CleanSpark, Inc. (CLSK) Earnings Call Transcript

Good afternoon. My name is Lacey, and I will be your conference operator today. I would like to welcome everyone to the Third Quarter of Fiscal Year 2026 Financial Results Conference Call for CleanSpark. [Operator Instructions] Thank you. I would now like to turn the floor over to Harry Sudock, CleanSpark's Chief Business Officer. Sir, you may begin.

Thanks, Lacey. Thank you for joining us today to review the third quarter 2026 financial results for CleanSpark. We encourage you to review our earnings results press release, which was issued today and is available on our website. Our 10-Q will be filed shortly. A webcast replay and transcript of today's call will be added to our website once available. On the call today, I'm joined by Matt Schultz, our Chairman and Chief Executive Officer; and Gary Vecchiarelli, our President and Chief Financial Officer. Some of the statements we make today will be forward looking based on our best view of the world and our business as we see them today. The statements and information provided remains subject to the risk factors disclosed in our 10-K. We will also discuss certain non-GAAP financial measures concerning our performance during today's call. You can find the reconciliation of non-GAAP financial measures in the press release, which is available on our website. And with that, take it away, Matt.

Thanks, Harry. Good afternoon, everyone, and thank you for joining us. To begin, I'd like to discuss how we're putting our operations and balance sheet to work as we continue our evolution into a diversified digital infrastructure platform. Following the close of this quarter, we completed an agreement that is the most significant transaction in CleanSpark's history, a 20-year triple-net lease at our Sandersville, Georgia campus with a high investment-grade leading global technology company. That agreement marks the start of our evolution into a diversified digital infrastructure platform, and begins the monetization of our power portfolio at an institutional scale. The lease represents $6.6 billion in contracted revenue to the business with the ability to scale to $11.6 billion with the execution of two 5-year extensions. Being a true triple-net structure, we expect near 100% net operating income margin. We have also forged a partnership with a vertically integrated mechanical, electrical and plumbing manufacturer and design builder and added 885 megawatts of commercial growth. Taken together, these developments bring our second mover advantage fully to bear. Many of you joined us for our investor update on July 14, and today, I would like to recap 3 things. First, I want to spend some time on why we believe this lease is not just a milestone, but among the absolute best commercial outcomes in this sector. Second, I'll update you on execution at Sandersville across our Texas portfolio under exclusivity and across the broader platform. And third, I want to talk briefly about our legacy Bitcoin mining business and how the optionality it provides us a strategic advantage as we enter this next phase of evolution and growth. Let me start with the market. The demand for compute continues to outpace the industry's ability to supply it. The constraint is not chips, and it is not capital. It is energized grid-connected power at scale, sitting on land that can support a full AI campus build. That combination is scarce today, and we believe it will remain scarce for years to come. The agreement is a 20-year triple-net lease covering the entirety of our Sandersville campus, nearly 250 megawatts of gross capacity and 175 megawatts of critical IT load. The initial term carries approximately $6.6 billion of contracted revenue with two 5-year extension options that bring the total to approximately $11.6 billion over 30 years. We expect an average annual NOI of approximately $330 million and we project our cost per critical IT megawatt at the $10 million to $12 million range, in line with the market for this generation of data center. Here are 2 things I want everyone to take away from those numbers. First, we believe, on a risk-adjusted basis, the economics are among the best in the space. Headline comparisons across data center deals can be misleading as different contract durations, different lease structures, different risk allocations and different financing packages can obscure actual returns. Next, this is a triple-net lease taxes, insurance and importantly, maintenance CapEx all sit with the tenant, which means our contracted revenue converts to net operating income at effectively 100%. But the benefit goes beyond that. The triple-net structure is a powerful risk mitigant on the finance side. Lenders underwriting this project are looking at 2 decades or more of clean, predictable high-margin cash flows backed by a high investment-grade counterparty with no operating cost volatility sitting between the rent check and the debt service. This is why we were able to structure this deal without credit wrappers and why we expect most of this build-out -- pardon me, we expect to finance most of this build-out with attractively priced project-based financing. Gary will speak to the capital markets backdrop, but the short version is this. The structure of this lease was engineered to make financing straightforward and support a high loan to cost, protecting our equity. We deliberately went after a longer duration lease. We deliberately targeted a triple-net structure, and we deliberately chose a high investment-grade counterparty as our tenant. Each of these choices was made with the strategic intention to protect and enhance shareholder value. As we sit here today, we have already fully funded the equity portion of the Sandersville project. On execution, the project is advancing according to plan. Site preparation work has been underway for months on a 122-acre greenfield parcel that we acquired earlier this year, which allows us to build without disturbing our Bitcoin mining footprint just down the road. We're working with our tenants' preferred engineering, procurement and construction management firm and general contractor, a partner with a proven track record and domestic manufacturing capability across key mechanical, electrical, plumbing and cooling components. That familiarity meaningfully derisks the supply chain and the long lead equipment time line. It's also another step in shifting construction out of the field and into the factory, reducing on-site labor and delivering a faster, more repeatable build. The first data hall remains on track to be ready for service in the fourth quarter of calendar 2027, and we've already ordered and prepaid for all data center long lead items for RFS. Let's turn to Texas. In conjunction with the Sandersville lease, our entire Texas portfolio, 718 acres and up to 885 megawatts of secured and planned capacity across our Sealy and Brazoria campuses is under exclusivity with the exact same counterparty. As I said on July 14, we are in an exclusivity window, not at the finish line. Still, the scope of this discussion reflects the portfolio dynamic we described to you last quarter. Sophisticated compute buyers are not looking for a single data hall. They want a growth path measured in gigawatts and a developer they can trust across a multi-decade relationship. Sandersville was the first chapter. Exclusivity across 885 megawatts in Texas shows how our counterparty views the rest of our portfolio. Outside of CleanSpark's work in Texas, there has been a broader policy discussion this week. As many of you have likely seen, Governor Abbott directed the Public Utility Commission of Texas and ERCOT to audit and verify all data center projects seeking a grid connection before they can move forward. While this may be seen as uncertainty or short-term setback, we don't see it that way. We commend the Governor for his thoughtful attention to one of the nation's most important power markets and the broader impact of ERCOT's upcoming actions. The operational rigor and financial commitments required to progress through the next phases of development and energization in Texas align exactly with how CleanSpark already plans and operates. While a formal batch process is now expected to extend beyond August 7, our campus quality, disciplined site selection and project readiness remain unchanged. ERCOT has notified large load applicants that final Batch 0 determinations will be delayed with the next key milestone, the August 20 PUCT open hearing. In the meantime, the state agencies are mobilizing to address the Governor's directive, which underscores the critical role our sector plays in the Texas economy. CleanSpark's historical growth through countercyclical acquisitions provides a strong basis for us to run a similar playbook in the ERCOT market. As an example, you may recall that we acquired the original Sandersville campus during challenging political and economic times for the prior owner. We anticipate similar opportunities might develop in Texas in the near term. It's critically important for everyone to know that our previously discussed exclusivity agreement with -- remains intact. We continue to invest in our Sealy and Brazoria sites and there is no anticipated change to our energization time lines. Across the broader portfolio, we hold 1.8 gigawatts of contracted capacity today with a high degree of confidence in growing beyond 2.1 gigawatts in the near term through the ERCOT review process, and we are evaluating multiple gigawatts of projects as we look to scale and diversify our digital infrastructure across the U.S. Our land and power engine has not slowed down because of this lease. If anything, this transaction sharpens our conviction. Powered land is the scarce asset of our area -- of our era and finding it, contracting it and developing it is what our team does better than anyone else. Before I hand it to Gary, I want to spend a moment on our legacy Bitcoin mining business and the role it plays as we evolve into something profoundly different. When speed to power matters to a utility, we have the ability to bring a paying load online faster than a traditional data center developer. That capability helped us build this portfolio, and it will help us to expand it. At Sandersville, mining will continue to keep the site productive until the day of cutover. Financially, mining and our Bitcoin treasury give us a set of levers that very few companies in this sector can match. Our HODL balance is not a passive position. It is a flexible source of capital that we can deploy in whichever way creates the most value. What that means in practice is that we are never forced sellers of anything, not our Bitcoin and critically not our equity. Gary will walk through the capital strategy in detail, but the principle is simple. Mining funds the platform, the balance sheet protects our shareholders and the AI business monetizes the portfolio for decades to come. I'll close where I always do, and that's with our people. The lease we signed in July reflects the grit, discipline and talent of a team that I am privileged to work alongside. From the operators in Sandersville and all across Georgia, Texas, Mississippi, Tennessee and Wyoming to the deal team that completed this transaction, a heartfelt thank you. Your tireless commitment to the mission, we all share has paved the way for our future success. And with that, I'll turn it over to Gary to walk through the numbers. Gary?

Thank you, Matt, and good afternoon, everyone. I will dive into the quarter's numbers and then talk about our new Sandersville lease and our capital strategy going forward. For the quarter, our revenue of $138 million reflected a modest increase of 1% compared to the preceding second quarter. The average revenue per Bitcoin mined in Q3 was approximately $72,000, a 5% decrease from the prior quarter, when the average revenue per Bitcoin mined was $76,000. However, we still managed to recognize an increase in revenue as our uptime and hash rate increased in the third quarter. Gross margin in the third quarter was approximately 38% compared to 40% in the second quarter as power prices had a modest increase of $0.01 over the prior quarter. GAAP net loss for the third quarter was $240 million compared to a loss of $378 million in the preceding second quarter. The primary driver in this quarter's net loss was unfavorable mark-to-market adjustments on Bitcoin balances, approximately $133 million. Comparatively, we recognized a larger unfavorable mark-to-market adjustment of $263 million in the prior quarter. Our adjusted EBITDA in the third quarter was negative $113 million. It is important to note that this adjusted EBITDA number does not adjust out the $133 million mark-to-market adjustment. When normalized for such noncash items, we produced $20 million -- a positive $20 million for the quarter. In the third quarter, our digital asset management activities once again delivered net positive cash returns. This quarter, we generated approximately $8.6 million in cash, which brings our total for the fiscal year to $25.8 million. Additionally, we sold the majority of our quarterly Bitcoin production, and through the efforts of our digital asset management team, we were able to recognize sales for Bitcoin 7% greater than spot. Before we open it to Q&A, I want to discuss our capital strategy. With the Sandersville lease announced and progress being made towards future lease announcements, we are quickly approaching a day with much more consistent, predictable and durable cash flows. Our job between now and then is to unlock significant shareholder value by continuing to educate the market on the evolution our business is undergoing and the progress we are making towards that goal. Last quarter, I mentioned our second mover advantage in the AI infrastructure space. As Matt discussed, we have seen this play out in the commercialization of Sandersville. The 20-year duration of our lease, combined with its triple-net structure and the high investment-grade credit quality of our counterparty put us in a rarified air amongst our peers. We are positioned extremely well to execute on time and on budget with strong returns for our shareholders. With respect to the financing for the Sandersville project, I want to be explicitly clear. Our plan is to seek project-based debt financing to cover the vast majority of the build out. Based on our estimated project costs of $10 million to $12 million per megawatt, we expect total cash needed for CapEx in the range of $1.75 billion to $2.1 billion. Recent financings have exceeded 90% loan-to-cost. However, we have the equity portion of the Sandersville project already funded, and we will not be raising capital via equity or equity-linked instruments for Sandersville. It's also important to note that we had total liquidity of $917 million as of June 30, including approximately $200 million of cash and almost 14,000 bitcoin. To follow up on Matt's remarks, we have a very positive view on the recent developments in ERCOT as we expect further accretive opportunities to present themselves. Prior to Governor Abbott's directive, there were a number of opportunities that came to us on the eve of the SB 6 attestation deadline requirement. This is quite telling, as we believe there are potential sites that either need capital or don't have access to capital. We have significant dry powder with our Bitcoin balance, which we would be willing to part with if and when the right opportunity presents itself. We will continue to be aggressive, yet disciplined in pursuing opportunities that secure a pipeline, which will deliver significant returns. With that, I'll hand it back over to Harry and lead us into Q&A.

Thanks, Gary. We will now open the floor to questions from the analyst community. Operator, please provide instructions and manage the queue for the Q&A session. Thank you.

[Operator Instructions] Your first question comes from the line of Paul Golding with Macquarie Capital.

Congrats on the progress in procuring all the long lead time items for Sandersville. I wanted to ask first on the $10 million to $12 million per megawatt CapEx range. Is there a benefit baked into that from the existing infrastructure at the site? And how should we think about that cost per megawatt versus, say, Texas, where maybe there's a little more greenfield development to do or maybe not? Just any color you could give on that. And then I have a follow-up.

Awesome. Thanks, Paul. This is Harry. So really important for our strategy for developing in Sandersville was to procure the 122-acre parcel a little bit down the road so that we could build out the data center as a true greenfield development. So when we think about all of the data center costs associated with the build out, those are really apples-to-apples versus any greenfield project in our mind. There is a very modest accretive value from the current deployment that's related to the substation being already built and energized. That point was a critical piece in the commercialization process because having landed power and energized power made that a particularly attractive campus. But when you think through the CapEx budget outside of the substation itself, really, this is a true greenfield build.

Great. And then maybe following on to that, you seem well placed, well positioned to leverage the liquidity position that you're in, and the opportunity presenting itself in ERCOT. How should we think about behind the meter versus grid-connected opportunities? It seems like you're encountering grid-connected opportunities given the commentary around timing in SB-6 and things that were presented to you. But just wanted to get a sense of the market, how you see that value proposition? Are you exploring behind the meter? Or is it not even necessary given what you're seeing around grid connected opportunities?

Thanks, Paul. I think that the fundamental macro perspective is that the explosive growth in the data center market and the relentless demand profile that sits behind that is going to require behind-the-meter deployments to achieve any of the -- even the bear or the mid-case data center growth numbers that are out there from yourself and from other shops that are looking to see what does this really look like over a 5-, 10- and 15-year deployment scale. It's something that we're very actively looking at internally. But like with everything, we move strategically and deliberately and we want to make sure that there is true continuity between the energy source, between the CleanSpark capability and between the customer needs so that we have full end-to-end project clarity before any of those solutions would come to market.

Your next question comes from the line of Greg Lewis with BTIG.

Realizing the opportunity in Texas has a long tailwind for you guys. You do have some other sites, right? You have Wyoming, you have Georgia, you have Tennessee. I'm just kind of curious, just given the fact that Texas seems to be in a little bit of a holding pattern. Could we see some of those locations actually move more to the forefront and have customers shown an increasing interest in those, just given -- realizing that this happened the other day. But it seems like Tennessee is becoming a hub. Obviously, you already have the 1 contract in Georgia. So just kind of curious how you're thinking about that 340 megawatts that's out there across those states.

Great, I really appreciate the question. This is Matt. I'll take that. I'd say given the scale of our Texas assets and how progressed our discussions are, both on permitting substations, the Batch 0 baseload process that we're in. I wouldn't exactly say that something is going to jump the line per se. But what I can tell you is the -- we've had a VDR open for a number of parties over time and the interest in assets like Washington, Georgia, for example, present an extremely compelling opportunity. It's 86 megawatts of energized capacity today. However, we've already submitted a line study at the request of MEAG and the local utility to expand that by up to 500 megawatts. So as we've progressed those discussions, we've also secured options on several hundred acres adjacent to the existing substation to support that build. Additionally, you nailed it. We have 100 megawatts in Wyoming. We have 60 megawatts in Jackson, Tennessee. We have other capacity in Ripley, Tennessee, as an example. So the -- I would say that the value of the assets outside of Texas for those in search of '27, '28 RFS went up significantly due to the uncertainty on the timing of this audit period that Governor Abbott pushed forward.

Your next question comes from the line of John Todaro with Needham & Company.

Congrats on the progress. I went through the slides, but we have been getting a lot of investor questions or I think we're just a little bit confused on Texas a bit here. And you did very explicitly walk through your unsigned Texas capacity. What still needs any ERCOT related approvals and what capacity in sites do not need any that you can go to sign?

Yes, John, absolutely. It's Harry. So the way that the acquisition process for the other 2 Texas sites went was very deliberate on our part. We are always thinking about risk management, first and foremost in the business. And so the way we went about acquiring those projects fit into the same framework. When we acquired the asset in Sealy, very importantly, that asset came with ERCOT approval already in hand in 2025. That changed in January of 2026 when the Batch 0 process was rolled out purely from a timing and requirement perspective. But what that site received out of the gate there was Batch 0 go status, which means the baseload designation. And so while we do require the final determination from ERCOT in order to interconnect that location, it has been given every green light every step of the way through despite needing the final check mark in order to proceed through that final energization gate. The second is in Brazoria. There, the 300 megawatts of additional Phase 1 capacity, those are functionally exactly the same as Sealy other than the fact that they were given the go status in 2026 rather than a 2025 overview. The second 300 megawatts in Brazoria are the ones that are -- is the segment of the megawatts that are subject to the final determination between baseload status and study load status though we still went through a very detailed transmission line assessment when we made the decision to acquire that location. And based on the transmission line capacity that we evaluated, we understand that site to have a tremendous amount of capacity, both upstream and downstream within that segment of the utility system. And so we're very optimistic, but that is the segment of the load that we are still waiting for baseload versus studied load status.

Understood. It's very helpful. And just a follow-up here on the demand environment. Matt, you commented a bit on it. But just wondering, is there any kind of hyperscalers or chip manufacturers that are getting more aggressive in looking at sites, any that have maybe pulled back and become less aggressive? Just any color there.

Yes, definitely no pullback. The inbound inquiries have ramped up in the last 48 hours, with questions surrounding whether or not there's any exclusivity on the Washington project. I'd say that's the 1 that has garnered the most interest of late just because of the capacity to scale in the short term.

Your next question comes from the line of Brian Dobson with Clear Street.

Congrats again on the deal. I guess, as you're considering Governor Abbott's memorandum, is that indicative of what we might see in other large call it, utility areas? And do you think ultimately, it favors established players, such as yourself?

Yes. So I think that -- let's take the second part first, which is that as we see the demand profile on the tenant side of the house continue to be relentless, the established players who have a long track record of procurement and development of energy and infrastructure assets are going to continue to rise to be the cream of the crop on a go-forward basis. When it comes to the proliferation of this type of audit or assessment, I think it's dynamic. I think that Governor Abbott is facing the largest interconnect application queue of anybody in the nation. And so his response is going to be tailored to the Texas-specific dynamics and the ERCOT specific dynamics that take that grid in particular. I think that ultimately, one of the things that is currently not well understood broadly in public is the economic benefit that data center projects represent to cities, counties and states across the nation and some of the technical specifics of data center deployments. There is dramatic technological improvement than 20 or 25 years ago in the way that data centers operate. Water usage has been crushed to near 0. The footprint that these data centers are operating on from a land use perspective is shrinking. And so I think that while there's still an education curve that the public is going through, when we look back in 5 to 10 years, this is going to look like another industrial revolution in America.

And if I may, Brian, just to add on to Harry's comments on the second part of your question about established players. I've engaged meaningfully with the senior site selection folks from a number of hyperscalers. And what seems to be the consensus is that similar to Las Vegas residential real estate 20 years ago, in Texas, a lot of the developers or a lot of the hyperscalers are encountering folks that may be put down a deposit on some land with proximity to transmission lines, and applied for status within ERCOT approvals. What we're seeing now is with the implementation of SB-6 now you have to put your money where your mouth is. And so a lot of these projects that may have real viability have kind of washed out because there's not the capital support behind them that's necessary. And we -- based on our engagement with the communities and with ERCOT, having the balance sheet to actively stand behind the guarantees and the attestation process puts us in a really favorable position going forward.

Yes. So ultimately, it probably spurs more responsible development. As you're looking at your portfolio, are there any regions where you think you'd like to enter in the, call it, medium term without being too specific, I understand?

Thanks, Brian. I think that one of the things that we really pride ourselves in is being dynamic and flexible. I think that the opportunity set across these power projects are really at the project level where the big opportunity evaluation process happens. And so I think that we have the diligence team and the relationships to be successful in essentially every American power market as it exists today, but really where the special sauce from our perspective is being able to take in an enormous volume of deal visibility and be the strongest selectors within that deal volume so that we are able to be hyper, hyper surgical and specific about where we make selections and a lot of that isn't about region or jurisdiction. It's even more granular than that.

Your next question comes from the line of Mike Colonnese with H.C. Wainwright.

First one for me. I'm curious, based on your discussions with your tenant at Sandersville, would you say that getting a final green light on Batch 0 for Sealy and Brazoria Phase I would be a gating factor for the counterparty to sign a lease here? Or do you think that's one of the gating factors? And the other thing, are there other development milestones that the counterparty is looking out for at either the Texas sites that would need to be achieved by CleanSpark for them to move forward?

Mike, thanks for the question. This is Matt. Just to give you a little bit of color, we had a VDR open that encompassed a number of our sites, and we invited a number of major players into that VDR to do an analysis. And what surprised us wasn't that, hey, I'm interested in Site A or Site B or Site D or whatever the case may be, the response was, I'll take it all. So our Sandersville tenant was also in the VDR, and as a function of the negotiations and understanding the tailwinds that we brought to the Sandersville project, they were aggressive in seeking that exclusivity and optionality on the go forward. Now we're still under exclusivity there. We're working with them constructively on a daily basis. But suffice it to say that should this slip beyond the -- if we get bad news on the August 20 PUCT meeting, there's so much pent-up demand for energized megawatts that I think that if they were to fall out, there would be another company rushing to the gates in short order because of the quality of the asset, because of the status in that batch process, but also because of the advanced nature of the work that we've done to prepare those for pad ready.

Very helpful color there, Matt. I was hoping you guys could provide a little bit more insight as to these opportunities that are coming across the table in ERCOT, maybe some of these underfunded or less legitimate projects. What are some of the characteristics that you'd be most interested in pursuing as it relates to the sites that are coming to you?

Yes. Great question, Mike. Some of the stuff that we saw in the past that we passed on lacked the high quality that I think is a standard for CleanSpark maybe had power that was in a later batch process that we passed on. But then as it got closer to that SB-6 period, the quality assets started to accelerate just because folks didn't have the bandwidth or the capital to be able to support those initial payments. So I think for us, nothing changes in light of Governor Abbott's position. We're still going to maintain the same rigor on site selection. It has to meet our criteria. We're not -- I think we talked about this back in the Bitcoin days, Mike, and that is we're going to find high-quality assets, and we say no more often than we say yes, rather than find powered land and then try and shoehorn something to fit there, we look for the opportunities that best fit the demand profile we're seeing.

Your next question comes from the line of Brett Knoblauch with Fitzgerald.

This is Nathan calling in from Cantor Fitzgerald on behalf of Brett. Congrats again on the deal. Matt, I thought it was interesting what you said the role of Bitcoin mining was not just in building the portfolio, but in potentially being a differentiating factor and how you'll continue to grow it. So could you just walk us through how mining potentially could be incorporated into future site acquisitions and development? And then would it be similar to Sandersville, where it's a stand-alone data hall separate from the data center? Or are there other configurations that you'd consider?

Yes. Great question. Thank you for joining. Tell Brett, we are not forgiving him for not showing up. So I think the interesting part about this is when a utility has available power or when there's a new generation that comes online, occasionally, these loads are offered to the market initially as an interruptible load. So to be able to raise your hand and say, I'll take that power with 90 hours of annual curtailment with the understanding that once we demonstrate sustainability on the new generation, it becomes a fixed load or a firm load. So Bitcoin mining enables that ability for us. Additionally, if it is firm power that is available to monetize today, for the benefit of the local utility and the community, they don't really care who's paying the power bill as long as the power gets paid. And we can drop in modular immersion-cooled Bitcoin mining technology inside of 90 days and start to monetize those megawatts that are otherwise going underutilized. We can then, just like we've done in many jurisdictions, acquire sufficient land behind that, that we can take our time to build a data center according to the basis of design of the end-use tenant and switch that power over if and when the data center is complete. It's important to note, the CapEx requirement for that has already been spent on many of these modular deployments, as we move out of Sandersville in a couple of years, we'll repurpose those immersion cooled pods and the ASIC to kind of plow the road, if you will, for future developments if and when those kind of interruptible or early energized asset opportunities make themselves available.

Your next question comes from the line of Stephen Glagola with KBW.

I have one for Matt and one for Gary. Matt, maybe could you provide more color on the expected time line for gaining clarity around the audit process? You referenced in your prepared remarks, the August 20 PUCT hearing as a key date. Like what should investors expect to learn by then? And maybe what are the next milestones we should be watching? And then Gary, you mentioned targeting high loan to cost on the project debt side. It'd be great to unpack that more and what you're seeing in the funding markets today around lender appetite and cost of debt? And maybe what has changed over the last few months?