Earnings Call Transcript - HUT Q4 2025

Operator, Operator

Good morning, and welcome to Hut 8's Full Year 2025 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 annual report on Form 10-K 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 our other 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 metrics 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 will begin with a moderated Q&A session with our CEO, Asher Genoot, followed by a detailed financial review from our CFO, Sean Glennan. Okay, everyone. So let's get started. So to start Asher, fiscal 2025 was a big year for Hut 8. We executed on several important milestones, including the carve-out of our legacy Bitcoin mining business and the execution of our first AI data-center transaction. So what were the guiding principles that enabled us to achieve these outcomes in your mind in 2025?

Asher Genoot, CEO

2025 was about rebuilding Hut 8 around capital efficiency and durable cash flow. So two years ago, we rebuilt the company from a first principles approach after our merger with Hut 8 and going public. And everything started with the electrons. We chase megawatts, not chips, and we want to control the power layer first. And so we don't view electrons as a commodity, but rather strategic assets, and the ABC carve-out shifted us from cyclical CapEx exposure to contracted infrastructure-like cash flow. So that was a big theme of last year. We also reallocated capital from volatility to long-duration agreements, and with ABC being able to self-fund on itself on the mining and Hut 8 providing the infrastructure. And then River Bend validated that model. We had a true greenfield development. We didn't convert a site. We developed the site from the ground up with a power-first thinking. It was really the first domino to fall under an AI infrastructure platform. And we focused only on what compounds: power control, scalable campuses, disciplined capital structure, and repeatable execution. So 2025 was about building the right foundation, and now we compound and scale going into 2026.

Operator, Operator

In your mind, what specific operational and strategic milestones were prerequisites before the business could meaningfully accelerate?

Asher Genoot, CEO

2024, as I shared a little earlier, was about restructuring. We started a company, we merged with Hut 8, and then I took over shortly after. And it was restructuring the business and creating the ability for us to create a foundation. 2025 was about building credibility and so credibility started with our shareholders. In 2024, when I took over the helm, our institutional ownership was sub-10%. We're at approximately 70% today. Some of our earliest shareholders, who invested in us when we started the company 5 years ago, still hold a large percentage of the stock. We've given shareholders disciplined capital allocation that they've seen through us in the years and with us in the public markets. And then, honestly, transparent execution, we told people what we were going to do, and we focused on getting that done and delivering a fully built solution when we execute, and we intend to do the same. We built credibility with team members. We scaled intentionally. We institutionalized processes while maintaining an entrepreneurial speed. I think that's critical. As you grow, you naturally build bureaucracy, and having the mindset to say, you know what, we will continue to operate as a lean culture, even as we build an institutionalized process. I spent a lot of my time thinking about that in terms of how do we scale at the same rate that we have historically. Credibility with financing partners is really important for us to secure Tier 1 lenders with JPMorgan and Goldman Sachs to lock in nonrecourse project financing. It was harder, but we believe it's the better path. We believe early investors in us like Coatue, they backed us before the theme was a consensus. I'm glad that they're happy. I meaningfully spoke on a panel the other week, and I'm glad that we're able to drive a good return on their investments. Lastly, credibility with our partners. That's local partners, for example, at River Bend with Entergy, Louisiana, West Feliciana Parish, the Governor's office and being able to deliver on them and our commitments to them when they took a bet on us. Acceleration only happens when credibility is established, and we would like to compound on that as the years go by.

Operator, Operator

So you were under a lot of pressure last year to talk about a deal and guide the market to when it would come, what it would look like, but we kept our cards very close to our chest. So can you talk about why we were patient in what we wanted to see come together first?

Asher Genoot, CEO

I'll separate my role in speaking to the public markets and our shareholders, and my role in running the operating business and making sure we're able to build that over the long term. In my role in the public markets, I knew what shareholders wanted. They wanted a deal. They want to know that we can build the data center infrastructure platform, and they wanted to see what that would look like and to build the first domino as a part of the platform. For the business itself, we weren't waiting to announce a deal. We were really building a fully locked and executable program. For us, we didn't feel like there was a reason to add public market complexity while negotiating a super complex structure with a lot of moving pieces. We didn't want to just announce a headline. We wanted a complete financeable program. We wanted demand secured, financing secured, execution partners aligned on construction, delivery, engineering, long lead time items. We wanted our power path defined, and we wanted risk allocated properly across counterparties. Instead of guiding towards a deal that's coming, which everyone knew we were working hard at, we optimized for building the right structure for the company in the long term. When everything was negotiated and aligned, we announced the full framework in one step. That discipline hopefully reinforces the credibility that we're building with our shareholder base. As we move forward, we're going to be really thoughtful about what we share with the market versus how we think about those impacts on our customers as well.

Operator, Operator

So our first AI data center transaction generated significant market attention. How should investors think about this transaction in the context of our broader strategic evolution without over-indexing on a single data point?

Asher Genoot, CEO

I think this is a fair market deal. It was designed to compound relationships and not to maximize one transaction. We structured that market-clearing economics. A lot of private deals getting done, we believe, are done in a similar range. We focus on long-term creditworthy counterparties that can grow with us. We built to scale the program beyond just the first phase across our customers and across our financing counterparties, execution partners, supply chain partners. We took a very customer-centric approach to how we de-risk execute and give them confidence, we're not optimizing for headlines. We're trying to build repeatable partnerships because that's going to be the secret sauce in our ability to grow and scale.

Operator, Operator

So let's drill down into River Bend, then. As we look at River Bend even more holistically as well, can you update us on the progress of some of those power expansion discussions with Entergy? And how is construction tracking?

Asher Genoot, CEO

Look, 2026 is 1000% going to be about execution and delivery. That's going to be the theme of the market, in my opinion. We're super high for its construction right now, it's tracking according to plan. We have tight coordination with Jacobs Engineering and Vertiv. Long lead time procurement is progressing and getting manufacturing delivered. Our customer engagement is really high with Fluidstack and Anthropic. We have many working sessions a week, multiple stand-ups on-site in their offices, and really strong collaboration as we move towards delivery. The 1 gigawatt expansion plan at River Bend, the power is there. It's not about if, it's about when. Now, we're optimizing on delivery time lines and cost scenarios in terms of collateral upfront to make sure the rate base doesn't get impacted. We're working through different structuring paths to maximize efficiency and speed, meanwhile solving for what we're looking for and also what Entergy is trying to solve for, for their constituents. The focus now is simple: execute, deliver, and de-risk.

Operator, Operator

So how should investors think about long-term expansion opportunity at some of our other sites like Corpus Christi? Obviously, we've been seeing an uptick in pushback with respect to data centers getting done. We've seen new rules proposed out of ERCOT. How should investors think about the expansion?

Asher Genoot, CEO

Something I've been saying for years is we're building an energy infrastructure platform on digital infrastructure. Our edge is power-first development. Oftentimes in markets that others overlook, and so our Corpus Christi site that we announced, we have an approved interconnect in ERCOT that’s increasingly valuable. It was put in before all of these recent changes in batch studies. It was put in 2023, and in a changing regulatory environment, permitted power and transmission access matter more than ever. The interconnect and the permitting matrix at Corpus Christi give us a structural advantage to be able to build quickly and get access to power rapidly. If we were using the traditional developer playbook, we wouldn't have a low redundant data center solution for Bitcoin security as our tool for our underwriting assets. We probably would have passed on this 1 gigawatt interconnect like many others did, but we didn't because we have multiple use cases of our megawatts as we develop the platform. It's similar to what happened about 1.5 years ago in Louisiana. When we first talked about the site, we started marketing the site, people thought we were crazy. I thought we were early, and I think we were right. We validated the thesis of what we believe in terms of the value and the assets in the power and how much you can get at scale in the local regulatory environments where you're building this infrastructure. You see that in Louisiana, you have other hyperscalers like Meta and AWS that announced projects there as well. I don't want the market to forget that alongside our River Bend announcement, we also announced a strategic partnership with Anthropic. They're a partner that we continue to look at opportunities with alongside other demand signals that we've built over the last two years as well. Our development pipeline has 8.5 gigawatts across various stages of development. We are energy developers first. We understand grid dynamics, regulatory shifts, and permitting realities. That gives us confidence that we can navigate the evolving environments. It's a lot easier to build a large-scale data center with the dollars we're bringing into these economies than developing Bitcoin mining facilities when no one wanted Bitcoin in their neighborhoods. We navigated through different environments. We've gone through changing ERCOT procedures for the last 5 years since we started the business. This is just another hurdle that comes along as demand continues to increase, and we need to continue to show our competitive edge in developing power assets.

Operator, Operator

So you often referenced first principles and value engineering and innovation as a core edge for Hut. Can you walk us through how Vega delivering 180 kilowatts direct liquid-to-chip at $455,000 per megawatt from scratch and our co-developed infrastructure design with Vertiv that we're using at River Bend. How do those two things reflect that philosophy and effectively position us for the future?

Asher Genoot, CEO

We think that no one has fundamentally challenged how data center infrastructure, I guess, now AI infrastructure is built. That's our opportunity. We reject the status quo. In the short term, there's a huge supply and demand imbalance that allows us to get deals done from a power perspective, and that's our competitive moat. In the medium term, differentiation will come from value engineering and infrastructure stack innovation as supply and demand reach equilibrium. Vega is a great example too because we developed 180 kilowatts per rack direct-to chip cooling technology earlier last year when NVIDIA was only at 120 kilowatts per rack. The reason we did it was to show the markets we could develop that type of infrastructure as you see on the screen here, but most importantly, we were able to develop that for $455,000 per megawatt. I'd like to tell customers when we show them the site that includes the office furniture, whiteboards, and everything and the fit-out in the data center. The reason we were able to do that is we built the site from a grass field. There were no legacy constraints. We figured out what areas of the infrastructure stack switch gears, PDUs, the rack conveying structure that we want to vertically integrate, design ourselves and contract manufacturer, what areas of the construction we want to self-perform and manage ourselves. That's what I'm really excited for. Once we can get the next couple of deals done and we get to a large multi-gigawatt platform in terms of data center, how do we think about the next phase of our competitive moat? That's around innovation and value engineering. The Vertiv partnership is a good example of our early stages into that. We co-designed the approach, the architecture with them and with Jacobs to show not only supply chain visibility, but how we take risk off of the site and speed the efficiencies of these developments, so we can have them in controlled environments for a lot of the infrastructure we're building and how we have long lead-time mitigation risk. I had the opportunity to keynote Vertiv's main event earlier in January this year because we believe that's where the edge will live in the second phase, as we monetize the power assets in our pipeline in this first phase. When demand and supply normalize, infrastructure efficiency will matter more and more, and I think we're perfectly positioned for that, and it's a story and theme that people haven't really delved under Hut today.

Operator, Operator

We're executing on large and complex infrastructure projects that require significant capital and coordination. With that kind of scale comes risk. How are we structurally mitigating down that exposure and protecting equity if things don't go according to plan? You talk a lot about being a credit guy. Can you elaborate on that?

Asher Genoot, CEO

I actually don't talk a lot about being a credit guy. Sean talks a lot about me being a credit guy. He says, for your age, you're really more of a credit guy than anything else when we talk about underwriting these sites. The experience I've gained from living in kind of the business and building infrastructure around Bitcoin has really rooted us in how we think about underwriting opportunities and building the business long term. We lived through cycles, for example, in 2022, where Bitcoin prices crashed, our only revenue; energy prices skyrocketed because of the geopolitical situation, and profits were squeezed while we had fixed debt and amortization payments. I chaired UCC committees of companies going through restructuring and bankruptcy. We were the planned sponsor of taking Celsius out of a business segment they had out of bankruptcy and turning it into a company. I learned firsthand meeting with creditors on what could go wrong when you're building in a hypergrowth environment. As we think about these projects, what I've always told the team is, if we have the privilege to sign a data center deal, that contract is a liability until we deliver and start generating cash flow. That's why we're so thoughtful in how we structured this deal. We didn't just want to announce a deal and then figure out the rest afterward. We wanted to announce a deal that had financing, execution, manual labor, steel for the buildings, long lead-time items, regulatory permitting, all secured when we announced it to the market. That was a key part of the timing and the things we wanted to get done. As we think about debt obligations, construction risk, power risk delivery, counterparty risk, those were all the things we've mitigated through long lead-time contracted cash flows, creditworthy counterparties, structuring nonrecourse financing, and really disciplined underwriting as we look at how we thought about the overall risk framework between us and our counterparties and what we were committing and what we were all receiving. Durable credit really compounds. If you want to build a business over the long term, it's about compounding value over time. I feel like over the last two years, we've done that, and we've already seen the rewards of those actions, and we're really just getting started. Another element is if you look at our balance sheet, I think we have one of the cleanest balance sheets in the market today. If you think about the debt structure that we have, we have three pieces of paper at the parent level. The first is our Coatue convertible note that we did almost two years ago. That's heavily in the money and most likely gets converted out this year. That's the only parent recourse piece of debt we have. The other two pieces are one with Coinbase, which is recourse against the Bitcoin only without parent resource, and the next one is the one we have with NextEra at King Mountain, which is recourse against our equity stake in that 50% JV on the Bitcoin mining facility and does not have parent recourse. As we really think about it with Coatue heavily in the money, we have no recourse debt on our balance sheet. We're bringing on great financing for the projects, but as we think about growth, we're considering growth in credit really well to manage these markets effectively and hopefully be able to grow faster.

Operator, Operator

So Sean is going to walk us through our results shortly. We're going to do a little Q&A with him. But let's touch on a few areas in our results. G&A, for example, what drove that this year?

Asher Genoot, CEO

Stock-based compensation aligns everyone with the long-term and long-term value creation. At the company, every single person has equity from our site-level team members to the CEO, myself and the CFO, Sean. Real ownership culture and skin in the game is something we've optimized from day one and continue to do so even in the public nature that we have. Total G&A was about $122.8 million this fiscal year compared to $72.9 million. Stock-based comp was around $57.8 million versus $20.8 million in the year prior. That has to do with a lot of the upscaling we had in the investment in our engineering, development, institutional infrastructure team. Cash SG&A only rose from $52 million to about $65 million in the year, and that includes all of those transactions we did from the carve-out of American Bitcoin to the deals we've done. Our belief is that we're building the team and investing in it for the future financial profile of where we're going. A lot of the dollars we're spending are on growth, not on managing the current business, not even executing River Bend, but on the growth of all the sites we're looking to execute and commercialize. We don't believe you can scale a multi-gigawatt infrastructure platform without building and training that team in advance. We are rightsizing for where we believe the company is going.

Operator, Operator

So before we wrap up with you and move on to Sean, is there anything we haven't covered that you want investors to keep in mind. Specifically, there's been a lot of internal discussion around moving from simply building infrastructure for AI to actually building infrastructure with AI. How are you thinking about that evolution? And why is it strategically important for us over the long term?

Asher Genoot, CEO

Yes. We're sequencing the business pretty deliberately. The next layer of our advantage is technological convergence. So Phase 1, which is about 1 to 2 years in my mind, is locking in the right deals, establishing durable counterparties, building the right financing framework, and monetizing our power capabilities. Phase 2 in the 2- to 5-year range is value engineering the infrastructure stack, driving costs down per megawatt, improving speed, efficiency, and repeatability. Phase 3 is more in the 5- to 10-year range, being at the forefront of how AI and robotics reshape infrastructure development. We're not just building infrastructure for AI; we have to be building infrastructure with AI and leveraging it to change how we think about charging our innovation cycles. We're deeply thinking about how AI integrates into our business. From a workforce perspective, increasing productivity across the organization, and from a design and engineering perspective, hundreds of thousands of engineering hours go into these builds. AI is fundamentally going to change the way that work can be done. We can model, simulate, and optimize every variable for capital deployed. In the build process, we're in the early stages of exploring robotics and automation embedded directly into construction workflows. This is not a distraction. It's an awareness of where the technology curve is heading, and we intend to meet it at the moment of real inflection when it meaningfully changes how infrastructure is built at scale, and that will be the next compounding layer of advantage after the first two I spoke about earlier in this call.

Operator, Operator

That's super exciting. With that foundation set, 2026 shifts us from prove to scale, right? So what should investors look to from us moving forward?

Asher Genoot, CEO

2026 is about execution and delivery, full stop. Converting the pipeline to additional contracted revenue, advancing power origination, delivering River Bend on time and on budget, and maintaining capital discipline, no trend chasing. The foundation is built. Now we execute and scale.

Operator, Operator

All right. Now let's move on to Sean. Sean, let's walk through the results. Maybe we can start with what defined fiscal 2025 performance?

Sean Glennan, CFO

Yes. Thanks, Sue. I think there are two main things that really define fiscal 2025. First, I want to talk about margin expansion and operating leverage, and second is bottom line results. On the first topic, revenue grew 45% to $235.1 million, driven primarily by our compute segment, while the cost of revenue grew by 24% to $107.8 million. This resulted in gross margin expansion from 47% to 54%. I'd also like to highlight sequential Q4 2025 over Q4 2024 results where revenues grew by 179% and gross margin expanded from 36% to 60%. Each of these data points highlights enhanced operating leverage in the business. In other words, the foundation is sound, and what we've set in place will compound over time. Next, moving to bottom line results: Net loss was $248 million, and we had an adjusted EBITDA loss of $135.4 million, compared to net income of $331.4 million and adjusted EBITDA of $555.7 million in 2024. Importantly, I think to note, that swing was largely due to a $220 million primarily unrealized mark-to-market loss in 2025 of our Bitcoin stack versus a $509.3 million gain in the prior year.

Operator, Operator

Now let's walk through segment by segment. Can you walk us through the power digital infrastructure and then compute segment results?

Sean Glennan, CFO

Absolutely. In our power layer, revenue was $23.2 million versus $56.6 million in 2024. Cost of revenue declined to $20.5 million from $21.5 million in the year prior. The revenue decline reflects the termination of our ionic digital agreement in managed services. This was somewhat offset by increased revenues in our Far North segment due to increasing power market tightness, which I think is kind of some of the fundamental underpinnings of the business. It’s showing up in other places, too. On digital infrastructure, revenue was $9.6 million compared to $17.5 million last year. Cost of revenue declined to $8.9 million from $15.6 million last year, and margins improved sequentially as Vega entered commercialization, and we transitioned to colocation-based payments from American Bitcoin. Finally, compute: This was the real growth engine. Revenue more than doubled to $202.3 million from $80.7 million the year prior. Cost of revenue increased to $78.4 million from $45 million in the year prior. This was driven by infrastructure upgrades, higher deployed hash rate, and a full year of steady-state operations of Highrise AI, which added $7.4 million year-over-year. Important to note, as we talk about operating leverage here, segment margins expanded from 44% to 61%.

Operator, Operator

Now let's get into capital structure and strategy. How are we thinking about our capital structure evolution?

Sean Glennan, CFO

Yes. I think 2025 was an incredibly important year for capital structure evolution. Spinning out our Bitcoin mining business through our American Bitcoin subsidiary shifted us from a very high cost of capital, high CapEx cyclicality business to a much lower cost of capital business with a lot more focus on infrastructure, low cost of capital, lower risk, and longer duration.

Operator, Operator

Okay. And then heading into 2026, what are some of your top financing priorities?

Sean Glennan, CFO

Yes. I think about this as what I spend most of my days thinking about. Looking into 2026, there are four main things I'm really focused on. One is protecting shareholder value through disciplined equity use. Two is minimizing enterprise risk; three, diversifying liquidity sources, including private markets; and then four, maintaining a strong balance sheet that allows for strategic flexibility and a path towards an investment-grade rating. One important note is we continue to evaluate all financing options, and we say no to a lot of things. We're not just trying to get capital wherever it's available. We're looking for the lowest cost of capital. I probably get 10 things across my desk every day, most of which I say no to because we want to drive that and innovate on capital structure rather than just following the market.

Operator, Operator

So then, to wrap up, how would you, as the CFO, summarize our position heading into 2026?

Sean Glennan, CFO

Yes. It's amazing to think relative to when I joined 1.5 years ago where we are entering 2026. We have greater scale, both from an exahash market cap and future cash flow position. We have improved margin durability, which I think the numbers speak for themselves, and declining cost of capital. I think the project financing we're working on with JPMorgan and Goldman Sachs is indicative of that. That's the lowest cost of capital that anyone in our sector has raised ready to support AI infrastructure growth to date. I think, as I mentioned in your first question, our capital structure is aligned with long-term value creation. Those are the things setting us apart, and it's very exciting to be in this position. We feel grateful to our shareholders for entrusting us with their capital as we go into 2026. With that, I think we'll go into Q&A.

Operator, Operator

Yes. Let's go into the Q&A here.

Asher Genoot, CEO

Well, Sue, as we look at some of the questions, we want to shake things up this earnings a little bit and go with a video style, so we'll get feedback from our investors if they enjoy today or not. If we keep the video format, I want to be able to hear people when they ask questions. We talked a bit about that. As we set up this first structure, Sue is going to read the answers people are submitting. We weren't able to allow for that with the platform today, but if people like this current format, then I really want to hear them and see them if possible. We'll look at that for the next earnings calls. Hopefully, you guys enjoy the new shake-up here and approach this from first principles as well. Our goal was to use this call to have you really get to know us, get to know how we think about the world, how we think about problem-solving. You can look at our financials and our business through our public filings, but the purpose of this call, when we broke it down from first principles, was speaking to our shareholders in a direct, honest, transparent way, and we hope this new format helps drive closer to that as well.

Operator, Operator

Okay. So let's get into it. So from Greg Miller at Citizens, will the company be defining what portion of its pipeline will be allocated to Bitcoin mining and what percentage will be allocated to HPC as the two represent very different value propositions?

Asher Genoot, CEO

That's very fair. If we look at our existing capacity under management, in the gigawatt that we're managing today, we have 300 megawatts of power generation that we've told the markets that we're selling to TransAlta, and we have closed that transaction. We have 700 megawatts of compute that currently support American Bitcoin. In our capacity under construction, we have 330 megawatts of utility that's River Bend Phase 1. We have a multi-gigawatt pipeline as you go further and further up the development cycle. Currently, the core focus is converting those sites for AI use cases. Having Bitcoin as an alternative use case allows us to continue to develop confidently in building the substations on the land we acquire in interconnections, knowing that we'll have a consumer there in all scenarios rather than just having risk development capital. I think that's a unique edge that we have. Our key focus around our current full development pipeline is AI utilization and development, and we're seeing more and more focus on just power at scale and location really being a lower factor in that.

Operator, Operator

Okay. So George Sutton from Craig-Hallum, can you give any detail behind the $163 million deposit for future sites?

Sean Glennan, CFO

Sure. As we look at developing some of the future sites, we have lots of land options. We're also procuring long lead time equipment at some of these sites. I don't want to get into detail as to how much dollars are for which sites; that would give away some of our secret sauce in this very competitive industry. But effectively, those dollars are allocated towards that.

Asher Genoot, CEO

One key thing to know on how we approach development, historically and moving forward, when we think about risk dollars out there, whether it be land options or developing, we've historically been very, very low upfront until there's real feasibility. A big portion of those long lead time items are malleable pieces of equipment we can allocate specifically around high to medium voltage breakers at the substation, different transformers once we set things down to 34.5 kV. Malleable infrastructure that we can allocate across multiple campuses. The investments we make at the early stage of development are really lower in terms of development, unless it's a kayak payment or an infrastructure upgrade that we lock in the power. As we see collateral payments increasing, we're also looking at other project-level financing and balance sheet borrowing we're looking at to drive down our cost of capital.

Operator, Operator

All right. So from Brett at Cantor, you guys effectively set the market with your Fluidstack and Anthropic deal. Can you talk about how pricing has changed since then? Do you think the next deal will see a step-up in economics?

Asher Genoot, CEO

I don't think we set the market. I think the majority of transactions in the market are happening in the private markets. Everyone is focusing on public companies, but feels like 80% plus of the transactions are happening in the private markets with kind of the private equity funded development platforms on the data center side. I think our deal was fair. It was structured appropriately. As we continue to talk to customers about the deal and deal economics, we think these are market terms in the private markets, and we've held ourselves to the standard of a blue-chip data center development company. I've brought in the partners to validate that thought process and that approach as well.

Operator, Operator

Okay. So from Stephen Glagola at KBW, a recent job posting pointed to a potential scale-up of the Highrise AI GPU platform from roughly 1,000 GPUs to 20,000 GPUs. Can you provide more detail on your growth plans for the Highrise AI cloud business? And how do you envision scaling that trajectory?

Asher Genoot, CEO

Hut 8, the parent company, builds in the power layer and builds in the digital infrastructure layer, front meter, behind-the-meter interconnects; obviously, we own power generation, and we build digital infrastructure on top of that, i.e., the River Bend campus we announced. In a lot of these deals, there's an opportunity where we can fund the compute as well. Funding computing and ASICs on the American Bitcoin side GPUs on the AI side have fundamentally different cost and risk profiles, which is why those businesses are separate companies, ABTC being a public company now on its own, and then Highrise still being a private company that is growing. What's unique about Highrise? We've been quiet about it because we've been building the foundation of that business. It's not just the story that we're managing over 1,100 GPUs; the story is we've built a cloud network and software stack. We offer bare metals with multi-tenant solutions. We announced this in one of our press releases in Highrise, but the current CTO of that business ran AI in the IDF for a decade and a half. Looking at different opportunities, there are opportunities in these data center deals where Highrise can provide the financing around the GPU stack, services and technologies that I can build on top of the chip stack. Highrise is our new cloud business. We haven't spoken much about it because we are more about talking about things when they come into fruition rather than what’s on the way. We are building the company and hiring people to the place we are going, which is where you see a lot of investment in talent into the business we are going to build and scale.

Operator, Operator

Okay. So another one from George here that I really like because I don't think we talk enough about this in the market. Anthropic is a major disruptor in the space. How important is our existing relationship with them as part of the Phase II and Phase III opportunities?

Asher Genoot, CEO

They're great. They're very open to thinking about things from a first-principles approach. It's not a company where we have to do it this way just because. It's a company where we can talk about what we are trying to solve for and what is the best way to be able to solve for that. If you think about Phase 1, it’s about giving additional capacity and getting additional power converted for our customers. Phase 2 is driving down costs with really thinking about value engineering. Value engineering is two ways: one is how do we engineer and more efficiently drive costs down for our existing infrastructure stack; the second is how do we think about the demands that a customer has, which is why we have Highrise, to understand the full stack from electron to compute, from megawatt to token. By understanding that, we can optimize the infrastructure to support what’s really needed; we can have open discussions, and discussions with Anthropic have been great in terms of solutions and malleability over how things are built to get to the final outcome. Last, in terms of AI and robotics, obviously, they're at the forefront of building the technologies that support all industries. We are excited to continue to build those relations and compound, but I'm really thrilled for Phase III. We have to build Phase 1 and 2 to have the privilege for me to work on Phase III. This year will focus on Phase I and really scaling up the data center platform.

Operator, Operator

So I've got one here from Kevin Dede at H.C. Wainwright. Trump in his State of the Union last night asked big tech to commit to building their own power. How do you think your customers, partners and Hut 8 will react should it become law?

Asher Genoot, CEO

It's a natural progression of where things are going. If we think about the overall sentiment and what should happen right now is when a data center is built, it should be net positive to the community and the environment and the actual energy grids that you're impacting. When we think about sites like River Bend or Corpus Christi, we pay for the system upgrades that pull the power to where we are. We commit to the capacity on the energy side. That's kind of table stakes in my mind in the world we're developing today, and smaller utilities have become infinitely more sophisticated on that, which is to see kind of collateral coming in. Some places have gotten over-indexed and will kind of come back to the mean. But in general, that's the sentiment. As power generation gets constrained, more demand comes onto the grid. You'll see more often not just island generation or bridge generation where you kind of wait for the interconnect, but when you're building a load asset, you want the redundancy from the grid. There is value in that. When building a generation asset, you want the grid and the demand that's there for the power. More often what will happen in the markets is people will bring load and people will bring generation. We’re trying to have a net equal impact on the grid, interconnect everything in the long term, so you have power generation increasing in the grid, and you have load increasing. A lot of that will be financed and capitalized through the demand of the end customers and users, and we think that's the best way to compete in the AI race on the global markets.

Operator, Operator

So from John Todaro at Needham, can you walk us through where you stand on the OSA negotiations with Fluidstack and more broadly, give us an update on construction cadence? How many data falls in the initial phase? Are you seeing any supply chain or contractor bottlenecks? Maybe we can talk about where we're at on the procurement side at River Bend.

Asher Genoot, CEO

Sure, happy to do so. When we announced the deal, everything we locked in from people, contractors, long lead time items, equipment. All of that was locked in. There was nothing open when we announced the deal at the end of last year. On the delivery and the execution itself that we mentioned and guided towards, in the beginning of Q2, we’ll have the first data center coming online, and then we’ll have a data center coming online every 60 days thereafter. There are four data centers in this data hall. As we think about the actual construction right now, everything is going really, really well. People are excited. There's a lot of local talent in Louisiana because of the heavy industry that was there before. That’s great. Jacobs has been a great partner. Vertiv is fully cranking on the long lead time items they’re bringing and procuring for this project. Overall, from a constructability and delivery perspective, we feel very, very good. We gave ourselves a healthy timeline to deliver this as well. We’re not crunching every single thing. We put buffer in. We hope to deliver earlier if we can. From a financing perspective, things are going very well as well. We announced we’re targeting 75% to 85% LTC at a SOFR plus 225 rate. We’ve recently improved that to 90% LTC at SOFR plus 240 to account for the increased loan-to-cost ratio. More project financing we’re getting on the projects. Something Sean and I like to talk about: even when a deal is done, we still like to further improve and figure out how to make it better. Overall, in terms of delivery, we feel very, very good. We’re currently in active negotiations on the OSA in terms of operations and delivery, but we have some time until we actually are operating in the campus, so working through those kinds of contracts now as well.

Operator, Operator

So maybe just to quickly piggyback on that from one of our new friends, Robert Boucai at Newbrook. In considering future deals, do you require credit enhancements as with Google on the Fluidstack deal? Would you be willing to have one of the LLMs be a counterparty? How much of a gating issue would this be?

Asher Genoot, CEO

I think overall, it’s really thinking about our overall platform we’re building and the exposure we’re taking on investment-grade counterparties on non-investment-grade counterparties and everything in between. Thinking about developing our platform, we want to make sure that the cash flows we have and we’re projecting funding the growth and expansion of our business are durable and reliable. Obviously, that affects cost of capital and financing and LTCs as well. As we look at future growth opportunities, we’re obviously optimizing towards investment-grade counterparties, but it’s not binary; it has to be this or that. It’s around risk allocation. With everything everyone’s investing in, including shareholders on this call, it’s about risk allocation. What percentage of your platform is on high growth, high-risk companies with a ton of upside? What percentage is on stability on the platform as well? It’s not binary; there’s always a place in the portfolio for high-growth companies as well, but it’s about percentage exposure we have to them. Every time we announce a deal, we’ll walk through the thought processes on those. Right now, we continue to focus on investment-grade counterparties we’re financing towards.

Operator, Operator

Okay. So a question from Mike Grondahl at Northland. Can you describe the demand environment and how it’s evolved over the last 90 days for HPC? Obviously, a few new dynamics have transpired in the market. How has that evolved in terms of your conversations with customers?

Asher Genoot, CEO

It's interesting to note that last year around this time, the news about DeepSeek created a lot of fear regarding vanishing demand. The markets were nervous, leading to a significant drop in Microsoft's stock. However, during that period, we actually observed ongoing demand and positive signals. There was a notable increase in demand towards the end of last year. Currently, with the rise of Agentic AI and a situation where Mac Minis are sold out nationwide, we see that the utilization of our cloud services is at unprecedented levels. The applications and use cases are steadily expanding. This growth will inevitably require increased computing power. Our current circumstances differ from last year as we have established much stronger relationships with various partners over the past two years. We've engaged in extensive negotiations on contracting and collaborated on engineering with multiple partners. For instance, River Bend was just one entity we worked with over a year and a half, navigating through various iterations. Those partnerships have never been stronger, and our credibility permits us to have open discussions about their long-term capacity needs, how we fit into that picture, and how we can support them. We're currently focusing less on the stock market and more on understanding our customers' demands and building competitive advantages, as that will drive our business growth in the long run. While the market may be filled with uncertainty, we are experiencing robust demand, and companies continue to expand, as evidenced by Meta and AMD's recent announcement for additional capacity. Overall, demand remains healthy. A significant trend we're noticing is the increasing constraints on power; utilities and transmission operators are reevaluating their strategies. This is a positive development, as many developers may lack the resources or capabilities to establish sites and are primarily looking to secure interconnect agreements to transfer them to buyers. Our access to financing inquiries is substantial, as we receive numerous propositions to consider sites for potential acquisitions. Clearing out some of the uncertainties could improve the development queues. In terms of land development, we are witnessing a divide where some areas resist having such developments nearby while others welcome them.

Operator, Operator

Yes. Agree, we support any initiative that helps trim some of the fat in these queues, and also education is key in markets where we see stakeholder pushback for data center development. Okay. From Greg Lewis at BTIG. What I'm sure a lot of people want to know is any update on the power that is under exclusivity and the steps and processes needed to move that into development? It doesn't seem like we saw a lot happen in the current queue.

Asher Genoot, CEO

Yes. We’re currently working through it. We obviously have a substantial capacity in development to commercialize and sign agreements because if you think about our stages in the pipeline, we progress from capacity under diligence, where we have a large energy origination development team that’s out hunting, negotiating, and looking for opportunities that make sense. We get into capacity under exclusivity. That’s where we're investing more dollars from a team perspective and resources perspective. In those scenarios, we have exclusivity, land options. We’re investing in legal resources, preconstruction resources, high-voltage transmission engineering resources. From there, we go to capacity under development when we're actually buying the land, locking in contractual agreements on power, and putting in the kayak payments or any collateral obligations. We have a strong amount of megawatts over 1 gigawatt in capacity under development that we’re working to commercialize, and capacity exclusivity will be following that. Timing capacity and exclusivity, we have exclusive access to those opportunities, whether it be land, power, or both. In that scenario, why go and spend the money if you still have option value, right? Work on commercializing the things you already spent money on.

Operator, Operator

Thank you. So from our friend, Chris Brendler at Rosenblatt Securities. On funding River Bend CapEx, any early read on the project level financing deal given the recent volatility in the market? How do you view your Bitcoin holdings as a potential source of funding for the equity portion?

Asher Genoot, CEO

We feel very good. The equity as of now is already fully funded because we’ve had to fund the projects, our deposits with Vertiv, and so forth. When the project financing completes, we’re going to get a multi-hundred million dollar cash out of the transaction, then we’ll fund our 10% ongoing. As we mentioned, we went from 85% LTC to 90% LTC in the financing. We’re pushing aggressively towards closing. JPMorgan and Goldman are committed; they wouldn’t have given us quotes and included us in the press release unless they saw it was just an idea. We’re very excited and committed, and we have connectivity at the highest levels. I’ve had lunches with the CEOs of the firms. So we’re thrilled to have partners not only for this project but for future projects to replicate this program. From a Bitcoin perspective, our balance sheet and our Bitcoin on the balance sheet at the beginning of last year were core to our execution throughout last year. Using Bitcoin on the balance sheet to tell customers, don’t worry about our ability to finance, was really important. This year, Bitcoin on the balance sheet is not our focus. It’s just an asset on our balance sheet. The reality is we’re going to remove Bitcoin exposure on our balance sheet as we move forward. How we do it is what we’re focused on right now. Our exposure will be through the equity ownership in American Bitcoin. We’re very good at creating value, and our focus is on driving down the cost of capital and continuing to create value by building businesses.

Operator, Operator

Great. Okay. So let’s talk about we touched on this a bit. Why don’t we discuss where the ...

Asher Genoot, CEO

We have like 31 questions in the queue. I think we’ll reduce that number when we get people to come on stage with us next time.

Operator, Operator

Okay. We talk about the importance of our energy origination team and diversifying the pipeline, not being overweight to a single market. Can you and maybe Sean talk about some areas where we are still finding pockets of opportunity? For example, in a previous conference, we discussed how we were interested in studying development in Pennsylvania. Maybe just talk about some areas we’re looking at outside of ERCOT.

Asher Genoot, CEO

We’re looking across the whole U.S. Every area, as we mentioned, power and land in a regulatory environment that allows for building this infrastructure at scale are the key pieces in building. Fiber has been less of a bottleneck; we've been able to bring fiber to a lot of campuses we’re developing and building. It’s following the power, taking a first principles approach to where there’s power. Using River Bend as another example because I think it’s our first fully vetted case study, we can do the same for future sites that we announce. That project was around the transmission lines. It was around the generation near that campus. We came together and pieced together multiple pieces of land that were held for generations as hunting properties by people. We built this 3,000-acre opportunity to construct a large-scale mega campus. We’re looking for similar characteristics as we explore the United States and its ability to scale power, build with a friendly regulatory environment that wants our projects there, and see the impact and benefits we can bring. We can have talent to execute and build projects at the speeds we’re looking for. The reason we’re scaling is we’re increasing the breadth and depth of our origination pipeline across the whole U.S.

Operator, Operator

Awesome. So...

Asher Genoot, CEO

We have many more projects similar to River Bend that have been publicized but not disclosed to the markets due to local zoning and panel requirements. Keep an eye on Hut 8 for updates.

Operator, Operator

Okay. We are coming up on the hour here. So we're going to do one more...

Asher Genoot, CEO

We have like 31 questions in the queue. So I think we'll reduce that number when we get people to come on stage with us next time.

Operator, Operator

That's right. So we talk a lot about the importance of our energy origination team and diversifying the pipeline of not being overweighted to a single market. Can you, and maybe Sean, talk about some areas where we are still finding pockets of opportunity? For example, in a previous conference, we talked about how we were interested in studying a development in Pennsylvania. Maybe just talk a little bit about some of the areas that we're looking at well outside of ERCOT.

Asher Genoot, CEO

We are looking across the whole U.S. Every area in the U.S., as we mentioned, power and land, and a regulatory environment that allows for building this infrastructure at scale is key in building. Fiber has been less of a bottleneck; we’ve been able to bring fiber to many campuses we’re developing.

Operator, Operator

Awesome. So...

Asher Genoot, CEO

In summary, we’re looking at the fundamental dynamics of power and land and aiming for areas where we can build effectively based on regulatory environments that are supportive of infrastructure development.