管理層發言
Good morning, and welcome to Hut 8's Second Quarter 2026 Financial Results Conference Call. Joining us today are our CEO, Asher Genoot; and our CFO, Sean Glennan. Following the presentation, we will open the line for questions. This event is being recorded and a transcript will be made available on our website. In addition to the press release issued earlier today, our full quarterly report on Form 10-Q is available at hut8.com, on our EDGAR profile at sec.gov and on our SEDAR+ profile at sedarplus.ca. Unless otherwise indicated, all figures discussed today are in U.S. dollars. Certain statements made during this call may constitute forward-looking statements within the meaning of applicable securities laws. These statements reflect current expectations and are subject to risks and uncertainties that could cause actual results to differ materially. Certain key risks are detailed in our Form 10-K for the year ended December 31, 2025, and are continuous disclosure documents. Except as required by law, we assume no obligation to update or revise any forward-looking statements. During the call, management may reference non-GAAP measures such as adjusted EBITDA. We believe these measures, alongside GAAP results, provide valuable insight into our performance. Reconciliations of GAAP and non-GAAP results are included in the tables accompanying today's press release available on our website. We'll begin with a moderated Q&A session with our CEO, Asher Genoot, followed by a detailed financial review from our CFO, Sean Glennan. Let's get started.
Good morning, everyone, and thank you for joining us. I'll start today with a conversation with Mark Eidelman, our new Head of Investor Relations, who joined us in June from NextEra Energy. Mark has spent the last several weeks speaking with the research and investor community. And I asked him to share some of the comments, questions, and observations that he has heard most. After our discussion, Sean will walk through the quarter and then we'll open up the line for questions. Thanks.
Investors often describe Hut 8 as a bitcoin miner that transitions to data center development. I do not think that framing is quite right. What is Hut 8 and what does power-first actually mean?
I think everything starts with one simple observation. Electricity is becoming one of the scarcest resources in the economy. Hut 8 is an energy infrastructure platform. We build large-scale digital infrastructure around scarce power. AI, bitcoin mining, high-performance computing and whatever comes next are applications running on that platform. AI happens to be the highest-value application today. Power-first is not simply a development strategy. It is the operating system for how we allocate capital, manage risk and build the business. So bitcoin mining was our first proof point. We learned how to source low-cost power, build infrastructure faster and more efficiently and operate assets at scale. AI infrastructure rewards those same capabilities but across larger deployments, longer-duration contracts and more financeable cash flows. The operating model has not changed. The opportunity has expanded. In practice, our framework is repeatable: originate power, secure site control and interconnection, commercialize with high credit quality counterparties, finance efficiently and build and operate against long-duration contracted cash flows. River Bend, Beacon Point and the financings we will discuss today are outputs of that same framework. We're not building a collection of projects. We're building a platform that repeatedly converts scarce power into long-duration contracted infrastructure assets.
That makes sense. Many companies now describe themselves as power-first or as AI infrastructure developers. How should investors distinguish capability from the client?
Power-first is not the differentiator; capability is. The differentiator is the ability to consistently originate, commercialize, finance and execute infrastructure around scarce power. That starts with how we allocate capital. We don't underwrite applications. We underwrite scarce power. Applications change, customer demand changes, technology changes. Our job is to preserve the flexibility to commercialize that power through the highest-value use case over time. Beacon Point is a good example. When we first invested in the site, we underwrote a bitcoin commercialization path because it offered attractive risk-adjusted returns when we weren't sure the location hit the requirements of AI workloads. But we never underwrote the investment around one outcome. We preserved multiple commercialization paths from day one. And as AI demand accelerated and locations started being more of a preference rather than a requirement, we commercialized the same underlying power through a higher-value application. We do not predict the future. We built the flexibility to adapt to it. Our first two AI campuses were not existing bitcoin mining facilities that we converted. And I think that's really important for people to understand about the Hut 8 story. So our first two campuses that we've announced were greenfield campuses that we originated from the ground up, commercialized with investment-grade anchored counterparties, financed in the investment-grade markets and are now executing through construction. At the same time, approximately 700 megawatts of our infrastructure supports our affiliated tenant, American Bitcoin. That demonstrates that we can commercialize power through more than one application. And the market has already provided meaningful evidence of that capability. We have three 15-year leases with investment-grade anchored counterparties in the last nine months alone; two of the first investment-grade construction financings for single-sponsored data center projects and multiple greenfield campuses advancing through origination, commercialization, financing and now construction. In infrastructure, capability is not measured by what you say. It is measured by what customers sign and what capital markets finance.
So one of the things I want to talk about is financing. I financed projects at JPMorgan and helped build projects at NextEra. One of the things I'm going to introduce is why I left all of that and joined Hut 8. I think the best way to answer that is to start with what I've learned over the last 20 years. Great infrastructure companies are not built around individual assets. They're built around repeatable systems that can consistently originate, commercialize, finance, build and operate infrastructure over long periods of time. That is what stood out to me about Hut 8. River Bend alone did not convince me. Beacon Point alone did not convince me. What convinced me was that both were produced by the same system. Projects can be replicated; systems compound. Before I joined, I spent a lot of time challenging you, Asher, and the team on the risks: power origination and interconnection, customer relationships and counterparty quality, delivery timelines and construction risk, capital formation and financing, repeatability and the long-term vision. You all had well-thought-out answers to each of my questions that demonstrated that you understood the core risks and were mitigating them effectively. I was not looking for every risk to disappear. Infrastructure is not about eliminating risk. It's about understanding it, structuring it and allocating capital accordingly. Having spent my career financing and helping build infrastructure businesses, I recognized the same characteristics I've seen in the very best platforms: disciplined capital allocation, rigorous risk management and a repeatable system for creating value. This is also a rare opportunity to help build an infrastructure company at the beginning of its journey around one of the most valuable resources in the economy — power. Large infrastructure platforms are built by repeatedly applying the same disciplined framework over many years. I believe Hut 8 is at the beginning of that journey. So Asher, the topic we hear most from investors is execution risk. Hut 8 has not yet delivered projects of the scale on this timeline for counterparties of this quality. What is the basis for your confidence in on-time delivery?
I think execution starts long before construction. People often think execution begins when you start forming concrete. Construction is the final stage of execution, not the beginning of it. By the time construction starts, a lot of the most important decisions should have already been made. That's why we think about execution as a system and not an event. It starts with disciplined underwriting, power origination, site control, permitting, engineering, procurement, financing, counterparty alignment and construction sequencing. Every one of those decisions is made to reduce uncertainty before we mobilize on site. Our confidence rests on three things. One, priority: delivering River Bend and Beacon Point on time is our number-one priority. Our reputation and the repeatability of the model depend on it. Two, discipline: permitting, procurement, site work, power delivery and counterparty coordination all run in a single integrated schedule with conservative assumptions. And three, demonstrated capability: we have energized industrial-scale capacity before and repeatedly. The application has changed, but the discipline required to deliver has not. Every campus we develop makes the platform stronger. It improves our engineering, supply chain, execution, institutional relationships and ability to deliver the next campus. Execution is not something that we hope for. It's something we design for. Every campus we develop makes the platform stronger.
Talk about campus. Let's talk about River Bend. Where does construction stand today at River Bend? And what are the key milestones between here and energization?
Delivery is part of our model investors can verify in real time. We're very pleased with where River Bend stands today. The team is executing well. Structural steel erection began in early June. The building foundations are expected to be completed before month-end, and that opens up additional work fronts that allow crews to move in parallel rather than sequentially. We began steel erection on the substation in mid-July. We are now beginning slab-on-grade pours across the auxiliary support yard and in the main building. None of that is accidental; it's what disciplined sequencing, integrated planning and one delivery schedule are designed to produce. Every milestone does more than advance River Bend; it strengthens our engineering, supply chain and execution capabilities and our credibility for the next transaction. As our customers and our partners look at how we execute, they build more and more confidence. River Bend is not only a building. It's a campus. It's building capabilities that will make every campus after it and every building on the campus much better.
Thanks, Asher. Let's check to Beacon Point. We announced a second Beacon Point lease last month. What does that transaction demonstrate?
Beacon Point Building 2 is important for a much bigger reason than just signing another lease. It's another proof point that our framework is repeatable. The progression matters. River Bend demonstrated that we could commercialize a greenfield campus with an investment-grade anchor counterparty. Beacon Point Building 1 demonstrated that the framework was repeatable, but with a different customer. Beacon Point Building 2 demonstrated something different: an existing customer chose to expand under the same commercial framework. Different customers, same operating model, similar lease structure, same long duration, contracted cash flows. The second Beacon Point lease is for 352 megawatts of IT capacity and represents about $9.8 billion of expected base term contract value. With that lease, the campus is now fully commercialized with a full gigawatt of utility capacity supporting contracted investment-grade cash flows. The customer chose to double its footprint at Beacon Point. We think that's one of the strongest forms of validation an infrastructure platform can receive. Customers don't expand because of presentations; they expand because they have confidence in the asset and in our ability to deliver. Beacon Point also reinforces how we allocate capital. We originally underwrote the site for a bitcoin commercialization path, but we preserved multiple paths from day one. When the market evolved, we were able to commercialize that same power through a higher-value application and build it from greenfield. We didn't change the asset; we just changed the application. At the platform level, Beacon Point now represents 704 megawatts of contracted IT capacity, roughly $19.6 billion of expected base term contract value. Together with River Bend, total contracted AI data center capacity is about 949 megawatts, representing roughly $26.6 billion of expected aggregate base term contract value, all produced by the same operating model in less than a year. Every commercialization expands the platform, and that's what compounding looks like for us. A key part of River Bend and Beacon Point is not only the signed lease but the executed financing behind it. Mark, when you look at the River Bend financing from the outside, what did it signal to you?
Sure. Well, let me start with what impressed me most. It was not the size of the financing; it was what the market agreed to underwrite. The transaction consisted of $3.25 billion of fully amortizing senior secured notes due in 2044, rated investment-grade, issued at the project level, nonrecourse to Hut 8 and backed by contracted lease revenues from a campus still under construction. Investment-grade markets have historically not financed construction-stage data centers, especially single-sponsor, single-asset projects. The rating agencies and fixed-income investors underwrote the contract structure, counterparty credit and backstop, Hut 8's delivery model and risk allocation for 16.5 years, covering the expected construction period and the entire 15-year lease. Having spent years on the other side of that analysis, I can tell you that credit committees do not finance ambition; they finance certainty and execution. That was institutional validation of the development model, especially the most rigorous currency there is committed capital at investment-grade pricing. It also proved the capital formation model. Each project raises debt against its own contracted cash flows, nonrecourse to the parent and fully amortizing. That generally ring-fences development risk and preserves capacity at the parent level. River Bend did more than finance one campus; it created a repeatable template for financing future campuses. Beacon Point then applied that template again, and on even better terms. Beacon Point financing executed on better terms than River Bend: a higher rating, better pricing and greater scale. What did that improvement tell you? And how does the capital structure support growth from here?
I think it tells us that capital follows capability. We did not get an investment-grade financing because we wanted it; we earned it through disciplined execution. When we first started River Bend, we went to the rating agencies and the investment-grade result was because of what we presented them, not because we went in expecting that. Beacon Point consisted of $4.25 billion of senior secured notes. The notes were rated one notch higher than River Bend, and they priced 20 basis points inside of River Bend. The offering was substantially oversubscribed with repeat investors returning and new investors joining. We pushed amortization from two years on River Bend to four years on Beacon Point. We did not copy the transaction; we structured every term from first principles and the result was improved ratings, pricing, scale and amortization. We didn't negotiate our way to better terms; we earned them through the way we structured and built the second project. The structure is also what allows us to scale. It's fully amortizing, so there's no refinancing wall at the project level. It's nonrecourse, so there's zero recourse debt at the parent level. And it's nondilutive to equity holders. Each project is designed to generate sufficient cash flows to support the related construction financing. Growth is not constrained by the corporate balance sheet, and we can develop multiple campuses at once. Together, River Bend and Beacon Point Building 1 represent approximately $7.5 billion of investment-grade capital raised for construction-stage development. Every successful financing expands the platform's ability to finance the next one. This is capital formation compounding in real time. Capital follows capability and better capital is earned through better execution.
Ash, let's shift gears a little bit. Behind the model of the capital, how do you build an organization that can deliver at this scale and keep delivering as the platform grows?
People ultimately determine whether a platform can compound over time. Organizations don't scale because they own great assets. They scale because they build capabilities that can be repeated, and people create those capabilities. We built the organization around the actual life cycle of a project, not around a traditional corporate org chart: origination, underwriting, development, financing, delivery, operations. We've been equally deliberate about the type of people we recruit. We want builders who take ownership, enjoy solving hard problems, think from first principles and want to build something that compounds over decades, not quarters. I've said before that this is more of a religion than a job, and like-minded folks really attract each other. I think of the team we assemble for a project almost like a group of Navy SEALs rather than an army. Everyone brings a specific skill set, everyone has operated at a high level in that function and they come together as one unit to execute from start to finish. We've invested in talent with deep backgrounds across power, development, infrastructure, procurement, project execution and capital markets. That investment shows up in SG&A, and we do not view it as overhead creep. We view it as an investment in platform capacity and capability because we're so focused on growth and scale. After our first two campuses, we now have a repeatable framework across design and engineering, supply chain, contracting, financing and delivery. Organizations learn; capabilities compound. Every campus improves the team and the improved team makes the next campus better. People are not separate from the platform; they are the platform, and they are the capability that compounds every other capability.
Thanks for that, Asher. Let's talk about the pipeline. Investors want greater visibility to the development pipeline. How do you decide what enters the pipeline and what changed this quarter?
I think it's a fair ask, and the first step is to frame the question correctly. The goal is not to build the largest headline megawatt number. The goal is to convert the right opportunities into financeable, commercializable infrastructure. We manage the pipeline like an underwriting exercise. A project must clear a series of gates before it moves forward: power scale and speed to power, interconnection certainty, site control and a path towards permitting, network access, customer demand, capital intensity and risk-adjusted returns. Every megawatt in the reported pipeline has already been tested against those criteria. That's what makes the number meaningful, not simply large. Through that lens, the development pipeline now stands at about 8.7 gigawatts, up approximately 300 megawatts from last quarter. We have 11 sites in the under-diligence and under-exclusivity stages, averaging more than 650 megawatts each. On average, those opportunities are larger than each Beacon Point building. We did not only grow the pipeline; we advanced it. The exclusivity stage increased by 200 megawatts as projects moved forward from diligence. Importantly, the reported number also excludes M&A opportunities, behind-the-meter power generation solutions and potential River Bend expansion where the tenant holds the right of first offer on the next gigawatt. The direction of opportunity flow is changing. River Bend and Beacon Point have led more developers and power producers to bring opportunities to us rather than the other way around. They see our ability to execute, our ability to finance at scale and our deep tenant relationships. We're having more inbound interest from developers who have a piece of land and interconnect and need someone to commercialize that for them. That's a sign that the platform itself is beginning to compound. We don't optimize for the biggest pipeline; we optimize for the highest-quality pipeline. Investors should underwrite the platform's ability to repeat, not only the next lease.
Last question for you, Asher. There was clearly a philosophy underpinning these answers. How would you simply summarize our philosophy for investors?
It comes down to a handful of principles we return to every day. Scarcity creates opportunity. First principles identify that opportunity. Optionality protects capital. Commercialization creates value. Execution earns trust. Capital follows capability. Platforms compound. Everything starts with power, and we have built the operating system that turns those principles into contracted cash flow. We are proud of what the team has accomplished, but we still believe we are very early. Every campus strengthens the platform. Every financing expands our capabilities. Every customer deepens our relationships and every great person makes the organization stronger. What I would encourage investors to underwrite is not our next project but our ability to compound capabilities over time, create projects that create earnings and compounding capabilities that create enduring enterprise value.
Asher, thank you. Sean, let's turn to the quarter's financial results. Investors can read the income statement in the 10-Q, so I want to focus this discussion on what the numbers say about the underlying business, the balance sheet and Hut 8's ability to finance growth. Revenue increased meaningfully year-over-year and adjusted EBITDA improved, yet the quarter still showed a significant GAAP net loss. How should investors reconcile those results?
Thanks, Mark. I think there are three key takeaways in our financials. One, the operating business grew. Two, margins expanded. And three, EBITDA improved. Moving to the P&L items themselves, revenue increased approximately 81% year over year to $74.9 million, while cost of revenue increased by approximately 23%. That produced gross profit of approximately $48 million and expanded gross margin to approximately 64%, compared with approximately 47% in the prior year period. Adjusted EBITDA, excluding digital asset mark-to-market movements, was $10.4 million. That compares with $4.2 million in the prior year period. The GAAP net loss of $177.1 million was driven primarily by a $138 million loss in digital assets. Bitcoin declined during the quarter while it had increased materially in the prior year period. So the year-over-year comparison is dominated by a non-cash mark-to-market swing.
Thanks, Sean. Let's go one level deeper. What were the most important drivers across power, digital infrastructure and compute?
Compute remained the primary operating contributor. Revenue increased to $72.5 million from $34.3 million, driven by an increase in bitcoin mined from approximately 308 to approximately 935. That growth reflects additional operating capacity following the commencement of operations at Vega and the re-energization of our Drumheller facility. Compute cost of revenue increased at a much slower rate than revenue itself, resulting in a segment gross margin of approximately 66%. That operating leverage is important because it demonstrates the earnings capacity of the current platform even before our contracted AI data center revenues begin contributing. Digital infrastructure revenue was $1.3 million, broadly consistent with the prior year period. Today, that segment still reflects the legacy base. Its financial profile changes materially as River Bend and Beacon Point data halls are delivered and the associated long-duration lease revenues begin coming online. Power revenue declined to $1.2 million from $5.5 million, and that is primarily because the prior year quarter included a full quarter of activity from the Far North portfolio, which we sold in February. That decline is therefore a function of portfolio management rather than a decline in the core business.
General and administrative expense increased substantially. How should investors distinguish between recurring overhead and investment in the platform?
Reported G&A was $76.1 million, compared with $30.2 million in the prior year period. Approximately $43.6 million of the increase was share-based compensation, so the majority of the year-over-year increase was non-cash. Cash investment also increased as we added talent and capabilities to support a much larger development platform. Salaries and benefits increased by approximately $4.1 million, primarily from additional headcount supporting growth initiatives, particularly in our energy origination group. We evaluate SG&A spending through a growth-first maintenance lens. The organization required to maintain today's operating base is meaningfully smaller than the organization required to originate, finance, construct and operate multi-megawatt campuses in parallel. That does not mean growth spending is unconstrained. We expect every investment in people, systems and capabilities to be tied to specific commercial outcomes: more high-quality power origination, faster project conversion, lower cost of capital, improved execution and stronger operating leverage over time.
Let's talk about the balance sheet. It looks very different. Cash and restricted cash increased to approximately $7 billion and total debt increased to approximately $7.6 billion. What is the right way to interpret those figures?
The first distinction is between corporate liquidity and project-restricted capital. At June 30, we had approximately $233.6 million of unrestricted cash and approximately $6.8 billion of restricted cash and cash equivalents. The restricted cash primarily represents proceeds from the River Bend and Beacon Point financings, and those funds are held in project accounts and can only be used for construction, debt service reserves and other specified project purposes. That is not excess corporate cash, and the related debt is not general corporate leverage. Similarly, the majority of the $7.6 billion carrying amount of debt consists of the $3.25 billion River Bend notes and the $4.25 billion Beacon Point notes. Those obligations sit at bankruptcy-remote project subsidiaries. They're secured by the applicable project assets and accounts and, importantly, are nonrecourse to Hut 8's parent company. The consolidated balance sheet has become larger because two of our three projects under construction are fully financed. Economically, each project is designed to service its own debt from its own contracted lease cash flows. That's the financial architecture we want: ring-fence project risk, preserve parent flexibility and minimize reliance on corporate equity.
Investors will also note that interest expense increased sharply, while interest income increased to $27.1 million. How should we think about the construction-period carry on these financings?
Interest expense increased because we closed $7.5 billion of long-duration project financing during the quarter. That's expected when fully funding two campuses before the related lease revenues begin. Importantly, undrawn construction funds are invested in short-duration instruments within project accounts. Those funds generated $27.1 million of interest income in the quarter, partially offsetting the interest cost on the notes. We also capitalized $5.7 million of interest into construction in progress during the quarter. The accounting therefore reflects three components: interest expense recognized currently, interest income earned on undeployed proceeds and interest capitalized as part of the cost of the assets under construction. We structured these financings to remove refinancing risk and secure the full construction capital upfront. There's a cost of carrying committed capital during construction, but we believe that cost is outweighed by the certainty of funding, protection against future capital market volatility and the ability to execute without returning to the market to build.
How did the Coatue conversion and the FalconX refinancing change the parent-level balance sheet during the quarter?
In May, Coatue converted approximately $159.3 million of accretive principal balance of its note into 9.7 million shares. That eliminated our only remaining parent-recourse debt. We also refinanced a $200 million Coinbase facility with a new $200 million FalconX term loan. The coupon declined from 9% to 7% as a result of the refinancing, and the facility is collateralized by bitcoin, not the parent. Those transactions simplify the parent capital structure. Excluding ordinary-course obligations, the parent is not obligated under the River Bend or Beacon Point notes, and the remaining significant financing is secured by a discrete pool of bitcoin. That matters because one of our most valuable corporate assets is flexibility. A clean parent balance sheet gives us the ability to fund early-stage development, absorb timing differences, pursue strategic opportunities and choose the right financing for each asset rather than being forced into whatever financing happens to be available at a specific point in time.
Beacon Point Phase 1 financing and the second phase are now contracted. What principles will guide financing Beacon Point Phase 2 and the broader development pipeline?
It comes down to four principles. First, asset-level self-sufficiency: we will seek to finance each project against its own contracted cash flows with risk generally ring-fenced at the project and no recourse to the parent wherever feasible. Second, optimization rather than repetition: River Bend established the market; Beacon Point improved on that execution with a larger issuance, a lower coupon, a higher rating and a later start to scheduled amortization. We will evaluate each next financing from first principles — the asset, the lease, the construction schedule, market conditions and investor demand. Third, disciplined use of equity: equity should fund the portions of the development cycle where it creates the most value — origination, site control, interconnection, design and other work required to convert an opportunity into a financeable project. Once contracted cash flows are in place, we want long-duration project capital to fund construction. Fourth, preserving liquidity across the portfolio: the model needs to support several campuses advancing at once. That means matching duration, amortization, covenants and recourse to the economics of each asset while maintaining capacity at the parent.
Thanks, Sean. To close, what should investors take away from the quarter from a financial perspective?
First, the operating business has strengthened: revenue grew, gross margins expanded and adjusted EBITDAX, excluding digital asset mark-to-market, increased year-over-year. Second, the capital formation model moved from concept to repeatable execution: we raised $7.5 billion of investment-grade, long-duration project financing for two construction-stage campuses with no recourse to the parent. Third, the parent balance sheet became cleaner: the Coatue note converted, the bitcoin-backed facility was refinanced at a lower coupon and the majority of consolidated debt is now matched to contracted project cash flows. Finally, the financial profile is in transition. Today's income statement is still dominated by compute and digital asset accounting. As River Bend and Beacon Point are delivered, the mix should shift meaningfully toward long-duration contracted digital infrastructure cash flows. Our focus is to manage the transition with discipline, execute the projects, protect the parent balance sheet and finance growth in a way that compounds value per share.
Thank you, Sean. That concludes our prepared discussion. Operator, please open the line for questions.
分析師問答
Our first question will come from the line of Stephen Byrd with Morgan Stanley.
I wanted to just dive into behind-the-meter generation and really just get your overall temperature check in terms of how desired is this by your customers? I guess this can really help to create much larger sites and move much faster, potentially. So it strikes me as a very good complement to the grid access that you have. Asher, you've spoken to this before, but just curious on your latest thinking: how likely is this in your view? How important is this to your customers to be able to achieve both the timing and scale objectives that they have? I'd love any comments you might have on that.
Behind-the-meter capacity will happen. We see the demand and the opportunities within our pipeline, and it's often the fastest path to power. Customers want it and the grids in the regions where we're building want us to bring additional power and, in some cases, help offset consumption on the grid. The reason we don't include behind-the-meter opportunities in our development pipeline is because we view those megawatts as less authentic for public reporting. If we have a piece of land with interconnection in place, from a pure land-and-interconnect perspective we could theoretically put as many megawatts as the land and interconnect can support. For example, River Bend could be a multi-gigawatt site if we counted behind-the-meter in that way. Including those would make our reported pipeline far larger than the 8.7 gigawatts we disclosed. We treat behind-the-meter similar to M&A: when they become real, executed or contracted, they become visible catalysts. We're working on many opportunities for both behind-the-meter and M&A across the team today.
Our next question will come from the line of Brett Knoblauch with Cantor Fitzgerald.
I know there was a letter yesterday that had a lot of people asking some questions. I'm curious to get your thoughts on it: to what extent is it a big-point grandfathered situation? And how does that change your view of where you're looking to grow the portfolio from a pipeline perspective?
Across the U.S. today, more politicians want to ensure that ratepayers and local voters feel protected. We saw the letter and we trust the legislative process. We're prepared to work with the PUC and ERCOT to implement any new process. We feel confident in the package we put forward during the batch process; many elements align with the points raised by the governor, including grid reliability, water usage, environmental considerations, noise, traffic, emergency and other community protections. We voluntarily participated in the PUC survey and provided detailed information on Beacon Point around water and power usage, both operating and under construction. We'll do the same with the governor's request. As we develop across the pipeline in places like Texas, Louisiana, Alabama and other southeastern states, there are some states where officials want the business of data centers but also want to make sure communities feel protected. That sensitivity affects how quickly and how aggressively we invest in certain regions. Overall, I think it's healthy for the U.S. development environment because it encourages better dialogue with communities and reduces the 'he said, she said' noise that creates fear among local officials. We'll continue to be thoughtful and engage communities to ensure development is responsible and aligned with local priorities.
Awesome. If I can follow up on River Bend: how quickly could behind-the-meter at that site get stood up? Would that come before additional power delivery from the grid? Walk me through how River Bend expands from here via grid, behind-the-meter or directly from energy.
Building 1 is being constructed through 2027 with halls handed over sequentially. For Building 2 to start delivering data halls, it would be on the back of Building 1, so you could think of end of 2027 as a timeframe for additional capacity to come online. For behind-the-meter generation, some solutions can provide power faster than the data center construction timeline, though not all. River Bend is a unique environment with a supportive state and local community that wants this business and a strong local workforce for skilled trades. The gas pipeline access is there and we've confirmed that, and we're working with Entergy around capacity. I see a world where you could see behind-the-meter generation working in concert with Entergy grid-connected capacity to accelerate and expand the site.
Our next question will come from the line of Darren Aftahi with Lucid Capital Markets.
On your exclusive energy basket in your release — the roughly 1.9 gigawatts — could you characterize where those sites are? Brownfield or greenfield? And how would you characterize geographic and community risk, given some of the governor's letter comments?
We're pretty diversified. The best way to think about the company is by how we develop. The pipeline we report is primarily greenfield opportunities: land plus interconnect. M&A opportunities and brownfield assets that are at different stages are not included in that pipeline. We have multiple teams organized into five regional pods, each covering a set of ISOs. Each pod has a budget and people focused on diligence and development. Under diligence, teams put in land options, interconnection agreements, studies, pre-construction work, site surveys and geotechnical work. In exclusivity, there's a line of sight on power in addition to land control and permitting. Local community support is paramount as a project moves from diligence to exclusivity; that's when we commit more resources. As a result, the pipeline is diversified across the U.S., across multiple states and ISOs, and not concentrated in a single region by design. We built separate pods to ensure we have the bandwidth and expertise to progress multiple opportunities in parallel. Internally, our pipeline is broader than what we disclose publicly because we set a higher disclosure threshold; we avoid 'bragawatt' numbers and only report opportunities that have cleared meaningful diligence gates.
Our next question will come from the line of Stephen Glagola with KBW.
Can you provide more detail on how you intend to fund the equity component associated with the Beacon Point Phase 2 lease? And Sean, what's your broader view on the funding markets today for project financing? Has availability changed over the last few months?
If we were to follow a similar structure to River Bend Building 1 and Beacon Point Building 1 — meaning long-duration investment-grade bonds at the project level — the equity commitment is within the balance sheet capacity we have and we've thought through equity dilution scenarios carefully. From Building 1 to Building 2, we focused on first principles and improved metrics across the board. For Beacon Point Building 2, we're applying that same first-principles approach to determine the most accretive and durable capital structure. We'll share more details in the coming weeks, but given our current balance sheet and the interest from financing counterparties, we feel confident in the options available to fund the equity component in a way that supports long-term value creation.
The market remains open and receptive to a lot of different paper. There's certainly a lot of supply coming, but there's also discernment among investors for quality leases, quality operators, quality developers and quality structures. That's why we are principled in structuring our debt deals: we want them to be attractive to the market and get strong receptivity. We've developed a pretty good following in the fixed-income markets, and for issuers who execute well, the market should remain open and provide good pricing. Ultimately, it will be issuer-by-issuer going forward, so we continue to focus on quality and structure.
Our next question will come from the line of Ben Summers with BTIG.
Asher, you mentioned M&A opportunities. What are you seeing in that market, and are there specific power markets where you're seeing more acquisition opportunities?
We have a lot of inbound inquiries every day from across the country. Roughly 70% of those opportunities don't meet our threshold and are a waste of time; about 30% are interesting. We've expanded the team to diligence and vet these prospects. Some developers have land and interconnect but lack capital or the tenant relationships to commercialize projects. Our platform and our three announced leases have generated reputational credibility, leading to more inbound interest. We typically don't pay the highest upfront price for development-stage land; our development capital is usually lower and more heavily weighted toward post-commercialization deployment. Many developers are willing to take backend economics in exchange for our ability to commercialize and execute. We look to structure deals where the seller and Hut 8 share aligned incentives: we bring commercial execution and they retain upside if the project is realized. Our ability to get quick indications of tenant interest has also improved, enabling us to focus on the right opportunities.
Our next question will come from the line of George Sutton with Craig-Hallum.
Asher, you talked about existing customers that have the right to new megawatts. Are you operating on behalf of some customers relative to M&A opportunities and going to market that way, or how should we understand those rights in context?
No, we are not operating on behalf of customers in the sense of proactively buying assets for them. For some opportunities we have rights-of-first-offer where customers get a first look and can elect to take capacity. For many of our potential sites we enter discussions with two to three likely tenants in mind and can get responses within days as opposed to weeks. Our tenant relationships are deep and active; we know their criteria and can quickly assess fit. For M&A, historically it required directional bets and significant development capital. We're more comfortable with greenfield where the cost basis is lower. Two changes in the past year are important: one, we can now get demand signals quickly; and two, many developers are willing to accept backend economics rather than requiring us to fund the full development upfront. That alignment makes certain M&A opportunities much more attractive.
Our next question will come from the line of Joe Vafi with Canaccord. Our next question will come from the line of Brian Dobson with Clear Street Equity Research.
At the risk of beating a dead horse regarding the governor's statement, do you think this might help wash out some of the weaker players in the ERCOT queue and favor established players like Hut 8?
I think many of these initiatives will do exactly that: reduce noise and surface the developers that have robust execution capabilities. There's a lot of speculative activity where people spend tens of thousands on interconnection studies and land options without having the capital or execution track record to commercialize a large campus. That speculative noise scares local stakeholders. We believe responsible development requires community engagement and demonstrating benefits to local economies. Our first site was a brownfield redevelopment of a former industrial facility in Niagara Falls, New York, and we've always tried to be thoughtful about community impact. Companies willing to invest the time to build community trust and demonstrate environmental and operational discipline are often the ones with the deeper development capabilities. So yes, these processes should favor more experienced and responsible developers.
Follow-up: there's concern in the broader market about CapEx spending from hyperscalers. Are you seeing anything that indicates they are taking the foot off the gas on data center development?
From our conversations, demand is robust. Tenants want capacity and often want it quickly; that has been the pattern for the last two years. Our perspective is biased because we've built deeper tenant relationships and are front-and-center in many conversations, so we see a consistent level of demand. Individual tenants will ebb and flow and sometimes pause, but we've seen those temporary pauses reverse. Overall, market demand remains strong.
Our next question will come from the line of Patrick Moley with Piper Sandler. The next question will come from the line of Will Cox on for Patrick Moley.
Specifically regarding your gigawatt diligence agreement with Anthropic, could you give us an update on your talks and relationship with the company and where this ranks in your priorities relative to the 50 megawatts under development, River Bend expansion, or movement of megawatts from exclusivity or diligence into development?
Anthropic is a great example of a customer with very large demand requirements. We work closely with them and have a good relationship. We're building the campus at River Bend for them, and we continue to explore additional opportunities and novel structures with them. They remain a strategic and engaged tenant, and we continue active discussions on future capacity and deployment options.
Our next question will come from the line of Chris Brendler with Rosenblatt Securities.
Congrats on the progress. A quick question unrelated to the data center business: can you give us an update on your ownership and current position in American Bitcoin given recent developments? And how are you thinking about the company's bitcoin holdings on the balance sheet?
We own roughly 54% of American Bitcoin today. American Bitcoin reported strong operating results, including the most bitcoin mined in a year even though bitcoin prices were down. Margins remain around 50% gross, only modestly lower despite price declines. The underlying operating business is strong, even if market sentiment and liquidity in bitcoin have pulled back. For Hut 8, bitcoin is a corporate asset like cash. We would opportunistically sell if there were attractive use cases to fund initiatives, but so far we have been able to finance projects without needing to liquidate bitcoin holdings. Going forward, our exposure to bitcoin will be primarily through American Bitcoin.
Our next question will come from the line of Nick Giles with B. Riley Securities.
There's a lot of dialogue around upward pressure on build costs. How much of your CapEx is already secured on contracted capacity? Are there further contracts to negotiate with suppliers? How has your procurement strategy shifted as supply chains tightened?
For the first building on each campus, long-lead items are fully contracted: 100% of long-lead items are contracted, general contractor agreements and subcontractors are in place, and pricing is fixed to align with financing. Building 2 was cheaper than Building 1, and we expect Building 3 to be cheaper than Building 2. We're focused on continuous improvement in design, construction and supply chain. We build deep partnerships with suppliers rather than one-off purchases. We engage at the CEO level with many of our suppliers and collaborate on innovation such as skidded equipment design. We haven't seen major impacts on lead times or allocation for our projects because we prioritize relationships and partnerships with vendors that share our long-term vision. We're continuously pushing to improve time-to-build and cost-to-build as we scale.
Our next question will come from the line of Allen Klee with Maxim Group.
On a site-level basis for the Digital Infrastructure segment, as the leases fully scale up, how do you think about gross margins and adjusted EBITDA margins?
You'll see those margins increase. Consider roughly $27 billion of contracted revenue equates to about $1.7 billion of annual cash flow. Because these are triple-net leases, operating costs are largely passed through to tenants, so a substantial portion of that $1.7 billion drops to the bottom line, which is why we showed roughly a 99% NOI margin at the project level. From there, your cost is primarily servicing principal and interest on project bonds.
That's right. Project-level margins have been essentially in the high 90s for these triple-net lease structures. We'll avoid providing forward-year guidance today, but we'll maintain a keen eye on SG&A to ensure we're investing for growth and not allowing overhead creep. Asher's point is correct: model the lease economics, the amortization schedules on project bonds and a disciplined SG&A assumption to estimate net cash flow.
Most people in our organization are focused on net new growth, not merely running the existing portfolio. If we were only managing the current public company and the three announced data halls, we could operate with significantly fewer people. The majority of our team is focused on originating and delivering new campuses, which we view as an investment in future scale and margin expansion.
This concludes our question-and-answer session. Thank you all for joining.