this cycle with a stronger balance sheet than most miners, and its financial services business provides a diversified revenue stream that most pure miners lack. But the fundamental principle remains: capital intensity amplifies cyclicality.
In my 2024 work analyzing transaction throughput on the OP Stack, I discovered a bottleneck in their state commitment processing that limited scalability during peak network congestion. I proposed a modification to the sequencer ordering logic that increased throughput by roughly 15%. The experience confirmed a principle I apply to all analysis, whether code or corporate finance: identify where the system is forced to wait, and you identify where its fragility lies. In Galaxy's case, the bottleneck is not the data center or the financial services business — it is the transition itself. The ability to manage both businesses simultaneously, without underfunding either, is the constraint. Every company in this position must optimize where capacity is applied, and the market will eventually reward clarity and punish ambiguity.
RISK MATRIX: THE NEXT TWELVE MONTHS
Let me lay out the specific risks that will determine whether Galaxy's infrastructure pivot creates or destroys shareholder value.
On the operational side, the principal risks are construction delays, power price spikes, and engineering overruns. A data center campus, once announced and partially capitalized, develops its own inertia. Construction schedules slip, interconnection timelines extend, and equipment delivery delays accumulate. None of these are fatal individually, but they compound into a later revenue recognition schedule and increased financing costs. In Q2 results, the earliest warnings appear in the balance sheet footnotes: increases in construction-in-progress, extensions in scheduled completion dates, or changes in expected power availability.
On the market side, the risk is valuation multiple compression. When the AI data center narrative cools — and every narrative cools eventually — companies that participated in the narrative without demonstrating contracted revenue will see their premium multiple revert. The speed and severity of that reversion depends on the strength of the underlying business. A company with strong recurring trading and asset management revenue can absorb a multiple compression without its share price collapsing. A company whose share price is built primarily on narrative will experience the full force of the repricing.
On the balance sheet side, I am watching for two signals. First, leverage. If Galaxy is funding its data center expansion through debt, its interest coverage ratio is now a function of both crypto market conditions and data center utilization — an uncomfortable combination. Second, equity dilution. If the company issues new shares to fund the expansion, existing shareholders bear the cost of that dilution, and the arithmetic matters more than the narrative. I have modeled this scenario. It produces a clear pattern: the share count rises, EPS drops before the infrastructure generates returns, and the stock trades lower for two to three quarters before the revenue ramp restores the earnings picture.
I will not be surprised if the Q2 report contains an announcement of additional financing. That would be a sign the board is serious about the expansion, but it will also be a headwind for the stock in the near term.
THE REGULATORY AND ESG FOG
As Galaxy's infrastructure footprint expands, so does its exposure to environmental, social, and governance (ESG) scrutiny. Public companies in Texas operating high-power facilities face increasingly vocal environmental concerns about energy consumption, water usage for cooling, and the emissions associated with natural gas peaker plants during peak demand. These concerns have already surfaced around bitcoin mining facilities in Texas. Data centers, whether AI or mining, attract the same scrutiny.
Beyond ESG, there is the question of permitting and grid interconnection. New data center capacity requires either physical connection agreements with ERCOT or direct arrangements with power producers. These agreements take time, face regulatory review, and carry financial penalties for delays. The market, currently focused on the AI growth narrative, systematically undervalues the length and complexity of power interconnection timelines. This is a recurring source of surprise in infrastructure transitions. When I analyzed the survival trajectories of public mining companies during the 2022 downturn, the single most important variable affecting their long-term viability was not the price of bitcoin — it was the cost and reliability of their power arrangements. The companies that had negotiated long-term fixed-price power contracts survived. The companies that relied on spot power entered distress. That data is public. That history is a dataset we have already optimized.
For Galaxy, the regulatory overlay is more complex than for a pure miner. The company simultaneously operates a New York-licensed financial services business, trades under SEC rules, and now operates physical infrastructure in Texas. Its compliance burden is the sum of the financial and physical regulatory frameworks. This complexity is manageable, but it carries administrative costs that are not always visible in the headline financials.
WHAT I'M WATCHING: THE SIGNAL TABLE
Let me be concrete about the specific signals that will determine how I and other careful market participants evaluate Galaxy's Q2 results.
First, the segment disclosure. If Galaxy introduces a data center or infrastructure segment with discrete revenue, margin, and utilization data, that is the clearest possible signal that the infrastructure business is maturing. If the report continues to fold data center operations into a "mining and infrastructure" line without separation, the business is still being treated internally as auxiliary rather than core.
Second, the CapEx guidance. A raise in forward guidance confirms the expansion is proceeding. A confirmation of prior guidance suggests the pace is stable. Any reduction will be treated by the market as an admission that the data center thesis has encountered headwinds.
Third, the balance sheet. An increase in construction-in-progress beyond current operational capacity indicates the expansion is physical. A rise in total debt with corresponding infrastructure asset growth suggests the company is leveraged to its own buildout. A share issuance announcement will create near-term dilution but signals confidence in long-term returns.
Fourth, management commentary on the call. The tone, specificity, and responsiveness of the CEO and CFO to infrastructure questions will tell the market whether this is a strategic commitment or a side experiment. A CEO who is fluent in megawatts, PUE ratios, and interconnection timelines is serious about infrastructure. A CEO who speaks about infrastructure in broad strategic terms is still at the narrative stage.
Fifth, and perhaps most important, the company's forward-looking statements about external customers. The first committed anchor tenant for the data center changes the entire risk profile of the expansion. It converts a speculative capital project into a contracted revenue asset. That event — when it occurs — will matter more than any single quarter's earnings.
THE PATH FORWARD: TWO SCENARIOS
I see two plausible paths for Galaxy over the next two years. In the first, the company executes the infrastructure transition flawlessly. It signs one or two anchor tenants, locks in favorable power agreements, and gradually moves from negative free cash flow to a stable infrastructure revenue stream. The financial services business continues to operate at its historical performance levels, and the market gradually awards the company a blended valuation — perhaps not the full AI premium, but a meaningful uplift over its pre-infrastructure multiple. In this scenario, GLXY becomes a genuinely diversified digital assets and infrastructure leader, and the equity story gains a resilience it previously lacked.
The second path is more sobering. The data center construction encounters the same delays, overruns, and liquidity drains that have affected infrastructure projects throughout the industry's history. The AI narrative cools before the company signs anchor tenants. The financial services business suffers periodic market downturns while management attention is diverted. Free cash flow remains negative for longer than expected, and the stock trades down to a pure crypto financial services multiple — potentially below where it traded before the infrastructure narrative began. In this scenario, the data center expansion becomes a permanent drag on valuation. It is the story of many companies across many industries that diversified into physical infrastructure during a hype cycle without a sufficiently committed customer base.
Which path Galaxy takes is not predetermined. It depends on the decisions management makes in the next two to four quarters — decisions that will be informed by the market's reaction to tomorrow's results and by the contracts and power agreements the company can negotiate during its expansion phase. The market, for its part, must decide whether it is willing to continue paying infrastructure multiples for capital expenditures that have yet to generate contracted revenue.
THE DECISIVE QUESTION
There is a question that sits at the center of Galaxy's Q2 2026 earnings release, and it is larger than the numbers themselves: is Galaxy evolving, or is it being carried by a narrative it does not control? Evolving companies act with a plan — they negotiate contracts, lock in power, secure tenants, and disclose progress proportionately. Carried companies speculate on the narrative's persistence, hoping that future valuation will justify current capital deployment. The difference is visible in the details. The details are in tomorrow's report.
I have been building financial models and auditing systems for nearly three decades, from the ICO audits of 2017 to the compositional risk analyses of 2020 to the infrastructure modeling that helped clients survive the 2022 bear market. If there is a single lesson I would leave you with, it is this: the question is never whether the narrative is attractive. It is whether the operating reality justifies the market price. The narrative is always attractive. The operating reality is what compounds.
Galaxy's Q2 earnings will not resolve the question. But they will provide the first serious dataset for evaluating the company's trajectory. For investors, the appropriate stance is not binary — buy or sell based on one quarter's numbers. It is to observe the specific signals and update probability estimates as the transition proceeds. The direction of the updates matters more than the starting point.
Tomorrow, the market will parse a financial statement and make an implicit judgment about a hundred billion dollars of physical and digital asset evolution. It will be listening for the sound of an engine turning — or the sound of a narrative thinning. In my experience, the two are rarely confused, but the market frequently chooses to ignore the difference until the evidence is undeniable.
TAKEAWAY: THE NEXT 48 HOURS AND THE NEXT 8 QUARTERS
The next 48 hours will establish the near-term trading floor or ceiling for Galaxy's stock. The next 8 quarters will determine whether the company's infrastructure pivot has created genuine enterprise value. The smartest approach for equity investors is to treat this quarter as a data point, not a verdict — but to hold management accountable to the metrics that differentiate a real infrastructure transition from a narrative slide. Watch the CapEx guidance. Watch free cash flow. Watch balance sheet footnotes. Watch the earnings call's tone. And most of all, watch for the first anchor tenant announcement. Until that event occurs, the data center expansion is a cost center with a story attached.
I have seen too many transitions in this industry, and too many repricings, to treat infrastructure announcements as value creation in themselves. What creates value is the contract. The customer. The locked-in power agreement. The disclosure that transforms a capital project into a revenue-generating asset. Until those appear, the prudent assessment is that Galaxy is in the early stage of a complex transition — with meaningful upside if it executes, and meaningful downside if it does not.
Galaxy's management has the opportunity, over the next two years, to prove that the infrastructure pivot is more than narrative necessity. They have the balance sheet, the relationships, and the regulatory positioning to do it. Whether they have the operational discipline to convert capital into contracted revenue is the question that will define the stock, the company, and ultimately the legacy of one of the most prominent names in digital assets.
As always in this industry, the proof is not in the story. The proof is in the infrastructure. The proof is in the data. The proof is in the gas.
The market is not a religion. It is a measurement device. It will measure Galaxy carefully tonight and tomorrow morning. The results of that measurement will be noisy, but over the next two years, the signal will emerge. That signal will be the answer to the question every serious analyst is asking today: does Galaxy's infrastructure story have a real foundation, or is it trading altitude on a narrative that the next earnings cycle will bring back to earth?
