All right, guys. Good morning, everyone. Thanks again for taking the time with us. We're back here, second round here of the day, talking with Bitfarms folks. We have Ben Gagnon and Jonathan here. I'm joined with my colleague, Jonathan Petersen here. We're gonna talk with the crew here. As usual, if you guys have questions, ping us here respectively on email or on chat if you guys want to weave in your questions, back to the team here. With that said, JP, do you wanna say a couple words and kick off our conversation? I'll plug back in here in a moment.
Yeah. Ben, Jonathan, thanks for being with us. It's been a while since we caught up, so I'm really excited to get an update from Bitfarms on what's going on. Maybe we could kick off with if you guys just wanna spend a few minutes and just update us on who Bitfarms is, where you guys are at in your strategy, and we'll take it from there.
Yeah. Thanks, Jonathan, and always great to speak with you and catch up. You know, Bitfarms has been a longtime energy infrastructure company. We've been operating for about eight years as a publicly traded entity. You know, in the past, we've always been focusing on monetizing our energy infrastructure portfolio of Bitcoin mining, but over the last year and change, we've really been focusing on 100% North American infrastructure and 100% HPC and AI infrastructure. We're no longer making any investments into Bitcoin mining. It's 100% focused on HPC and AI. You know, we are actually just about to complete our re-domiciliation, which will be effective tomorrow. As of tomorrow, Bitfarms will no longer be a Canadian company. It will be a U.S. company, and it will also be called Keel Infrastructure.
I think that reflects the changing nature of the business. I think it reflects exactly how we wanna position ourselves in the business as a pure infrastructure developer and owner. We're not trying to compete in the HPC and AI compute or the cloud space. We're really focusing on enabling hyperscalers and neo cloud customers to scale, to deploy on time, without, you know, and operate without interruption.
Okay. All right. That makes sense. Can you talk to us a little more about, like, where your site locations are, kind of some like, I guess, how much power you guys have and kind of at what locations and how you're positioning yourself in this environment?
Yeah. We've got a really exciting and unique portfolio. We've got about 2.2 GW of energized, secured, and expansion capacity spread across Pennsylvania, Quebec, and Washington. The vast majority of our pipeline is in Pennsylvania, where we have three sites. Starting with our smallest site, we have Sharon, 110 MW in the western side of the state, very close to Pittsburgh, Cleveland, and the border with Ohio. The next biggest site we have in Pennsylvania is our Panther Creek site, which is our current flagship development. We have 350 MW secured. There, we have potential expansion capacity to take us beyond 400, maybe upwards of 500 MW at that site, and that is about three hours away from New York and Philadelphia.
That one is attracting probably the vast majority of interest from the hyperscalers and neo clouds given its premier location. The last site that we have in Pennsylvania is a site called Scrubgrass. It's actually located about 30 minutes east of Sharon on the west side of Pennsylvania. This is our largest site that we have a potential GW of capacity, and that's broken out into two different buckets. There is a 750 MW detailed load study, which is under active processing with FirstEnergy now. We should have good visibility on that probably in the Q3, Q4 timeframe, and how we can convert that over to firm service.
We're also working to develop an additional 550 MW of on-site generation with nat gas at that location, bringing the total capacity up to 1.3 GW. Beyond that, we have a small site in Washington. It's about 18 MW that we're actively developing with Turner Construction, Vertiv, two of the biggest names in the industry by far. We have 170 MW in Quebec, where we're really focusing on 96 MW in Sherbrooke. We have three Bitcoin mining sites in Sherbrooke, a 48 MW, a 30 MW, and an 18 MW, and we are working to consolidate the power from those three Bitcoin mines onto a new campus that we will be developing in the proximity of those three sites. That's kind of the overall pipeline that we have and the areas that we're focusing on first.
Okay. All right. There's been a lot of, you know, new entrants to this data center or kind of the traditional data center or kind of hyperscaler-focused world. I guess, how are you guys trying to differentiate yourself versus some of the other players? You know, where do you guys think Bitfarms has an edge?
I think the real edge that we have is in our locations. You know, I think one of the things that people forget about this space is that, data centers are still a real estate play, and in real estate, location is, a huge source of the value. You know, clearly a megawatt in West Texas is worth, a different amount than a megawatt in Manhattan and a megawatt in something like the northern territories in Canada, right? A megawatt is not a megawatt. Where a megawatt is and when a megawatt is available is really what determines that megawatt's worth. When you look at, most of our peers, you know, you'll see that there is a very, very strong concentration in Texas and in the South. Whereas our concentration is entirely in the northern states, Pennsylvania, Quebec, and Washington.
We don't have a single site below 40 degrees north. What that means is that we're closer to these major metropolitan areas. We are in cooler climates, which means higher efficiency, which means lower OpEx, it means lower CapEx, and it means more critical megawatts are driving more revenue on site than they would be in a place like Texas. It's really our locations, which I think is our unique aspect. Although we don't have the largest pipeline amongst the peer groups, we do have the largest pipelines in each one of those three high-demand markets with huge barriers to entry that I just mentioned, Pennsylvania, Quebec, and Washington. You know, we're positioning ourselves as regional leaders in these key markets.
Right. You know, when we talk to, you know, various players, it seems like there's a debate on whether or not latency matters with data center locations. Clearly, you guys are in a location that's near population centers. I mean, how much is that part of the conversation when you're talking to potential tenants?
That's really tenant-dependent because it's use case dependent, right? When it comes to training models, the latency is not really a big driver for most of these customers. But when it comes to inference, it's certainly a very big, important requirement for these customers. Inference is where most of the demand is going to come from over the next couple of years. There's obviously been a huge emphasis on training of the models. When you look at, you know, the actual applications of AI, that's all gonna be coming out of inference.
When, you know, you or I are working with Gemini or Claude or whatever, that's the inference that's taking place. That's the applications. Those are the ones that are really generating the revenue, and their loca- like, location and latency matters. Fortunately for us, given our great proximity to places like New York and Philadelphia, you know, with our Panther Creek site in places like Moses Lake, our great proximity to Seattle and Portland and Vancouver, those give us a really distinctive advantage in inference.
I'd say the sort of activity you see in Ohio is in and of itself a commentary on moving, you know, west or across the northern states and where you can see significant pockets of demand up north.
Got it. Okay. Sorry, Julian, you're on mute, Julian, if you're trying to talk.
Sorry. You're right. I was on mute. I apologize about that, JP. Appreciate it. Ben and Jonathan, thank you guys again. I want to sneak a few in here if I can. Look, taking it back from a high level real quickly, how do you think about, you know, you're repositioning here, you know, re-domiciling, et cetera. What are the milestones now with this inflection or pivot here, that one should expect from you guys? You talk about this premier site in Pennsylvania. Look, I don't disagree with your characterization. In fact, I'm fascinated to see the success that you leverage with Panther and elsewhere into, like, tangible data points. But again, people talk about time to power. You guys have had the site for a minute. How would you set expectations where you are in this progress, A?
B, really critically, you know, Pennsylvania's seen a bunch of pushback on sites. You guys probably saw this, like, high-profile Montour development recently. How does this combination of additionality and pushback on new sites help you guys? I imagine it's a net benefit as a kind of quasi-incumbent. Tackle that directly, if you don't mind, guys. I appreciate you guys entertaining, like, a bigger picture here.
Yeah, of course. Happy to answer. It's a great question, although there are a few questions in there. If I forget one, you might have to remind me. The first part of your
Yep
The question around kind of what are the milestones and what are the catalysts for investors this year, you know, we'd actually talked about it quite a lot on our quarterly earnings call this morning. I think it's the number one question that investors are looking for is: What are the milestones? What should we expect, and what are the catalysts? The answer is that there are basically three catalysts that investors should be looking for. The first is a continued execution on the site advancements with permitting and construction. The real reason why that's a catalyst is because that's effectively the kind of closing conditions necessary for signing leases.
When you look around at market valuations for the peers in our space who have signed leases and who have unsigned leases, there's a big discrepancy between the two. It's not really necessarily a market. It's not necessarily assigning that value because they've delivered facilities where the revenue started. They're assigning that value, and they're driving that re-rating just from having the signed leases. The first catalyst that investors should be looking for and expecting is lease execution, which should come shortly after the NTP milestones across those sites. The second catalyst that investors should look for is we currently don't or we believe the market's not really assigning much or any value to our expansion capacity.
When you look at the value of our 2027 secured megawatts relative to our peer group, we're really trading, I think, in line with those peers who haven't signed leases. Almost two-thirds of our portfolio is in our expansion capacity that we believe is being assigned little to no value. Converting over the expansion capacity into secured capacity by firming up the detailed load study at Scrubg rass into an ESA or confirming the plans and the gas availability and the contracts for the on-site generation at Scrubg rass, for instance, securing the additional power at Panther Creek through the ISA or through the additional power allocation, we think that's gonna be the second major catalyst that investors can look for this year. The third catalyst is really gonna be delivery in 2027, when the development risk on these projects falls away as the sites get commissioned, and that drives another catalyst event.
Julian and Jonathan, the way I think about it is we have now shared a point of view with, you know, our shareholders and investors. Like, we're in the market commercial discussions with potential customers. Implicit in that is that we have a point of view on when each of those sites is gonna be shovel-ready, right? Because you wouldn't want to go to market with something that you didn't think was gonna be, you know, shovel-ready at the time you sign a lease, because that's where you get real value. From our perspective, as you look at our progress over quarters, this is actually a pretty big milestone in and of itself, getting to the market with commercial discussions, because in and of itself, that gives, or implies a point of view on the timeline for securing the last items we need to turn shards.
So, so I-
Let me ask you this about. Go ahead, go for it.
Okay. I guess as you have those discussions, I mean, what, you know, when can you get something online today? Is Panther Creek able to come online by the end of 2027 at this point, or are we into 2028 on some of your sites?
Yeah. It's a great question, Jonathan. You know, 2026 is a year of execution, so it's executing against permits and leases and site construction. 2027 is gonna be a year of delivery. So that's gonna be the year where we believe that every site is commissioned and starts generating revenue. You know, it could be that there is going to be multiple phases at Panther Creek, but it's still fully within our timeline expectations that Panther Creek turns online in 2027 and starts generating meaningful revenue.
Our point of view, look, generally is that if you look back at how we've spoken about the business over a number of quarters, some of which I was not here for. The way we would summarize execution is that it's been going smoothly, and the pace of change is accelerating.
Okay. I guess clearly you guys are in negotiation. Well, it seems like you're in negotiations with potential customers. I mean, what are some of the things they're looking for you guys to do for them to get comfortable to be able to sign a lease? Like, what are some things you've done over the last, you know, six-12 months to get yourself to a point where you're confident you can sign something this year?
I mean, really it's just been the continued execution and de-risking of the sites. That's the advancement on the permits, things like securing zoning, advanced work on or continued work on the architecture and the engineering. We've also engaged a number of external advisors. You know, we've got financial advisors and data center industry advisors that help us in this process. Every time we have, you know, progress on site, the demand spikes again. You know, for instance, we cleared zoning in February and, within a few days of clearing zoning, our inboxes were again, you know, flooded with inbound demand because what we have is so scarce, people are actively seeking it out. You know, we don't need to be out there trying to push our sites out to customers because they're all coming to the sites seeking what we currently have.
One of the things customers really value is time to market. Power capacity and an energized site that arrives sooner is better and is worth more to a customer. If you look back over the past six or 12 months, what you see us having done is deliver exactly on what we promised, which is our getting ourselves through that development process and arriving at a point where we can start the commercial discussions because we implicitly have a point of view on completing the process across each of those three sites.
Okay. I guess, talk about how selective you guys are and what sort of tenant you're looking to sign with. Are you restricting yourself to only an IG-rated hyperscaler? Are you willing to look at, you know, larger enterprises, neo clouds, hyperscaler guarantees? Like, there we've seen a lot of different structures out there. Are you guys limiting yourself to any of them?
I'll start simple and Ben will have a broader point of view. We view the investment grade rating either directly from your customer or via a credit wrap that is carefully constructed to be back-to-back with the lease. Credit wrap is a pretty broad term. Some will work, some won't. We view those as almost table stakes for a larger project because it's really expensive to finance without that. You know, you need a lease that compensates you for that cost of capital. We're focused on investment grade, either directly or via credit wrap. It's what we hope to achieve. It's the nature of the discussions we're having now. That's just a much lower cost of capital. That all flows through to shareholders. I mean, the more we can take down that cost of debt financing, the more there is for our shareholders.
Yeah. I think Jonathan summed it up pretty well. I'll just also add on top of it that, you know, when we're having these conversations, a lot of our sites are sizable sites. Really the demand and the appetite to take down a site is really focused on the hyperscalers and the neo clouds that do bring that kind of credit quality. For you know, smaller sites, you know, it could be a much smaller interest, like a much smaller neo cloud who'd be interested in that. We would still then be seeking some sort of a wrap or a backstop or something to help support the development of those sites and make sure that our contracted revenues and the site, you know, forward looking out five, 10, 15 years, is clear, consistent, and de-risked.
Got it.
Hey guys, I'm getting a few questions coming in from the lines here, and again, folks, please keep them coming. With that said, a number of people asking how, let's talk about data center pushback in the States. Where are you in tackling those issues specifically? And just to come back to the roadmap you articulated a second ago, where are you in getting approvals on, you know, upsizing what you guys have? You talked about that, I think that was number two on your list a second ago, Ben Gagnon, if I had it right. You wanna speak to that real quickly? Like where, like, what's your visibility on that, making that happen?
You know, we've got a pretty good line of sight on expanding power capacity at a number of our sites. You know, when it comes to the Scrubgrass site, there's the two buckets that I spoke to earlier. The detailed load study with FirstEnergy is currently active. That process takes roughly nine months, so we're expecting to have visibility on that in kind of the Q3-Q4 timeframe. Beyond that, you'd have to take a few more months to convert the detailed load study into an ESA and firm service.
The great thing about the Scrubgrass location is not only is that a really sizable application for power with 750 MW, there's a very high probability of success because there's so much power on that existing transmission line that even the power peaker plant that's located adjacent to the substation just 3 mi north of our site is almost never turned on and is more frequently used as a load bank than as a power peaker. It just kind of gives some support to how much capacity is on this high voltage line. The second part of this in terms of the ease of bringing those 750 MW and our confidence in them, it's only 3 mi to connect from the substation where we'd be interconnecting to our site at Scrubgrass.
Those 3 mi would utilize an existing transmission line and easement that we already own. You know, we don't need to go out and seek permits. We don't need to go and rework the paths over and over again. Everything's in place. The construction is incredibly simple. The distance is incredibly short, and the timelines are really easy to manage from there. When it comes to the nat gas component, which would add an additional 500 MW-550 MW of onsite generation capacity, we should be, you know, fairly clear in the coming months in terms of what the engineering would be to go and build a 4 mi connection. Just 1 mi north of the substation where we'd be connecting in with FirstEnergy, there's about the Tennessee Gas Pipeline.
This is the second largest natural gas pipeline in the United States. We've been actively working with engineering firms to develop and do the pipeline that will connect from that site onto our location to deliver 550 MW of gas. Three of those 4 mi would utilize the exact same easement, the exact same area that we have the existing transmission line, and would build the expanded transmission line. Again, incredibly simple with regards to infrastructure development, because there's almost a very short path with almost no barriers to development. We should have good visibility on both of these later in 2026. With regards to expanding capacity at the Scrubgrass location, there's two buckets there as well. The first is the conversion over of our ISA into firm service.
We have non-firm service in an ISA of 60 MW. We're looking to convert that into firm service. And really this is just a regulatory matter. The transmission line is there. It's currently active. We own and maintain that transmission line. We have the substation. There's no money that needs to be invested. There's not an inch of transmission line that needs to be constructed. It's really a matter of getting a stamp on some papers from regulators to be able to convert non-firm service into firm service. That could happen at any time this year. Then beyond that, we also have an additional expansion capacity with PPL, who's looking to develop new generation in the area.
You know, with energy bottleneck being the primary limiter on growth for HPC and AI, it was obviously a very welcome call for us to receive from PPL, "Hey, we're looking to invest in new nat gas generation to support data centers in your area. How much more capacity can you take on site?" That's the call that we received. Obviously a very good call to receive in this market. That's gonna take a little bit longer to develop because it's new generation, it's new capacity, it's new transmission. Most of these customers that we speak to, especially the hyperscalers, what they want is an immediate pathway to development, and they wanna be able to scale that site out for years to come. That makes a really, really attractive location for these customers that we're talking to have that ability to continue to expand over the years.
Julian, one thing we think about is, you know, if you read in terms of process, getting things through, if you read Governor Shapiro's four principles or you read the statement that came out from President Trump and the federal government, we're already doing all of those things. We are already working to all of those principles because our power is secured at each site that we have in the market right now.
A point of fact, what we're working on are the expansion opportunities, which are terrific upside for our customers and for us. If one already has an ESA and a site that is connected to the grid under firm capacity, if capacity is available, we'll get the call about whether or not we can take it on our site. It's very, very different than stepping into line at the queue for interconnections and for capacity. We're in a completely different place. Much more advanced. It's one we really like.
Awesome. Just a follow-up here. I mean, I got multiple people pinging me here, so real time here on. This is what happens. You guys have a real time call earlier. You've got results. You have an opportunity. Appreciate you guys back to back and working with us on this.
Yeah.
A number of folks pinging me. It's sort of all around the same nexus of timeline. Again, I perhaps opened Pandora's box by starting off with that pivot towards, you know, the milestones and timeline. People really zeroing in on 2027 and what you realistically can get done. What are you assuming here, more specifically, I think is the best way to say this, in the cadence of 2027, right? Considering that we're sitting here in 2Q 2026, of when some of this stuff comes online. Like you said a second ago, 2027 is like that inflection year. What are you specifically assuming comes online when, in as much as I think people are really kind of zeroing in on what you say is kind of a tantalizing opportunity, a year out?
Yeah. You know, I appreciate that, and timelines are obviously a big interest to all investors in trying to understand their expectations. You know, the timeline is probably gonna be the fastest on Moses Lake, so that's the smallest site that we have. It's very likely to be the first site that comes online in the first half of next year. Then with Sharon and Panther Creek is probably going to be coming online in the second half of next year.
I don't wanna get into you know, specifics on the exact timelines, but once we've cleared NTP, once we have those signed leases in hand, that's the point when we're gonna be able to give very clear guidance on specific timelines for when the projects are gonna be built to come online. You know, that's something that investors can look forward to, you know, over the coming, you know, couple of months, as we achieve NTP and we work towards those signed leases.
Julian, I'd also say we really view this year as an important inflection point as well. 2025, we discussed the intention to pivot to the HPC AI data center space, and we've executed everything that we needed to do to do that. We're only in North America. We're gonna be re-domiciled as a U.S. company and part of all the U.S. indices, et cetera. 2026 is about the commercial inflection point, coming to market with sites that will be ready to turn shovels at the time they are leased and executing on those this year. Even getting to that point from where we started nine months ago to say we're now in commercial discussions is, you know, almost transformational progress. Following this point, we intend to deliver in 2027 on the actual commitments we're making this year with customers.
Awesome. All right. JP, I'll let you jump back in here.
Okay. Yeah, I wanted to come back to that on-site power generation that you're doing. I'm just curious, the conversations with potential tenants, like are they viewing that as, or you guys view it as a way to get power online quickly, and then you flip over to a grid connection over time, and maybe that on-site power generation is for building the next phase? Like, how does that fit into it long term? And how willing are hyperscalers to sign a fully behind the meter, you know, type site?
I'll differentiate between sort of two aspects of that question. One, when we give capacity guidance around Panther Creek at 350 MW, that is grid capacity. We are not relying on any behind the meter capacity for that power. Similarly, when we talk about capacity available at Scrubgrass, that is grid capacity only. We are not assuming the use of the on-site plant to support the data center. We think that is the right way, it's the conservative way to talk to people about what's gonna get delivered on site. There's use for those power plants, but our capacity numbers are based on firm grid capacity. The expansion opportunity at Scrubgrass is both interesting and tangibly advanced in an important way around potential co-located CCGT generation, given what Ben spoke about and the proximity to the Tennessee Pipeline.
Like, there are not so many places where one can build a lateral that short for 500 or 600 megawatts of generation capacity on a colo site. Again, when we talk about megawatts, that is grid connected firm capacity. It does not include the plants. At Scrubgrass, we have both the firm connected grid capacity and the, I'd say, pretty tangible potential to co-locate a CCGT.
Okay. All right. That's clear. That's very helpful. Then, you know, pre-signing a lease, can you talk about the capital that you guys have put into the site, the capital that you've spent on. I assume you've been putting down deposits on a lot of the supply chain, that you need to secure. Like I guess, where are you guys at in that, and, like, what is required? I guess today, like where do you have to put hard money down, so a tenant's confident you guys can deliver?
It's an excellent question. First, important takeaway is we think there's a lot of value in having a strong balance sheet and a good liquidity position. Our plan is designed around making sure we get through leasing and well beyond with comfortable levels of liquidity and not needing to access the markets for additional capital. Great opportunity comes along, of course, we'll look at it. The point is the plan is constructed so that we do not need more financing to get through leasing and well beyond. That's critically important because the cost of capital falls dramatically, right, once you have that in hand. We're being really deliberate about that on behalf of shareholders. It's why we're keeping liquidity.
Soft costs to finish up what one needs to do to have a project be shovel-ready, they're just not that high. That's pure development. It's people sitting in conference rooms, quite frankly. That's, you know, you could spend if you worked really, really hard, you might be able to spend 5% of a project's total cost on soft costs, right? You could spend quite a bit more depending on what long lead time items you want ordered. What we're seeing is that long lead time items are actually sometimes not expected to come from us, but sometimes the customer expects to deliver them because they have a very specific point of view on what they want. You know, that'll be part of a sort of final CapEx timing and shape. Again, right now we're gonna get through leasing. It could be well beyond without needing any external financing.
Okay. Do you have any thoughts on the build cost per megawatt on what you do, or will that be customer dependent?
That's gonna be customer dependent. It's a great question, including because of the point, you know, we just discussed, which is different customers are gonna have different requirements f or the data center. They have different design specs, like, what it costs to build depends on what the customer wants, and then the lease will work against that build cost, right? That it actually makes sense for us to go ahead with the project.
Yeah, I think that's one thing that the market probably doesn't understand is how much difference there is between like a Google data center, an Amazon data center, a Meta data center, a Microsoft data center. You know, a data center is not a data center the same way a megawatt is not a megawatt. There's huge differences, and while standardization is starting to come into the industry in a more aggressive way, there's still huge discrepancies amongst most of these customers around what kind of equipment they need, what kind of, you know, redundancy that they wanna do, how they want to position themselves, or what kind of, you know, features that they want out of that site. It could get really pedantic or it could be very, you know, high level. It comes throughout the entire stack.
Yeah. While we right now need to know what our customers want to estimate build costs, you can certainly look at our peers who are announcing leases and data centers of different flavors and get a sense for what the market views as build costs right now.
Okay. I guess before we move off financing, can you talk to the extent you have a insight into it? Can you talk about the construction loan market right now? There's obviously been you know, a lot of news articles around private credit and all that. It's been a little while since we've seen any of your peers raise capital. I mean, they did a couple months ago, but the world's kinda changed since then. Any thoughts on like where financing costs are today?
Yeah. It's really interesting. It's part of why we're prioritizing working with investment grade counterparties or counterparties that can bring a credit wrap to the table that's durable and back-to-back with the lease. That enables a company to either A, work in the conventional bank-led construction and project finance market, which has the advantage of being you know very flexible on the draw, well-established, a lot of liquidity or as we've seen in the market, you get new paths of capital formation being created. Now, the high yield market is becoming quite competitive economically with the conventional construction market. Sure, each has different advantages. High yield offers much more flexibility because a construction loan is very covenant intensive. High yield is much more flexible. You have some negative carry costs, and each company will decide for themselves what's the right structure.
From a market perspective, the costs of those have converged in a really interesting way, you know, right? There's a pretty modest spread we see generally between the two. It widens and narrows, but we think there'll be more innovation on the debt side as well, and our plans haven't relied on the private credit market. I think if one is trying to finance something a little off-piece or smaller, maybe that's an interesting market. You don't wanna. I'd say it's not our first choice for financing a large data center project.
Okay. I guess on the general contractor side, do you guys have a general contractor lined up at your sites that is gonna work on the construction side once you sign a lease? Just talk to us about that process of getting to that point, 'cause labor is very constrained right now.
We've got great partners across all of these sites, and we're actually, you know, unveiling our new website tomorrow for keelinfra.com. On that new website for each project that we have, we do list the partners that we have for each one of those sites. So everyone's gonna be able to go onto the website and see, you know, who's our owner's rep, who's our GC, who's working on architecture and engineering. What you're gonna see on that website is, best of the best names.
You're gonna see names like Turner Construction, which is number one. You're gonna see names like Corgan Architecture and Engineering, which is number one. You're gonna see names like Vertiv, which is number one in supply. You're gonna see best-in-class names, and all of that information is gonna be there on the website for investors to be better informed about who we're working with and what that means for kind of timelines and expected outcomes.
I'd say just two supporting points. One, we think it's worth the investment to work with the best companies in any given area because that's what our customers are used to seeing. We think that matters. Two, this is again, there's a difference between a Texas megawatt and a Pennsylvania megawatt because we're not competing for precisely the same person.
Yeah.
To be honest.
Okay. If there's any part of the supply chain that is most likely to cause delays with data center development, what do you think it is right now?
You know, even if I could give you an answer to that, Jonathan, it would probably be different tomorrow. That's just the nature of how fast this market is moving. You know, I think that the key thing for any developer is making sure that you stay agile and flexible in these developments and you've got robust supply chain capacity. Even if there's an issue somewhere, you're always able to flexibly pivot elsewhere. I think we've got the procurement team in place. I think we have all the right partners in place, that we don't currently expect any sort of issues right now, but we're always planning for the worst and making sure that we've got those redundancy plans in place because who knows what tomorrow, you know, holds.
Okay.
Hey, guys, just another question coming through here just related to that. We've certainly heard from day one a lot of talk about this. What about liquidated damages? How do you think about, you know, being subject to those kinds of commitments here? Again, that obviously comes with contracts in tandem, but how do you think about, you know, the risk side of damages in the event of any delays on what you're looking at right now, especially performing for 2027?
This comes into a huge part of our strategy on how we go to market and why we've been so, you know, consistent in saying we need to de-risk these sites through permitting and through execution, and that's really when we're gonna be able to get comfortable locking in timelines. We understand that, you know, there are significant, you know, challenges associated with timelines and with deliveries and the, you know, one of the biggest things that anyone can do is make sure to de-risk and advance the sites as far as possible before you commit into those leases, because that's just gonna help you to minimize the timelines, minimize the uncertainty, and execute with conviction.
You know, there's other things that you can negotiate with some counterparties, but I think really, you know, we address this through strategy and making sure that we're not locking ourselves and committing ourselves into specific dates until all of the pieces of the puzzle are put together, are lined up to help minimize and manage those risks.
We spent just a minute on Bitcoin. Can you remind us, like, how much mining revenue do you still have coming in? I don't know if you guys disclose. How much Bitcoin do you currently hold, if any? Or do you sell it as soon as you mine it?
We haven't been actively selling our Bitcoin. We have a little under 2,500 Bitcoin on the balance sheet right now, fully, you know, owned, unencumbered. You know, our view is that as we continue to wind down the Bitcoin mining exposure, we'll also continue to wind down the Bitcoin exposure. You know, we're not a Bitcoin company anymore, Jonathan. We're Keel Infrastructure. We're purely focused on HPC and AI. We're gonna be looking to continue to wind down that position with, you know, discipline into strength so that we can try and capture every value, all the dollars that we can, so that we can reinvest them into HPC and AI.
Well, I guess my question is, like as you transition, you know, and into Keel Infrastructure, like, I guess, how much revenue, how much cash flow can come in by just kind of bleeding those miners dry over the next year or two? 'Cause I imagine a year or two, you just won't have it anymore. Just curious, like how much more revenue is there? Realizing I'm asking you to project the Bitcoin market, which isn't possible. Just as a run rate, how much should we think about that?
I'll jump in here, you know, kind of at a high level, then I'll let Jonathan maybe provide a little bit more insight. You know, for us, we plan the entire company and the balance sheet around not needing to rely on the Bitcoin mining revenues to fund the business for OpEx or CapEx. We have, you know, raised a tremendous amount of money last year with the convertible, which really positions us well to execute this year. Regardless of what's happening with Bitcoin economics, you know, our site projects are unimpacted, our ability to execute is unimpacted, our team continues to run, and we've got all the financial liquidity and flexibility to move as fast as we want and as fast as we can. Jonathan, do you wanna add anything to that?
I'd just add that when we talk about being really comfortable with our liquidity position through lease signing and well beyond. That assumes that we are not making any tangible cashflow out of the Bitcoin mining business. We're being really conservative in how we build our liquidity plans.
Okay. One other kind of random question here towards the end. You have sites in Canada, so just curious how, like, data sovereignty and privacy, like, if that's impacting any sort of leasing discussions you might have north of the border versus in the United States. Is there, like, a different kind of customer that might exist up there, different regulatory environments around what the data center's being used for?
It's a great question, Jonathan, and data sovereignty is a big topic north of the border. It's, you know, true at the federal level. It's also true at the provincial level. There's legislation in place which really means that a lot of Canadian enterprise, and especially a lot of Canadian government agencies, have to have their data in Canada, and in many cases have to have their data specifically in Quebec. Given that this is a really high barrier to entry market, I mean, just in the last year, Microsoft, Amazon, Google have all been denied power applications for new data centers in the province. Having the power that we have, having the ability to develop HPC and AI data centers is a huge value add.
You know, from the market information that we have from CBRE and others, it does translate into higher lease values than the United States because there is a stronger drive to enforce the demand to stay there with legislation, and then there's just significantly higher barriers to entry to actually grow and deploy those megawatts. It does translate into higher rates overall.
Okay. If we think many years into the future, and we can end on this one, so your sites are more northern, right? Canada, northern part of the United States. Like, as you kind of think 5+ years into the future and where you might want to acquire land sites, are you guys kind of writing off the Texas and Southeast markets and letting other people compete there, and you view your differentiator as being in places like Pennsylvania and Washington and kind of some of those more higher population areas?
You know, our view has always been we wanna go to energy markets where the energy capacity is far in excess of the demand and the transmission capabilities, right? Our view is always that those were the areas that would give us natural hedges on escalating energy prices, and it would also ensure that there was the most natural economic incentive for us as a developer and for the local communities to have a win-win sort of scenario. That's how we, you know, grew in Quebec. That's how we expanded into Washington. That's what we did in Pennsylvania and other markets. We'll continue to take that same philosophy and that same approach, right? We specifically avoided Texas in the past because we didn't want to be in, you know, markets where the only
You know, the power was incredibly volatile, and the only real application that we saw was the energy trading that came from Bitcoin mining. We wanted to invest in areas where the energy had applications beyond Bitcoin mining. That's what we did. That's what we'll continue to do. It doesn't mean that we'll stay out of Texas forever. The markets are constantly evolving. They're constantly adapting. We're always gonna be looking for those best opportunities to add powered land and continue to grow the portfolio, but that's kind of the strategic framework that we operate under. Let's find the sites that have more than one application. Let's find the sites that have natural hedges in place. Let's find the sites where there's the biggest win-win between us developing the sites and those areas being able to monetize more energy than they otherwise could.
Yeah. For us, as a practical matter in the timeframe you're talking about, it's what we know really well, which is Pennsylvania, and then moving west across the map. Some of those are in different RTOs, but we're gonna, we try and be very conscious of where our experience lies and when it doesn't, and our strategy is built around executing on matters where we feel like we have a distinct point of view that we can use for shareholders in a positive way.
Perfect. It's very clear. Guys, thanks so much. Really appreciate your time. I think it's a great place to end it. Ben.
Yeah
Jonathan, I really appreciate you guys being here.
Perfect. Thank you, guys.
Thank you, Pete.
Thanks to-
Good luck with Keel.
Thank you, everyone, for tuning in.
All right.
Thank you.
All right. Have a good day.
Bye.