Today. Thank you for standing by. Welcome to Ionic Digital's Investor Day. At this time, all participants are in listen only mode. After the speaker's presentation, there will be a question and answer session with sell side equity research analysts. To ask a question during the session, you will need to press star one one on your telephone. You will hear an automated message advising your hand is raised. To withdraw your question, please press star one one again. Please be advised that today's conference is being recorded. I would now like to hand the conference over to your speakers today. Richard, please go ahead. Good morning. Welcome to Ionic Digital's Investor Day. Before we begin, a brief reminder. Today's presentation contains forward-looking statements, including statements about our contracted revenue, growth pipeline, capital plans, and the proposed direct listing. Actual results may differ materially. These statements are subject to the risks described in our registration statement on Form S-1 filed with the SEC, which you should read in full. We undertake no obligation to update them. This presentation also references adjusted EBITDA and other non-GAAP measures. Reconciliations are in the appendix. With that, I'll turn it over to our Chief Executive Officer, Andy Stewart. Good morning, everyone. Thank you for being here. I'm Andy Stewart, CEO of Ionic Digital. I know these presentations can be a bit analytical, even dry. Hang with us because what you're going to hear is a great story and really excited to tell it to you. Page three lays out how we'll spend the next 45 minutes, who Ionic is today, the growth already embedded in the footprint we own, where we see the market heading next, how we're governed as we enter the public markets, and the financial foundation under it all. First, let me introduce the people you'll hear from today. On page four, you will see that we are a data center-first leadership team assembled over the past year, each with decades of experience building and operating infrastructure at scale. Chris Hickman, our CFO, brings two decades across digital infrastructure and energy, including leading an energy infrastructure IPO and leadership roles at Crown Castle. Antonio Piraino, our Chief Strategy Officer, spent two decades at the intersection of data centers, cloud and AI, including as a CTO and a leading advisor at the Uptime Institute. Mark Lambourne, our Chief Development Officer, has spent more than 30 years developing global data center platforms at firms like Digital Realty. Richard Carson, our General Counsel, brings 30 years of energy and capital markets legal experience across multiple public listings. We were brought together over the past year for one purpose: to complete Ionic's transition into a data center-first company. Let me start there, with who Ionic is today. Turning to page six. We are a digital infrastructure provider for high-performance computing and artificial intelligence, that's a very different answer than you would have received just 18 months ago when this company was primarily focused on mining Bitcoin. We're not the first Bitcoin mining platform to pivot to HPC and AI data centers. Many making this pivot are crypto operators stepping into the data center business, while we are data center operators repositioning a crypto asset, and I believe that distinction shapes everything you'll hear today. Ionic was formed in 2024 to acquire the crypto mining assets of Celsius Mining out of bankruptcy. In mid 2025, our board looked hard at the power assets we owned and asked, "What is the highest and best use?" The answer wasn't mining Bitcoin. It was hosting high performance computing for AI, the largest source of infrastructure demand in the market. We made the call to reposition the entire platform for it. In a market where a lot gets announced, what truly matters is the ability to deliver. Deliver energized capacity and put it under contract. That's why we call ourselves a power now provider. We have market-ready assets producing revenue today and a pipeline of sites with secured power in attractive markets. Page seven gives you a snapshot of Ionic today. We control market-ready utility capacity totaling 822 MW of power, concentrated in the high-demand West Texas market with more than 300 MW energized today. Our anchor contract is a 10-year triple net lease with Nscale, roughly $2.6 billion of contracted revenue, it's backed by an NVIDIA guarantee for the first five years, Microsoft, importantly, has strategic interest in the site. Delivering on our strategy is a data center-first leadership team that has built and operated infrastructure at scale through every prior shift in computing. We believe we're doing it from a position of real financial strength, debt-free with significant liquidity. We recently raised $400 million from a group of leading institutional investors, which only reinforces this strength. I believe this combination is what sets Ionic apart. Contracted revenue, embedded growth within a footprint we already own, an experienced team, a clean balance sheet, a clear forward strategy. Few companies making this transition can point to one or maybe even two of these. We have the full complement. Turning to page eight, before we look ahead, I want to take a moment to recognize how we got here and acknowledge the shareholders who have been with us from the start. Many of them came to own our shares through the Celsius reorganization after four-plus years of uncertainty and frustration. Reaching a transparent public market with real liquidity is a meaningful step, one we're mindful of reaching on behalf of people who have waited so long for it. The best thing we can do for those shareholders, for all our investors, is execute on what's ahead, that's exactly what the rest of today is about. To level set on our industry and the opportunity, I think the most important thing to understand about this market is that demand has only ever moved in one direction. Each wave of computing, the cloud, model training, now inference and agentic AI, has required more power than the last, we don't see that changing. Wall Street research has U.S. data center power demand roughly doubling to about 66 GW by 2027. The difference this time is that supply can't keep pace. Newmark estimates the gap between demand and anticipated utility supply at roughly 50%, and that's not a gap that closes in a year or two. When demand doubles and the power to serve it falls half short, the binding constraint stops being announcements and becomes physics. Interconnection queues that run for years, transformer and equipment lead times, permitting. You can see it in the pipeline. The disclosed U.S. data center development pipeline has reached 241 GW, but only about a third of that is actually moving. The rest is waiting because of those exact constraints. That's what a power shortage looks like on the ground. Those constraints don't ease up from here. They compound. Every one of them does the same thing. Over time, it moves value towards whoever controls energized power in the right markets. That's the position Ionic has built, and it's a big part of why we secured our triple net lease with Nscale. When they came looking, we had 234 MW of low-cost, grid-connected power, energized and available in the right market with a credible path to 700 MW. In a business where power is the scarce input and most capacity is years away from delivery, that's what let us move when others couldn't. We're not waiting on the demand. We're not waiting on the power. We have both today. Page 10 shows our assets sitting in the middle of the West Texas data center boom. The named projects around us, Stargate, Crusoe's Microsoft AI Factory, Meta's El Paso campus, a Google-anchored site, and others, make this one of the most concentrated clusters of hyperscale and AI investment anywhere in the country. It's worth being specific about who is converging here. The projects on this map are anchored by the most demanding buyers of the compute in the world, names like OpenAI, Microsoft, Meta, and Google. When those names commit multi-gigawatt capital to a single region, it tells you the demand for power-ready capacity in this market is real. It's worth asking why the largest operators in the world are converging on this particular region, and the answer comes down to total cost of ownership. West Texas offers among the lowest all-in costs to build and operate large-scale compute anywhere in the country, and the power, the single largest operating cost, is the biggest driver of that. When you're deploying gigawatts, the price of power determines the economics, and this is where those economics are most favorable. Our assets sit inside that same low-cost basin and provide the source of that advantage. Our power at Ward County comes from existing grid interconnection, among the lowest cost power available, and runs around $0.05 / kWh. That's a meaningful structural edge. A number of the projects on this map are turning to off-grid or self-generated power because grid capacity at scale is hard to come by. We already have it at a cost that's among the most competitive in the country. The difference is that much of what you see here is still announced or under construction while ours is energized and generating revenue today. There's also a tax dimension to that cost advantage. Our Ward County site holds qualified data center status with the state of Texas, and under our lease, the associated sales and use tax advantage is passed through to Nscale. To put rough numbers on it, the exemption covers the state's 6.25% sales tax on qualifying data center equipment. On the order of $60 million of savings for every billion dollars of qualifying spend. On top of low-cost power, our tenant captures a meaningful tax benefit that lowers its all-in cost even further, which makes a site like ours more competitive for exactly the demand on this map at no cost to Ionic. Importantly, we don't have to build our way into this market. We're already in it with the power and the economics that will matter for years to come. That surrounding demand is exactly what our uncontracted capacity will sell into. The 377 MW we expect to bring to market at Ward County and the Midland sites we're converting. We're not hoping this demand shows up. It's already here at scale around the assets we already own. With that backdrop, let's look at our existing footprint on page 11. Our flagship asset in Ward County, where we own 136 acres with capacity to scale to 700 MW. Today, we have 234 MW energized and contracted to Nscale on a 10-year triple net lease with total contract value of approximately $1.9 billion. We also control 112 MW of grid power capacity across four sites in Midland. Those sites are mining Bitcoin today while we complete the pre-development work required to convert them to HPC and AI data centers. The Nscale lease is the anchor for everything you'll hear next. Before I get into how it's structured, let me pull our whole footprint together into one picture because it frames everything that follows. Here on page 12, the sites I just walked you through roll up into a single number and a single trajectory. 822 MW of market-ready or near market-ready utility power. Start at the left, where we have 234 MW energized and under contract at Ward County. From there, the growth builds in clear secured steps, all within our existing footprint. The incremental 89 MW at Ward County is expected in the second half of 2027. The remaining 377 MW that takes Ward County to its full 700 MW and 112 MW across our four Midland sites, with another 10 MW to come, all positioned to convert to AI workloads. You arrive at 822 MW of total utility power with 346 MW of current capacity today. I'd like you to take two things from this page. First, the composition. This is grid power we already control on a clear and staged path to energization, secured capacity inside a footprint we control. The gating items from here are utility build-out and grid energization. Mark will speak to how we're working through those, but we believe the capacity itself is very well-positioned. Second, the shape of our growth. You can see it building in layers across the page. It maps to how we run our business. A contracted base producing revenue today, embedded growth already sitting inside this footprint, beyond that, the next frontier of demand this kind of capacity is built to serve. Those are the three levers I'll keep coming back to. With the full footprint in view, let me turn to the asset and the contract at the center of it, starting with the company on the other side of that lease. Some of you know Nscale well. For those who don't, I think it's worth a moment because in a contract like this, the strength of the lease ultimately comes down to the strength of the counterparty behind it. Shown here on page 13, Nscale is a full-stack, vertically integrated Neocloud. They deploy the GPU infrastructure. They run the orchestration layer. They deliver AI cloud services end-to-end. They recently raised $2 billion at a $14.6 billion valuation with a group of investors behind it includes NVIDIA, Dell, Citadel, Jane Street, and Fidelity. A customer pipeline that continues to expand. This is a serious, well-capitalized partner and one I believe that will be a durable counterparty over the life of the lease. The strength behind this contract doesn't stop with our tenant. NVIDIA guarantees the first five years of rent. Microsoft has a strategic interest in the expansion capacity. When we look at the demand underpinning Ward County, it isn't resting on any one name. It runs through the entire chain behind it. That is a meaningful part of why we have the confidence we do in the asset. With that context on the counterparty, let me walk through the lease itself. Turning to page 14, Ward County began as a 234 MW Bitcoin operation. In October of 2025, we signed a lease to Nscale that converted that power into a 10-year, $1.9 billion triple net contract for an HPC and AI data center. Since then, Nscale has been fitting out the structures we already had in place and building full turnkey data centers on the site. The ability to scale this site to 700 MW was important to Nscale. They hold a right of first refusal on any additional power we secure here. I want to be precise on that point because it matters. It is a right of first refusal on the site, not a call option. I think it speaks to the quality of the demand behind this site that Nscale has in turn passed its right of first refusal onto its own customer, Microsoft. Earlier this year, Nscale came looking for additional power and ways to grow the site. We amended the lease to add an additional 89 MW as that capacity energizes, expected in the second half of 2027, bringing total contracted power to 323 MW and total contract value to roughly $2.6 billion. The economics are really what matter here. Base rent is $65 / kW per month, approximately $183 million of annual contracted rent on our currently available power. That rent goes cash pay in August. We secured it for about $10 million of capital. The return on capital we put into work is exceptional. Equally important is the durability of those cash flows. NVIDIA guarantees the first five years or $860 million in total. Because the lease is triple net, Nscale carries the operating costs, the taxes, the insurance, and the maintenance, not Ionic. That's what holds operating margins in the 98%-99% range. Another feature that's easy to overlook, if the lease isn't renewed, the data center infrastructure reverts to us, so the build-out being financed on our site accrues to Ionic over time. One other note on that $65 rent because it's worth being precise. Our rate is priced on gross power. You'll see other operators quote higher per kilowatt numbers, but those are typically on a critical IT basis. They come from building and operating the entire data center. Hundreds of millions, sometimes billions of dollars of capital behind them, often financed with debt. On that same critical IT basis, ours is roughly $90 at a 1.4 PUE. The rate was never the point. Under this contract, the tenant funds and builds the facility, not us. We unlock that contracted revenue without the build-out capital, and we get the revenue years earlier because there's no build cycle to wait on. The rate is just an input. What matters is how little capital sits behind it. Before I show you the growth, let me spend a moment on how we think about capturing it, because the structure we put on each site is central to the return, and it's where a lot of our discipline lives. Page 15 lays out the ways a site like ours can be monetized, ordered by how much capital and risk we take on. On the capital- light end of the spectrum is a powered shell. We develop the foundation, the walls, the roof, the floors, essentially construction through the building envelope with utility delivered to it. We stop short of the full mechanical electrical fit-out needed to actually operate the IT load. That is the tenant to complete. It's often the fastest path to revenue, the lightest on our capital, and it carries the least construction and operating risk. In a typical powered shell, we'd invest on the order of $1.5 million/ MW and target a low to mid-teen unlevered cash yield. From there, the model asks more of us and is often referred to as build- to- suit or turnkey, where we build to a specific tenant's requirement and earn a contracted yield on that cost. A turnkey build runs roughly $9 million-$13 million / MW and can generate a 15%-18% unlevered cash yield. A higher headline return than a powered shell, but for six to nine times the capital, a longer build cycle, and materially more risk. The Nscale lease sits in a category of its own. Because we monetize power and structures we already had in place, our capital was a small fraction of even a typical powered shell deal. The returns on it, frankly, are exceptional. It's not a deal we expect to replicate often. It's the kind of outcome that's possible when you already own energized power in the right market and can move quickly. We will make the powered shell versus turnkey call on a site-by-site basis, weighing the capital, the risk, and the return for that specific asset and that specific customer. We'll move up the curve when the incremental return clearly justifies the incremental capital and the risk. What we optimize for is not the highest rent or even the highest yield. It's the best risk-adjusted return on capital we can deploy. A low to mid-teen yield for a fraction of the capital at a fraction of the risk and faster time to revenue can compound more shareholder value than a higher yield that ties up 6x- 9 x the amount of capital. Having the full range available and the experience to execute on each option is what lets Ionic make the best decision for shareholder value. Let me pull our growth story together on page 17. We think about growth across three pillars. The first is our contracted base, it's where we've just spent most of our time. The Nscale lease at Ward County, 323 MW and roughly $2.6 billion of contracted revenue signed and beginning cash receipt next month. The second is embedded growth, the remaining capacity that takes Ward County to 700 MW and the four Midland sites we're converting. This is our path to the 822 MW we walked through earlier; it's embedded within our existing asset base. The third is the frontier, inference and agentic AI, where we believe the next wave of demand is heading, exactly the kind of demand this distributed power capacity is built to serve. The third pillar is where the market is going, it deserves real attention. Antonio has lived at the shift from inside the data center industry, he's the right person to take you through where demand goes next and why we're positioned for it. Antonio? Thank you, Andy. Turning to page 18, I want to spend a few minutes on where data center demand is going, because I think it changes how you should think about where Ionic is headed as well. I've spent most of my career on this exact question as a CTO of ScienceLogic, subsequently at the Uptime Institute, where the understanding of compute demand was really headed and how that impacted the infrastructure, the data centers themselves, what the requirements would be. That was really the job assigned to me. Let me tell you where I think it's now headed. 10, 20 years ago, a 2 MW facility was a serious building that was needed by all enterprises. We started to see the boom of internet applications, e-commerce applications, this massive push towards the cloud as things got more and more centralized in regional hyperscale campuses. As we entered the world of natural language processing, large language model training pushed us again to gigawatt-scale campuses in remote locations where everything was highly centralized. Models learned from massive data sets. There were huge requirements for data and storage, compute as well. That's an enormous market, it's well-served today by established, well-capitalized leaders. I want to be clear that we are not trying to compete here, as brilliant as that industry is. That architecture, the architecture of gigawatt- scale remote build facilities for training, is not the architecture that serves what comes next. We believe what comes next is inference and agentic AI, especially for the enterprise. The numbers behind that shift are substantial and becoming well-known. We've seen incredible growth over the last year in particular. Inference demand is projected to grow going forward roughly 35% a year through 2030 and agentic AI is going to be even faster still on the order of 78% growth year on year. Together, over that period, they're expected to take the data center market from around $389 billion to roughly $700 billion by the year 2030, and in some cases, some projections even further. That is the wave we're positioned for. Let me be clear about the distinction, because I think it's the part that matters most, and you can see it laid out in the table on this page. Training of data is episodic. Models learning from massive data sets with large storage requirements, compute requirements, gigawatt scale in remote locations where it's least costly and highly efficient to operate, and it tolerates a lot of latency. Inference and agentic workloads are a very different thing entirely. They're always on. They don't go to sleep. Responses are required immediately at sub-100 MW scales, close to high density environments, and they live or die on latency and the need for enterprises to have a highly secured, highly sovereign, highly governance sensitive facility. An enterprise agent making a real-time decision, therefore, cannot afford to wait on a round-trip multiple times to a remote campus hundreds of miles away. Proximity to the user, to the systems that it's interacting with is a requirement. A training campus and an inference and agentic site are not the same building. They're not the same building in size. They're not the same building in scope, location, and proximity. They're very different products. This one has to be near the enterprise. Turning to page 19, there's a second force at work here, and that's data gravity and sovereignty that I've mentioned before. Today, if we look at the world, roughly 83% of enterprise data still resides within their own customer environments, inside the business, not out inside that centralized public cloud environment. Enterprises are increasingly unwilling to write their proprietary data and intellectual property to a distant facility that they don't control, that they don't have the governance control over, especially in a highly regulated environment. What they really want is a sovereign contained environment close to the data, close to the people using it, close to the systems that they serve, especially these very highly interactive applications. It's a large and fast-growing need. We've seen this need exploding over the last year. The sovereign AI infrastructure market is projected, therefore, to reach roughly $730 billion by the year 2035. Amongst them is Nscale, our anchor tenant, which was built for sovereign-grade AI infrastructure. This is the demand that we're seeing directly in our facilities as well. This architecture for the next wave is not really the gigawatt campus out in the desert. It is the next generation of contained sub- 100 MW environments that are agile, dynamic, and distributed inside the major metro areas and key secondary markets where enterprises actually operate today and where their data exists today. That's exactly what we're targeting going forward. I want to be deliberate about how we describe it. It's not a single segment or a buzzword, but the broad shift towards distributed latency- sensitive inference, and agentic workloads in whatever form it ultimately takes, because these things are shifting all the time. We're not betting the company on one label catching on. We're positioning for the direction of the market that it's clearly moving in. The reason this is credible, we believe coming from us, is straightforward. We are data center people first and foremost. We understand power density. We understand the need for a mix of air-cooled and liquid-cooled system, the mix of CPU and GPUs, and the specific power and networking requirements of this particular era. Because we've built and operated through every power shift in computing, we believe that we are well situated to take on this next generation of construct. This isn't a new market that we're learning about. It's the natural extension of what our team has done for decades. First-mover advantage and positioning in the right cities with that operating heritage behind us is the opportunity that we believe sits in front of us. The person to give you more confidence in this regard and our ability to deliver exactly on this is Mark Lambourne. Mark has built and operated data centers through every era I've described, and he's going to walk you through how we turn secure power into energized, revenue-producing capacity and why this team is the one that executes. I'll hand it to Mark to take you through it now. Thank you, Antonio. I've spent more than 30 years developing data centers through the evolution of the industry that Antonio has just described. Andy walked you through the ways that we can deliver a site. Powered land lease like Nscale, a powered shell, or a build- to- suit. I think the most important point is that we can actually execute all three and that most operators in this space today can't pivot. They're built around a single delivery model. We can address them all. We understand the key drivers in today's market, power density, cooling architecture, permitting, interconnection agreements, and the emphasis on commissioning. That is for any given site and any given customer so that we can choose the structure that produces the best risk-adjusted return rather than forcing every deal into one model that we happen to know. The only way to have that flexibility is to have the experience across every one of those models, and that's exactly what this team offers. You can see that looks like Ward County on the image in front of you. We own 136 acres where you can see the data hall cluster and the substation on site and everything progressing. The transmission switch yard and the main substation that you can see are TNMP's utility infrastructure and not ours. While the substation and the on-site improvements are owned by Ionic. The site is presently energized at 234 MW today. Because the Nscale agreement is a powered land lease, Nscale is actually building and operating the data center itself. This means our capital obligation to deliver the contracted capacity is modest relative to a roughly $2.6 billion contract. Nscale has moved fast on their build. They've energized the first building, topped out buildings two, three, and building four's roof is close behind. Building one is actually on track to be completed by the end of August, which lines up with our August cash rent commencement, and building two's data hall is expected to be ready and be energized with power by the end of September. This is real physical progress that is going on the ground. In this area, power costs run about $0.05/ kWh, among the lowest available. As we expand and access the additional 377 MW we expect to bring to market, we believe this is going to be highly attractive to many customers. On the expansion itself, work is already moving. We've made the substation operational, we've done the improvements needed to activate the current 234 MW, and we've executed our EPC contract with Qualus. In addition, we've already ordered the three transformers that the expansion requires, and delivery is expected in December, ahead of the additional power energization date that we expect in the second half of 2027. We've also launched our pre-development planning for phase two, and we are having a real advantage in this particular situation as our pre-construction partner already has on the order of 700 people mobilized on this site across two shifts for phase one. This provides a trained workforce that knows the sites, knows the conditions, and what's needed for deployment. We intend to carry that momentum straight into the next phase rather than go rebuild it from scratch. Our next phase at Ward County gives us a real runway of near-term growth, which is shown here on page 21. The original 700 MW facility extension agreement that was executed with Texas-New Mexico Power was done so in 2021. It was studied and approved by ERCOT in 2022. We believe we're extremely well-positioned with ERCOT's new Batch Zero process, which I will cover in a moment. As we secure power, we intend to take that incremental capacity to market and will do so with the framework that Andy laid out earlier. As the incremental power energizes, annual contracted rent steps up from $183 million to approximately $251 million. To get there, we expect to invest approximately $40 million this year to upgrade the substation and get the necessary equipment to bring on the 89 MW, and then roughly an additional $24 million beyond that into 2027 to take the site to its full 700 MW, which as stated, we expect to complete by the end of 2027. $64 million of incremental capital to unlock the entire capacity. When I step back from Ward County, here's what I see. We own the power, we control the expansion, and we're converting both into long-dated, contracted de-risked cash flow with very little capital behind it. An excellent position to be in. Turning to Midland on page 22. Midland is our second asset, and it features four sites, East Stiles, Garden City, Rebel, and Stiles. Collectively, they total 112 MW of existing capacity today across about 60 acres of land, two parcels of which we own and two of which we lease through 2032. All four of these sites are generating cash flow through mining today. More importantly, every one of these sites has the land and the power to host a new data center serving inference and agentic AI workloads. We will use the same discipline data center first approach that we have proved at Ward County while continuing to mine during the development. What that means is there's no binary choice here. We actually are earning while we convert. In the near term, we have 25 MW of idle capacity at Rebel, which we added at the end of last year, which is available for both fleet upgrades or early development. We've also got an additional 10 MW coming online at East Stiles by 2027. That takes our total pipeline to 122 MW. Our conversion review work is already underway. We have launched our pre-development across the four sites at the start of June, and we are now working through site-by-site layouts, utility infrastructure review, and doing the standard due diligence that is necessary for us to pursue a phased development plan, which we plan to present to our executive management team at the end of this month. We're not waiting, though, on a finished plan before we go to market. We know that there is demand, and we're already fielding inbound inquiries from prospective tenants and working through their technical requirements. The disciplined sequencing, though, we're following is deliberate. We want these sites brought to market the right way and not the fast way. Everything I've just walked you through, the expansion at Ward County, the conversion at Midland, ultimately runs through one process, and that is to become real. In Texas, that process is ERCOT. The path every megawatt of new load has to take, and it's where deep execution experience matters most. Let me just spend a moment on it because it's where we are genuinely well-positioned. Getting a new load energized in ERCOT is a multi-step process. You request to connect, the planning group reviews and then signs off. You build while completing the required validation, and then you're cleared to energize. It's detailed and sequential, and moving through on schedule takes real expertise and knowledge. At Ward County, we're already well down that path. Our facility extension agreement with TNMP for the full 700 MW was actually executed in 2021 and approved in 2022. The contractual and planning foundation has been in place for years. Our load has further been qualified by ERCOT as eligible to operate as baseload power through the Batch Zero process, which is an important distinction. Baseload classification reflects the kind of steady, high utilization demand that ERCOT plans around. The transmission upgrades that serve our site are part of TNMP's broader build-out, not work we're waiting to initiate on our own. Here's where the expansion stands and what's next. We're completing our information package for TNMP, coordinated with Baker Botts and our engineering team, and expect to submit that package by July 10th. TNMP then carries that into ERCOT's Batch Zero classification process, and we expect ERCOT to publish its conclusion and classifications in early August. As already noted, we are already included in the Batch Zero baseline calculation, which puts us in an extremely strong position. Based on that, we anticipate confirmation of the additional 466 MW of utility power that takes this Ward County asset to its full 700 MW, with delivery expected by December 2027. That timing, just so everyone is aware, is gated by transmission infrastructure as there are two substations presently under construction by Oncor and TNMP. That is their responsibility to deliver, not anything on our side. Finally, let me close on how we grow beyond our current footprint into the inference and agentic opportunity that Antonio has already described. We have developed an acquisition strategy that we believe will drive long-term value and will review and acquire new assets opportunistically. Every site we will pursue has to clear the same high bar. It has to be in the right place, location, the major metros and enterprise hubs where this demand lives. It has to be the right size, sub- 100 MW, and more importantly, below the large load threshold in each geographical area we focus on. It has to have a credible near-term path to energization, ideally with a power contract interconnection agreement already in place. It has to come with control of the land and access to the infrastructure that matters: power, water, cooling, and fiber. More importantly, it has to clear our expected return threshold. Bluntly, if a site doesn't meet all five criteria, we pass. When one does, we run it through a process that we've done many times before. We initiate, we do due diligence across every dimension with third-party validation. We will acquire on disciplined terms. Then we will commercialize it quickly to the highest return demand. Let me just wrap up where I began with this team. The experience that we bring is what puts us in an absolute unique position to win, even in a competitive landscape. A great deal of capital is chasing land and power right now, and it is a competitive process. What I believe sets us apart isn't the willingness to pay, it's the ability to execute. Sellers, landowners, utilities, all want a counterparty who will actually energize a site and not just tie it up. That's a much shorter list than the list of people who can write a check. We offer flexibility on structure that a single- model buyer can't, and we move with the discipline of a team that has run this process many, many times before. That's what will win sites, and it's hard for just a pure financial buyer to match the level of experience this team has. Here's the point I'll leave you with. We're not building our first data center while we learn. We are scaling something that this team has done for decades, and we're doing it from a real position of real financial strength, which Chris will share with you. First, let me hand you over to Richard Carson, our General Counsel. Thank you. Good morning. I'm Richard Carson, general counsel of Ionic Digital. My role is to help Ionic enter the public markets with discipline, governance, disclosure, compliance, and public company readiness. I've spent more than 30 years as a corporate attorney and executive across energy, capital markets, restructuring, and public company governance. I've served as counsel on multiple public listings, providing direct experience for most of what this next chapter asks of us. Turning to page 26, I want to cover three things: how our governance is built, the experience of the team taking this company public, and why this listing is structured to align with our shareholders. First, governance. Ionic has a seven-member board. Six of them are independent. Our CEO is the only management director, an independent director, Elizabeth LaPuma, serves as the chair. That separation of chair and CEO is deliberate. Management runs the company, and the board independently oversees it. Our three key committees, audit, compensation, and nominating and governance, are fully independent with written charters, and our audit committee chair is a designated financial expert. We operate with one share, one vote. A single class of common stock, no dual class structure. Our recent private placement preferred stock will convert to common stock at our listing. As a result, every shareholder's voting power matches their economic stake. Second, the team taking us public has done this before. Between us, Chris and I have taken several companies public and provided ongoing public company leadership for several others. Public company readiness isn't something we're improvising at the listing. It's been built deliberately ahead of time. Internal controls, SOX compliance, SEC standard financial reporting, and audit-ready financials supported by an independent auditor. Many companies can list, but fewer are genuinely ready to operate as a public company on day one. That readiness reflects a management team that has operated in the public markets across prior roles and built this one to that same standard. Third, this listing is structured around our shareholders. A fair question is: why a direct listing rather than a traditional IPO? The answer goes right to alignment. Many of our shareholders received their shares through the Celsius reorganization and waited years for access to a public market. Our goal was to give them that access now. Just as important is how we're able to do it, because we're already funded, and we don't need to raise capital in this process, so we're not diluting our existing shareholders to do it. An IPO sells new shares and raises capital. A direct listing, on the other hand, simply registers existing shares for resale and lets trading begin. In our case, we will trade on Nasdaq under the ticker IOND. The market sets the opening price through Nasdaq's processes. The company does not. This approach puts existing shareholders first. Access to liquidity without dilution. The result is a governance structure and a team built to hold this company to a public company standard. With that, I'll hand our presentation over to Chris for the financial overview. Thanks, Richard. Turning to our financial highlights on page 28. Our financial strategy rests on a few simple principles. One, build long-term contracted revenue. Two, keep the balance sheet strong and flexible. Three, allocate capital with discipline. On revenue, we focus on long-term contracted income from strong counterparties, providing predictable recurring cash flow. Once Ward County reaches its full 323 MW of contracted capacity, we expect $251 million of annual contracted revenue with 98%-99% operating margins. On the balance sheet, we operate today with no debt and meaningful liquidity. Approximately $630 million between our treasury assets and the proceeds of our recent private placement. That strength is deliberate and provides capacity to be opportunistic, to move quickly on the right opportunities. On capital allocation, we deploy against a clear set of priorities. Expanding our existing sites, converting Midland to AI use, and pursuing disciplined acquisitions in new markets. The goal is durable, high-return growth. Taken together, that's a high-margin, contracted, capital-light cash flow profile with a balance sheet built to support growth. Turning to historical financials on page 29, you can see the transition come through clearly in the numbers. Full year 2025 total revenue was approximately $141 million, substantially all related to our mining business, and only 5% from digital infrastructure leasing. However, looking at the first quarter of 2026 annualized, 85% of our revenue was generated from digital infrastructure leasing, reflecting the start of revenue recognition under the Nscale lease. Over that same period, adjusted gross margin moved from 38% to 88%. The same shift shows up in gross profit. Quarterly adjusted gross profit was about $16 million in the first quarter of 2025, and approximately $25 million in the first quarter of 2026. That improvement is almost entirely the increase in contracted lease revenue. One timing point worth noting, we began recognizing lease revenue on a straight-line basis in December of 2025, but the monthly cash rent under the Nscale lease doesn't commence until August of 2026. The GAAP revenue inflection is already in the numbers, while the cash inflection is still just ahead of us. On the balance sheet, the most important recent event is the private placement we've just closed, summarized here on page 30. We raised $400 million at a $53 per share price, a $2.4 billion post-money valuation. We raised it from a group of tier-one institutional investors. The anchors are Oaktree, Attestor, and Sachem Head, joined by Citadel and Weiss. The securities convert one-to-one into common stock at listing, and there's a six-month lockup, with the ability to sell above $70 a share during that period. As part of the transaction, the investors received roughly three million five-year warrants struck in steps from $64 up to $87. The warrant strike sits at a meaningful premium to the $53 these investors just paid. The lowest strike is roughly 20% above it, the highest more than 60%. This capital only earns a return as the equity appreciates from here, alongside the rest of our shareholders, not ahead of them. This is sophisticated capital taking a long-term position with no special governance or information rights after listing. It's worth remembering what these investors were investing in. Not a concept, but the same assets, the same lease, and the same growth path we've walked you through this morning. They did due diligence, and they committed at a $2.4 billion post-money valuation. The view we've laid out today isn't ours alone. Between this raise and our existing treasury, we hold roughly $630 million of liquidity on a debt-free balance sheet, which provides substantial capacity to execute the plan in front of us. Now, turning to page 31 and the earnings power of the Ward County lease and the potential in our existing pipeline. The Ward County lease produces about $183 million of annual contracted NOI beginning at the August 2026 delivery after the initial ramp. The 89 MW expansion adds another $69 million, which brings total contracted Ward County NOI to roughly $251 million, a 37% increase in fully contracted to Nscale. I'd point out that those figures are before any contribution from the remaining 377 MW at Ward County or from the Midland conversion. Both of those are shown on this page as illustrative upside on top of the contracted base. They're not in the contracted numbers. The illustrative bars on the page reflect $130/ kW rate against the $65/ kW we contracted with Nscale. Not a forecast, but a sense of the embedded potential. As we progress through our development opportunities, we'll make that powered shell versus turnkey call site by site on yield cost exactly as Andy described earlier. Importantly, we expect this to scale with real operating leverage. Our existing margin at Ward County runs 98%-99%, and we expect to bring our existing pipeline online with only marginal increases to normalized G&A. As that contracted NOI grows, very little of it is expected to be absorbed by overhead, which would result in a company-wide level EBITDA margins potentially expanding as we scale. The result is $251 million of contracted NOI with Nscale and a much larger pool of embedded upside in the remaining 377 MW and the Midland conversion. Same footprint, same playbook. That's before any success from our third pillar, the inference and agentic frontier Antonio described. Finally, a moment on how we deploy capital. Our priorities are to fund the Ward County expansion, convert our Midland sites into HPC and AI data centers, then, as Antonio and Mark described, target inference and agentic demand by acquiring and developing sub-100 MW sites in the major metro markets. We also hold roughly 2,800 Bitcoin in treasury, worth about $192 million at the end of the first quarter of 2026. That Bitcoin is part of where Ionic came from, not where we're going. We don't view it as a long-term position, and we treat it as flexible source of funding that supports growth while keeping our balance sheet debt-free. We expect to sell it to fund the build-out of our own sites, which we view as a higher and better use of that capital. At Midland, we'll keep mining while it remains profitable as we develop those sites into HPC and AI data centers. Our strategy is to be a pure play HPC and AI data center company. On dividends, given the size of the market in front of us, we believe the best use of our cash flow is to reinvest it in that growth opportunity, so we don't expect to pay a dividend in the near term. That said, we'll always evaluate the best use of our capital and allocate it wherever it creates the most value for shareholders over time. The result is a company with a debt-free balance sheet and the liquidity to fund a significant amount of near-term growth. The plan we've laid out for you today. To wrap up, we have a contracted high margin revenue base, a debt-free balance sheet, and a capital-light model that produces strong returns as we grow. We're entering the public markets in a strong financial position, funded, contracted, and ready to execute on the opportunity in front of us. With that, I'll hand it back to Andy. Thank you, Chris. In closing, let me bring this back to where we started. This morning, we laid out three pillars, a contracted base, embedded growth inside our own footprint, and the agentic frontier ahead of us. You'll see the full case summarized here on page 33. Let me answer the question I opened with. Why Ionic specifically? I said most companies making this transition can point to one, maybe two, things that matter, and that we have the full set. Here's what I mean. We are positioned in a once in a generation AI and HPC cycle, and unlike most, we have energized capacity today, not years from now. We have a contracted base behind it, 323 MW and roughly $2.6 billion under contract, triple net lease with Nscale, supported by an NVIDIA guarantee and Microsoft demand. That base scales a clear path to 700 MW at Ward County with a Midland conversion right behind it. We can deliver power any way the market wants it. Powered shell to turnkey. We choose a structure that earns the best risk-adjusted return. We have an operator-led team that has done this work for 30 years through every prior shift in computing. We do it all from a position of financial strength. $251 million of contracted NOI once ramped, debt-free with strong liquidity, further bolstered by our $400 million raise at a $2.4 billion post-money valuation. None of those on its own is the whole story. It's the combination. It's all of them working together. It's difficult to replicate, and I believe that is what will create durable value for our shareholders over time. I mentioned at the beginning this might get a little dry or very analytical. I appreciate you staying with us because I think the story resonates, and it is really amazing. Thank you. I'll be happy to take your questions. Thank you. As a reminder, to ask a question, please press star one one on your telephone and wait for your name to be announced. To withdraw your question, please press star one one again. Please stand by while we compile the Q&A roster. Our first question comes from Richard Choe of JP Morgan. Your line is open. Great. Thank you. Thanks for all the detail. Just to kind of summarize and set expectations a little bit, how should we think about the delivery and cash ramp? You mentioned August, but into next year and as we go through next year, then maybe a little longer-term question, how should we think about the remaining 377 MW when that could be signed? Would it be post the Batch Zero process? Would it be closer to when the substations or the additional substations are up and running? How should we think about that also? Thank you. Great. Thanks, Richard. This is Chris. We'll take those in reverse order. We'll let Andy start with the 377 MW, then I'll touch on the cash ramp. Perfect. Richard, thank you for your question. We expect to hear from ERCOT on or around August 7th in terms of their decisions for Batch Zero. Batch Zero will be divided into two groups, Batch Zero Baseload and Batch Zero To-Be-Studied. Based on all of the documentation feedback from our outside counsel, and other conversations, we expect to be in batch zero baseload. Batch zero baseload would give us absolute confidence that we'd be getting the additional 466 MW, the first 89 MW of which goes to Nscale. A full 466 MW when it becomes available at the end of 2027. Once we have confidence that we're getting that power, we will begin our market outreach. We'll talk to hyperscalers, neoclouds, large enterprises. We fully expect to take that site to market in a turnkey or build-to-suit solution, which would maximize total return on invested capital and believe that it'd be possible to actually execute a contract before year-end. It takes 12, 18, 24 months to build. As you've seen in the market, many of these contracts are being signed well ahead of final delivery. To reiterate, we think we'd be able to execute a contract before year-end on that additional power, but we want to make sure that ERCOT has confirmed that we do in fact have it. On the cash ramp, Richard, as we mentioned in our prepared remarks, cash pace begins in August. To break it down for you, it's again, it's about $175 million of annual GAAP revenue annually really through the term of the lease, because of the straight-line nature of that contract. The third quarter cash component is $6.5 million. The full year 2026 number would be approximately $30 million. It ramps up and beginning getting to a good run rate in Q1 2027, to about $34 million-$35 million. It's got $135 million of cash revenue in 2027. We hit the full normalized run rate in 2028 at $182 million of cash revenue in 2028. Great. Thank you. Yep. Thanks for your questions. Thank you. Our next question comes from Michael Donovan of Compass Point. Your line is open. Hi. Thanks for taking my question. You discussed access to power being a priority when looking at Are we looking at the meter opportunities, or would you be open to behind- the- meter? Great question. Behind- the- meter is getting a lot of publicity in the market right now. We don't believe it's a panacea, and there are only a couple of operators who truly know how to do that. Our focus is going to be on grid power. Grid power for our expansion into major metros for agentic and inference. We think that is a better long-term solution. It's also a bit unrealistic to think you'll get the licenses to build natural gas plants in and around those major metros. We would be open to a conversation on behind- the- meter, it's with a very small handful of power companies who really know what they're doing. Just to reiterate, our key focus is on grid power. We think that's a winning strategy for the long term. Thank you. A follow-on, if I may. Can you discuss the type of counterparties that are showing the greatest interest at your Midland site? Great question. We haven't officially taken those sites to market, yet we've had over a half dozen inbounds for them. We've had interest from neo clouds that are looking at those sites for multiple use cases. We've had interest from technology firms as well, we've had interest from, I'd say, network connectivity-focused entities that are moving into cloud services. Three, even four different use cases, all of which need access to grid power quickly. We're excited for being able to convert those, excited for what the team can do with them. It's not only one use case we're trying to solve for. Appreciate it. I'll hop back in queue. Thank you. Thanks, Mike. Our next question comes from Marc Bachner of Mizuho. Your line is open. Great. Thank you. Just thinking about the contracting for the turnkey opportunities, how will your parameters differ versus the powered shell opportunity across those five parameters that you guys discussed, in terms of the sites' location, size, and everything like that? Let me touch that on a high level and then ask Mark to jump in as well. In terms of parameters for a turnkey build, I think it's incredibly important to understand the risk that the developer takes on, not just the headline number in terms of what the rate is. Right now, a lot of talk around of an $X billion deal or $X / kWh —or per kilowatt, excuse me. We will certainly maximize revenue and yield, but it's really important to understand the risk that goes along with building a turnkey data center as it relates to potential overages, potential delays in delivery date. We'll be very cognizant in managing a contract that touches on those. I think it's important for Mark to weigh in on the complexity, frankly, of building a turnkey data center and how we're thinking about it to manage that risk. Marc, we're taking three different approaches, depending upon the specific client. We will look to do a build- to- suit if the client has a very well-defined specification that we can address in a specific time period. We will take a design build approach where there may be a need for a collaboration with that specific tenant. Over time, our intent is to develop, plan, complete our own internal Ionic specification, which we've just started to commence, so that as we move into that third pillar that was explained earlier, that we will have a variety of different modular approaches that we can provide and discuss openly and transparently with the specific clients that we're looking to deal with. Great. Thank you very much. Thank you. Our next question comes from Jon Petersen of Jefferies. Your line is open. Great. Thank you. I just wanted to clarify if you could help on Ward County. Can you talk about what you do own at the site and what Nscale owns? Do you just own the land? Do you own the physical buildings around it? I see this point in here that you assume ownership of the data center infrastructure if the lease isn't renewed. Does that allow you to potentially raise debt capital on the value of those assets that Nscale has invested in? Just to clarify, we own the land. We own and control the substation. The land we own is the 136 acres, the 50 or so acres that Nscale is building on today, plus the 86 acres on the back half of property. We retain ownership over all of that during the lease Nscale is doing the fit out of the four or five existing buildings that are there. We have the foundations poured and the steel infrastructure in place. They're spending $2+ billion doing the data center fit out. I think what we said in the presentation is if Nscale does not renew the lease after 10 years, we retain ownership of those buildings and anything attached to the buildings. Essentially, a fully built-out data center. If your question is on raising debt, we have the ability to raise debt against the lease, particularly with the NVIDIA guarantee for the first five years. Even beyond that, to borrow against the overall lease term, we would not be borrowing against the assets inside of the data center. Got it. Okay. The $65/ kWh base rent, is that a year one rent, or is that the average over the 10 and a half year term? That is year one through five. That, importantly, I'm not sure if we distinguished, that is on gross kilowatts. That's not on net. A lot of statistics that you'll see in the market are based on a net kW rate. It's inclusive of PUE. Ours is on a gross. After year five, that rental rate increases by 3% per year. Okay. And Jon, this is Chris. And then- I'll just add. That $65 is the number when it's fully ramped. They are ramping into that run rate, kind of cash revenue base that I mentioned. Okay, that's very helpful. Yeah. Then the 10-year extension at market rate at the end, I guess, how is market rate determined? Market rate is defined in the contract. It is basically their ability to replicate that rental rate for that kind of property in that market. Somewhat generic terms for the industry. If we're not able to agree on that, then we would go through arbitration. We fully expect we'll be able to renegotiate that rate at a competitive number in year 10 and 11. Okay. That's all very helpful. Thank you. Thank you. Our next question comes from Brandon Nispel of KeyBanc Capital Markets. Your line is open. Hey, guys. Thanks for all the detail here. Just wondering, sort of looking beyond Ward County and Midland, can you give a sense of how many sites you're looking at in terms of really developing a land bank that you can really formulate a growth strategy longer term? Thanks. Hey, Brandon, thank you for the question. Admittedly, we are a little bit early on. The team has come together very, very quickly over the last couple of months, and we have prioritized Ward County, prioritized working with ERCOT, and making sure that Nscale gets off without a hitch. That said, we expect to look at hundreds of sites, literally hundreds of sites over the next couple of years because we know how hard it is, and that when you look at 100, you maybe have three or four that actually pan out. Between the cash flow we have coming in from the Nscale lease and the capital we raised, we think that that allows us to acquire a couple of properties per year. I think as Mark said, we're targeting sites that are below the large load requirement of the utilities, and we're seeing price points for these sites range anywhere from $200,000 to maybe $1 million/MW. If we're buying roughly 50 MW sites, that gives you a sense for how much capital we might spend over the next year or two. Thanks for the color. Thank you. Our next question comes from Ben Chamers of BCIC. Your line is open. Hey, thank you guys for the update, and thanks for taking my question. I was just wondering, is there any advantage you think in Ward County and in the Batch Zero process, given that it's an expansion capacity versus putting on a new site or acquiring a new site? Yeah. Great question. We absolutely believe that we are in an advantaged spot because it has existing load. We were actively mining Bitcoin, using the full 234 MW for several years before the Nscale lease. We think that positions us really well, especially when ERCOT is looking at, is it a real load or is it a speculative load? We believe that's also a notch in our case for getting in. Mark, anything to add? Yeah. I'd further add to that, Ben, that we have an excellent relationship with TNMP, who are the TSP on this site. We've had weekly calls with them because we are an existing customer. As we referenced, we've already got that 234 MW energized for a second time now as a data center use with Nscale. TNMP are very excited about this project, as you can imagine. Have worked really hard and collaboratively with us to ensure that as we've re-put back in all of the paperwork, that we have crossed all of our T's and dotted our I's. We're very excited about where we stand on the project. Super helpful. Thank you, guys, for the color. I'm going to jump in for a second. This is Chris Hickman. There was a lot of questions that came in from our annual shareholder meeting yesterday through the day. One very common question from our legacy shareholders was around the process of our direct listing and the shareholders' ability to access, unlock, and sell. We wanted to take a moment during the Q&A session to address those. First and foremost, existing shareholders can register their shares currently with Odyssey, that process can happen now. Once our S-1 is deemed effective, those shares will then be able to be transferred into a brokerage account. Between now and then, existing shareholders will need to set up a brokerage account in order to transfer those shares. We will be posting information on our website and make it available to existing shareholders as to what this process exactly looks like. Brokerage accounts can be set up currently, but the shares can't be transferred into a brokerage account until the S-1 is deemed effective. Once we begin trading, a few days after the S-1 is effective, those shares then are available to be sold from your brokerage account upon the first day of trading. We are targeting trading later this month. Before the end of the month is our target. Again, we'll continue to assess market conditions and SEC readiness as we make that decision. Didi, we can maybe check to see if there's other questions from the analysts. Thank you. As a reminder, if you have a question, please press star one one on your telephone. One moment for our next question. Our next question comes from William Johnson of Canaccord Genuity. Your line is open. Hey, thanks for taking my question. I know the acquisition strategy deliberately, that you're kind of looking at sites that are below large load thresholds. Just kind of on the existing Texas footprint, I was kind of wondering in the future sense, how you think about ERCOT tightening on power costs and just kind of energization timelines and how those may be pushed out or affected over the next two to five years. Yeah. I think a lot will be more clear on August seventh when ERCOT makes their announcement. For our property, if and when we're put into batch zero baseload, there really is very little risk in terms of when that energization happens. Getting that additional power is dependent on two transmission line upgrades, one by TNMP, that'll be completed imminently, the other by Oncor. That Oncor transmission line upgrade is already in flight and scheduled to be completed next year. All of this is public and on the Oncor website. When those two are completed, this part of West Texas will have the ability to take in the additional power. We see very, very little risk in getting the power. Thank you. Where your question is really applicable is on the group that's not considered baseload, that's on the to-be studied, and that is where ERCOT is going to have to make decisions in terms of allocating power, potentially curtailing power, or people maybe not even getting access to the power until 2030 or beyond. I think ERCOT really bifurcates this market into two segments, and the baseload segment is where we believe we'll be, and we have confidence in terms of power delivery. Thank you. Our next question comes from Mark Schultz at Raymond James. Your line is open. Mark, your line is open. Please unmute. Mark, please unmute. We have no further questions at this time. I'd like to turn it back to Andy Stewart, CEO of Ionic Digital, for closing remarks. Thank you, Didi. I just want to thank everyone for being on this call, the research analysts, all of the shareholders. We know our shareholders have been through an ordeal to get here. We're incredibly excited to bring the option of having liquidity in Ionic, and I think as you could hear in our voice, we're incredibly excited for the Ionic story. We have just really great things ahead. We appreciate everybody's time. Also important to note that the investor presentation will be available on our website, so you can go there for further information. We're excited to continue to tell the story as we march ahead to our direct listing. Thank you very much. This concludes today's conference call. Thank you for participating, and you may now disconnect.