TeraWulf - CFO Patrick Fleury Q&A - Anthropic Deal Announcement!

We sit down with TeraWulf CFO Patrick Fleury right after the Anthropic announcement to break down the deal terms, the financing structure, and what it means for long term shareholder value. We also unpack why TeraWulf is selling the Abernathy JV stake, how site selection drives terminal value, and where power markets like Texas can mislead investors. 

• overview of the two announcements and how they fit TeraWulf’s priorities 
• rationale for selling Abernathy and recycling capital into wholly owned sites 
• Hawesville lease structure with Anthropic including 20-year term and escalators 
• investment grade credit support framework and why it matters for project finance 
• yield on cost versus WACC and why leverage discipline drives equity value creation 
• valuation walk-through using NOI, cap rates, and per share impact 
• PUE strategy and designing capacity around peak versus average conditions 
• negotiation complexity and why leases and SLAs keep getting longer 
• terminal value thinking based on latency, population radius, and grid power economics 
• ERCOT realities, interconnection queues, and the risk of power assumptions 
• who owns technology refresh costs and why maintenance capex assumptions get mis-modeled 
• labour constraints for electricians, regional execution, and EPC partner selection 
• pipeline outlook for Morgantown, Eastern Kentucky, and potential Europe expansion 

Let us know in the comment section below if you're currently holding shares of WULF and what you thought about yesterday's news!

Anthonys Info:

Website:

https://www.powermininganalysis.com

X: 

https://twitter.com/cazenove_uk

Patreon:

https://www.patreon.com/Cazenove_UK

Bryce's Info:

X:

https://twitter.com/McnallieM

Patreon:

https://www.patreon.com/McNallieMoney

Merch:

https://shop.mcnalliemoney.com/

Website:

https://www.mcnalliemoney.com

Business Inquiries:

Email - powermininganalysis@gmail.com

Transcript

Welcome And Big News Setup

SPEAKER_01 0:00

Hey guys, welcome or welcome back to the channel McNally Money, the official home of power analysis. In today's episode, Anthony Power and I are pleased to welcome Patrick Fleury back to the channel once again. He's the CFO at TerraWolf, hot off the heels of yesterday's huge anthropic announcement. We've got a lot to talk about. Before we get into it, take a second, smash the like button, guys. Big help to myself and the channel. If you're not already subscribed, McNally Money, feel free to join. And let us know in the comment section below if you're currently holding shares of Wolf and what you thought about yesterday's news. With that being said, let's get into today's interview. All right, guys, away we go. A highly anticipated interview on the show today. We've got Patrick Fleury back by popular demand and hot off the heels of yesterday's massive doubleheader announcement, including a landmark deal with Anthropic. Uh, Patrick, CFO at TerraWolf, of course. Thank you for making the time and great to see you. Yeah, thanks for having me back. Hey, we're looking forward to this one. We saw the news came out yesterday. We were covering it on the podcast. We said hopefully we can get a member of the team on. And sure enough,

Breaking Down The Two Announcements

SPEAKER_01 1:13

here you are. Can you walk us through at a high level the two components of yesterday's press release? Uh, and then we'll obviously dig into the details.

SPEAKER_02 1:21

Yeah, absolutely. Look, super excited about all this massive lift from our entire team. Uh, so yeah, let's let's dive into it. So, touch on Abernathy first. Um, first of all, I just want to say we have maintained an excellent working relationship with Fluid Stack and all the folks involved there and at our Lake Mariner site. As you know, Fluid Stack is our tenant at Lake Mariner. We're effectively married, you know, we're with them for at least 10 years. Um, and uh yeah, just remain super supportive of them. The sale for us was was really, you know, gener it was really a focus on our priorities. And so if you kind of back up, what does the sale do? Sale results in uh cash return to us. So if you remember, we put 450 million in, we're getting about 530 and change out. So that's like a 20% return on invested capital. That allows us to recycle capital into wholly owned projects. As we've grown, we've added 2.5 gigawatts right across three sites in the past five months. And all of those sites have optionality for future on-site power generation and expansion. And so when we kind of looked at our priorities, the JV structure just required a lot of time, particularly from me and my team, because TerraWolf was the financing lead and um Fluid Stack was the development and operational lead. And you know, right now, time is our entire team's most precious commodity. So I want to focus and we want to focus on larger sites like Hawesville or eastern Kentucky for us or Morgantown, that those sites, right? A 500 megawatt site generates over 10 billion of value for Wolf shareholders. And I'll get into that in a minute. So that's number one. Number two, if you consider the velocity of capital and what I would think of as like bang for your buck in each of these deals, at Abernathy, right, we invested 450 million for 84 critical megawatts. So that's basically like around five, just over five million per megawatt. Okay. At Hawsville, we're gonna invest, call it about 850 to 900 million of equity at our typical kind of 80-20 debt equity, and I'll get into that in a minute, too, for 401 critical megawatts. So that's 2.2 million per megawatt, right? So that's like two and a half X bang for the buck in a wholly owned site. So that's what we like. That's where I want to focus on. You know, that really moves the needle for uh wolf shareholders. So now that's the Abernathy deal.

Why Sell Abernathy And Recycle Capital

SPEAKER_02 4:22

Let's let's now turn to Hosville, okay? The key elements of the Hosville deal, number one, direct lease with anthropic, who's the end user of the compute. Number two, 20-year lease term with annual escalator. Three, uh, framework for IG credit support, and then four, you know, mid-teens unlevered yield on cost. Okay, so let's let's dive into each one of those. Direct with anthropic. Super proud of that, right? Like we have a terrific relationship with them. They're an amazing company to work with. I feel really privileged to be riding side saddle with them. They're one of the fastest growing companies in the history of the world. Their products are incredible if you've ever used them. And I think you know, they've got you know some visionaries in that company top-down. So could not be more proud of that. Um let's dive into in the press release, we said that there the the uh lease payments will be backed by investment grade credit support. Okay, and we chose that language very carefully, and I want to get into that because one of two things is going to happen. First off, Anthropic, just like OpenAI and others, uses three primary chip providers: Nvidia, Amazon, and Google. One of those three will be providing the chips for this data center. So we will either do something similar that we've done in the past, where one of those investment grade providers provides credit support to the project because they're selling tens of billions of dollars of chips into the project. Or if we decide not to do that because Anthropic gets public, similar to SpaceX, and similar to SpaceX has a strong investment grade credit rating, then you don't need the investment grade credit wrap, you know, from one of the bigger guys. And honestly, this point I think is really important. As you guys know, we've always tried to be front foot forward with all of our partners on developing new structures. And the space is moving at light speed, it's maturing, it's becoming more financially um credit worthy, all the counterparties are. And so I think there's a natural evolution in the financing markets and in the credit wraps. Because when you look, you know, Amazon, Google, you know, Microsoft, NVIDIA, they don't want to wrap deals if they don't have to. That's not their business. And so they're doing it right to kind of get their ecosystems off on the right foot. And now a lot of those ecosystems are like, you know, they're able to swim on their own. So I do think there is a natural evolution here that happens. It's not, we're not quite ready for it now. But if Anthropic goes public and has a multi-trillion dollar market cap and a really strong investment grade rating, which could happen pretty quickly, then we might be ready for it. So again, we this is just a I think there's two points that I want to call out that again, I want to tip my hat to Anthropic and to us on being creative and just saying, hey, we have time. And by the way, I want to come back to one thing. As you guys know, folks were kind of scratching their head when I went to the equity markets preemptively and issued equity. Okay. That was for two reasons. Number one, it was to give us an advantage negotiating with the final customers to get to the finish line on a lease because it was funded. The hardest part of the deal was funded. And two, it was precisely this point where I wanted to have enough runway to let things play out. So now, for example, I can fund this project through the end of the year. I do not have to go to the equity markets, I do not have to go to the debt markets, and I can let the evolution of the space, including anthropic potentially being public, play out. So again, that was on purpose. I just couldn't tell you exactly why I was doing that at the time, but that's the purpose. So now we are in the catbird seat. I do not have to go to the equity markets. The equity markets, by the way, are a mess over the past week, for and we can get into that in a minute for reasons that I think are kind of foolish, but whatever. It is what it is. The good news is I don't have to, I don't have to go to the markets. So that is one

Hawesville Lease Terms And Credit Support

SPEAKER_02 8:54

piece. Now let's talk about economics, again, which I we are really proud of. Okay. Total contracted revenues of about 19 billion. That means average annual revenues of you know 930-ish million, 85% net operating income margin here. This is a gross modified lease, it is not triple net. So again, if you kind of stick with that guidance that I've gave given in the past, you know, that means average annual net operating income of about 800 million. The lease also has two five-year extensions. If you assume those extensions are exercised, it's over 33 billion of revenue, right? I mean, these are huge numbers. So when you when you dig into, and we haven't you know put this in writing yet, but I'm gonna tell you now. If you remember, as we've kind of moved forward in time, we've given CapEx guidance. Our original guidance on Core 42 was like six to eight million per megawatt. Our guidance on the fluid stack leases at Lake Mariner, eight to ten. Here we're saying 10 to 12. Okay. Um, if you take the midpoint of that guidance on 401 critical megawatts, it's 4.4 billion, right? 401 times 11, 4.4 billion to build. That means, again, if you take the NOI number I just gave you, about 800 million of NOI on 4.4 billion, right? That's an 18% yield on average. Like we've told you in the past, we try to go get mid-teens yield on cost from our customers. And that is simply because, and I think this is lost on a lot of people, so I want to be really clear. When I go finance these projects, I I was saying internally where I say, just because you can doesn't mean you should. What I mean by that is just because you can finance a project with 95 or 100% debt does not mean you should. Because if you have no equity layer in a project or in your company, when you stub your toe, it becomes an amputated leg. Okay? If you have a comfortable equity layer, that's not the case. You have room to operate. So you will likely see us kind of stick with reasonable equity layers in our deals. Does that mean 10% or 20%? Yeah, probably somewhere in that range. But I don't think you're gonna see Terror Wolf go do a 95 or 100% debt financing. Because again, even with 10 or 20%, it's an amazing return on equity. And let me explain why. The reason why the credit support is so important is when you think about whack, like what is my weighted average cost of capital? If I finance 10 to 20 percent, let's make numbers easy. 20% of a project with equity, my cost of equity is probably 25%. Okay? So that's five points to whack. Okay, 20% times 25 is five. 80% of the debt or of the financing is coming from debt. If that debt is six percent, okay, eighty times six is called 4.8. Let's round up. Let's say it's five. So five points from equity, five points from debt means my whack is 10%. By definition, my lease yield on cost has to be greater than my whack for there to be equity creation and equity accretion. So I, you know, I'm watching some of our, or I'm not gonna use the word peers. I'm watching other folks in the space execute leases where it looks like the lease is a pretty low yield on cost and your whack is effectively in excess of the yield on cost. By definition, that means there's no equity value creation, period. So until you can lower your whack, like you're not creating equity value. So again, that, and we are very transparent with our customers about that. That's where we are today. And by the way, if my whack changes, then my yield on cost that I'm seeking can come down. But that is the focus for us. And so if you think about where the value creation here really is, right? If I'm getting that back to what I said, about 800 million of NOI and 4.4 billion of capex, that's about an 18% yield on average. If that now, when it's fully leased up and built, trades at a 6% cap rate, just take 800 million of NOI divided by 6%, and that's roughly $13 billion. So something that cost me $4.4 billion to build is now worth $13 billion. That's $9 billion roughly of equity value creation to Terrawal shareholders on about 600 million fully diluted shares. That's $15 a share. It's just as simple as that. And so that business and doing it at big sites like Hawsville, 500 megawatts to 1.5 gigawatts, that's where I want me and my team focused. Um, and that's what we're super excited

Yield On Cost, WACC, And Value Math

SPEAKER_02 14:23

about. There's one more element of this deal that I think is super creative that I want to again tip my hat to Anthropic and to um my operations team working on. And that is if you look here, right? We have we said the site is 482 gross megawatts, 401 net. If you just do that quick math, it's a 1.2 PUE. Yeah, which for that part of the world, right, it's not Lake Mariner, where you're on the shores of Lake Ontario. So, you know, we've gotten some questions, and guys are like, well, how are you doing that? And again, this is where, you know, again, Sean Farrell, Nazarkhan, like my whole team, and and their experience in building the data centers at Lake Mariner and working with Anthropic. What we're doing is we're basically overbuilding relative to peak PUE. So meaning the peak PUE is probably four to six weeks, right? Where it's really hot between you know noon and 8 p.m. Um, where you know the critical megawatts are constrained. So we basically structured the agreement um to create extra capacity that'll be available like 95% of the year. So Anthropic will be able to get more compute out of that site. So you're not just building to what is the worst PUE, you're building to more like what is a more realistic, like average PUE throughout the year absent those like peak times. So I think that's like sounds pretty obvious to you right now, but it you know that I think hopefully it's only because I'm explaining it that way. But you know, that took a lot of brain power and I think really creative answer um from our operations teams and and anthropic to get more out of uh out of less. So that's a quick description of the deals. Um, you know, happy to kind of dive in if you got more questions.

SPEAKER_00 16:13

No, that's that's that's a that's a great uh that's a great summary of of what what's uh what's come out yesterday, Patrick. I I mean before we just get into those numbers because they they are extraordinary numbers, we we talked to you after the first deal, and you talked to me about the time you spent in that room with your team, the negotiating, the five or six months round the clock. Since that deal, I mean and what you've learnt through that deal and subsequent deals, can you sort of like give us a flavour of how the discussions with Anthropic and the timeline um to get to to where we are

Deal Speed, Site Selection, And Terminal Value

SPEAKER_00 16:49

now? And bear in mind, and I believe um Van Eck had on their register that you know you that there was anticipation that that both um uh Terra Wolf and BitGear were going to announce in the by the end of the second quarter. So the surprise of the announcement isn't as big a surprise because that's what the expectation was. But certainly from a from a time frame, are you seeing the fact now that you're you're getting far more experience and your teams, you know, they understand the the sort of like the the negotiations are going in a more seamless manner, or are you still finding every new client is going to be you know go through that challenge again uh to get the right deal for both parties?

SPEAKER_02 17:28

Yeah, great question. So uh I mean if you take if you just back up for a second to answer that question in a roundabout way, Anthony. You know, we've been working on Hawsville uh, you know, probably since fall of last year. We closed on it, but we were ready to close kind of before year end, um, but the seller had to work through some some some things. And so, you know, we closed February 2nd. So if you think about it, from February 2nd to now, that's pretty extraordinary, right? To kind of to A buy a site and then lease it up in that amount of time. So that um, yes, I think is a function of a couple things. Um one, first off, like the process there was very robust. And so lots of different parties involved in in diligencing and and bidding on that capacity. I I was surprised by that. I I think my operations team, you know, who's you know smarter than I am, they weren't. Um, and part of it is you know, when we look at sites, you know, we discuss terminal value constantly in in our company. And I can go into like it's terminal value on customer, who's the customer, what's the right multiple, what's the right credit support, um, terminal value on site, meaning we draw a 300-mile radius around the site, and we look at how many different cities and the total population within 300 miles, because that's about three milliseconds. Hawesville has 20 million people within three milliseconds, six NFL cities. Honestly, not all that different. Lake Mariner, like you draw the same thing, you've got New York City, Toronto, and Boston that again, 20 million people. So, our view is like those sites where you have that type of connectivity to that many people, um, and you're buying grid power. Our view is those sites are gonna have more terminal value than you know, a site in the middle of nowhere that's using behind-the-meter reciprocating generators that has a power cost of $100 to $150 a megawatt out. That I just think you might get leased up, but in 10 years, like we don't see the terminal value there being durable because the reality is having been in the power business for 26 years, the US will get its act together, we'll it'll be all hands-on decked, we'll build more solar wind, we'll build more SMRs, we'll build more nukes, we'll build more gas fire generation over the next five to 10 years, without question. So we'll rise to the occasion, right? Because it will, it this is about national security as well, and we'll build more generation. So if and and the average price of you know wholesale generation in the US, no matter what market you're in, it's probably around 80 bucks a megawatt hour. So if you're in a place where the price of power is like 100 to 150 long term, that's not going to be competitive, period. So, but back to back to the sort of competitive dynamic here, really competitive process. We wanted it that way because there are some folks, right, that we haven't danced with yet that we did dance with in this process that really liked this location. And then what do you know? We come out and announce Eastern Kentucky, which is a 500 megawatt site, ramps up to a thousand. So it starts about 500 in late 28 into early 29, goes up to 1,000 in 2030. It's actually a like almost the same exact location, 300 mile radius, like the same amount of people. So now we take a bunch of those bridesmaids from that process and we just move them over to eastern Kentucky and we're off to the races. I think that's lost on a lot of people because if you think about it, right, between Lake Mariner and Hawtsville, I have roughly a thousand megawatts contracted. At Eastern Kentucky, where I had a really robust process like just west of there, and now I've got folks working on that, that literally doubles the size of the company again. So I just basically doubled the size of the company with Hawtsville, and now I can basically double it again by getting another thousand megawatts, because my view is whoever takes that first 500 megawatts is gonna take it all. So that's pretty extraordinary. Like, that's why I wanna focus our team's time on those assets that really moved the needle for Wolf and its shareholders. And so back to your point, sorry I'm getting there, Anthony. But look, these processes have become hard, you know, people ask me all the time, why does it take so long? Like these, you know, the lease is 150 to 250 pages. The SLA addendum is another 80 pages, right? Like this stuff is hard, it's complicated. Um, folks on both sides are getting more sophisticated, right? I think if you kind of looked at the evolution of our leases internally, they're very different. And the level of sophistication and complication is only increasing, right? So um, yeah, and it and and again, and then you you have to take a partnership approach to these things, right? It's not like like anthropic, as an example, like they want us to succeed, they need the compute, right? So you gotta find a balance, right, when you're doing these things where you know you're locking hands and and you're in it to win it together. It's not, you know, somebody that's that's you know not constructive on a partnership basis, right? And so, or that's playing like gotcha, like that, it just doesn't work. You there's a lot of um, you know, there's just a lot of changes along the way as you're designing these projects because our customers are getting data inputs every single day. And and to the extent that we can, you know, modify designs every single day along the way as we're delivering capacity to make it more efficient for the customer, we want to do that. So um, yeah, so I think Anthony, to your point, you know, you kind of find your cadence with your partners, um, and you try to find that with potentially new ones as well. But that it just you know it takes reps. You just have to go through it uh a certain amount of time. And, you know, thankfully for you know, for us, it's different. Like we have like one or two sites we're focused on. You know, our big partners like anthropic or you know, any of the hyperscalers, if you think about it, like they're focused on, you know, a dozen at one time. So that's the other thing. It's like it's hard to get momentum, right, on both sides. Like I have to want to do it, but then my partner, you know, who's got a lot more going on than we do, has to want to do it as well. Um, but again, that's why I think Anthropics is such a privilege to work with them because they are super creative, they are very uh just a terrific partner. Um and I think you know the results of what I just went through kind of highlight some of that creativity.

SPEAKER_00 24:32

Just to clarify on the on the 19 billion that you mentioned at the start of the conversation, um is that excluding the escalator formula or does that incorporate the escalator formula? No, it's not. That excludes it. So going back to when you said 900 um million per year in revenues would suggest one of the from what I'm looking now is one of the best um contracts that we've seen out there in the in the domain at the moment. I mean, if we go back to some of the previous contracts, that people were talking about you know one 1.9 million per megawatt, getting towards two, then we've we've seen a couple breach just breach two. This one, and I I looked at what Cant put out yesterday, and they came out with the same number as me, um, at 2.37 million per per megawatt based on 19 billion at 401 megawatts over 20 years. I mean, are you seeing so that the the the because of the demand for power, those that can provide it and those that have got a back, you know, a background in delivering are going to be able to attract great rates going forward because it looks like they're starting to increase. And this came out from a few CEOs saying we don't want to sign any deal, we want to sign the right deal. And you know, you guys have come out now, and and I think this is for me, this is probably one of the biggest deals I've seen there. And I've looked at some of the companies, you know, like uh Applied Digital, and they've got hyperscaler deals, and theirs is coming out at about 1.6 to 1.7 million so per megawatt. And so there's a there is a disparity in some of the deals out there, but it looks like you you've got a cracking deal on paper there.

SPEAKER_02 26:11

Yeah, look, I'd say thank you. I mean, I think Paul Paul said rightfully so, we won't you know we want to sign the right deal. Look, I I know there's a tendency to compare all these deals across one another, you know, they're they're all different in some regards, right? And I, you know, again, I I I I don't really care what other people are doing. Like, I just told you like how we like the financial standards that we hold ourselves to and what needs to happen, right? You know, what other people do is their own prerogative. Uh but yeah, look, uh there's just so many elements that that go into these deals. But to to break apart your your your questions, Anthony, 19 billion is total across the 20 years. Obviously, a longer lease term, if you have an escalator, results in like you know, uh higher numbers, and those numbers compound significantly, as we all know. But it but so is how does it compare to like you know the other recent ones in the space? You know, I don't I think it's hard in a vacuum to compare them all, but look, the the power is like the two scarcest commodities, in my view, in the space are power and capital. Yeah, like for everybody, not just for me, and maybe other than like you know, the four five big hyperscalers that can go issue debt. You know, one of them's in the market today, issuing um a bunch of securities, including a 40-year bond, which is pretty incredible. But the that's my the two focuses for us are power and capital. If you have both of those, like if I have both of those, then I can succeed.

SPEAKER_01 27:47

Hey, uh, two-parter for me, one from the subscriber base. I know you mentioned you're still uh going through the review process in terms of who will provide the chips. They were asking about whether this will be Grace Blackwell or Vera Rubin compatible in terms of the build-out. Do you have any insight on that at this point, Patrick?

SPEAKER_02 28:05

Uh nothing that I can share other than I would say again, the designs are getting more sophisticated, right, with us and our partners. And there's a couple of tweaks, um not as many as you would expect for different technology. So, for example, like at Lake Mariner, we are now running AMD, NVIDIA, and TPUs. So we got it all. So we we you know we have experience running all of those. Um, I think the only one we're not running is is Amazon currently. And so we you know, we know what each of them requires. We know the capex differential between each of them. I would say the one big difference in capex kind of across the board um that we're seeing is some require UPS, uh, some don't. That is you know about a million dollars a megawatt in in capex differential right there. So that, but other than that, I think we're getting to a point where um you know we at least we internally feel good on the design basis that we can kind of design and meet, you know, anybody's technology.

SPEAKER_01 29:17

Gotcha.

Texas Grid Reality And Latency Tradeoffs

SPEAKER_01 29:18

And then the second one here, uh, talking about uh something I think you actually brought up on the channel initially, the ERCOT market. So I know you mentioned you guys got a great return on the Abernathy site. Anthony and I also felt this was strategic in a geographic sense. We've heard from yourself, we've heard from Ben Ganyon that look, Texas is great for training, it's good to get permits, there's a lot of power, but as the industry evolves and gets more into inference, that latency becomes a lot more important, that uh geography close to populations, like you just alluded to. Can you talk to us about that shift and how that impacts the pricing power of maybe your sites versus someone located in in ERCOT?

SPEAKER_02 29:58

Yeah, so just to be clear, so we, you know, we still we love the Abernathy site, um, don't get me wrong. And it's it's not in ERCOT, it's in Texas, but it's in SPP. Big difference. Okay. So I always like to frame ERCOT for people. And yes, there is a reason we're not there. I will say we have looked at six to 12 very large, like anywhere from a gigawatt to five gigawatts of both grid and behind the meter sites in Texas, and we've not done one of them. And I can get into why. Um, but just high level for people that have not been in the power business for 26 years. Now I'm really gonna bore you, so I'll try to make it interesting. To put into perspective what is happening in Texas, okay. California, the state of California on a standalone basis is the fifth largest economy standalone in the world. California has an 85 gigawatt roughly grid. Texas, okay, has a roughly 95 gigawatt grid. And it's because it's a lot peakier, right? You have three months a year where it can be like over 100 degrees for for literally three months. And so that demand is super high in the summer, and then it kind of falls in in the shoulder months and and in the winter, absent some you know, crazy storms. In Texas, you have over, and I don't even know what the exact number now is. I think it's around 340 gigawatts, but let's just say 300. You have over 300 gigawatts of large load interconnection requests on a 95 gigawatt grid. That will not happen in my lifetime, period. Like, again, like nukes, like the biggest nukes in the US are like 2,500 megawatts, right? Like I mean, just putting that into perspective, like that is a massive amount of power. So literally not gonna happen. So I think unless, and this is why the focus on this whole batching process in EarthCot is is rightfully um so on like batch zero, because I think if you're not in batch zero and you have a line of sight to like, I mean, good luck. Like, yeah, again, I think it'll be like decades until you get your power. So, yeah, we're again like SPP and that Abernathy site, we love it. I think it's a great site. Wish Fluid Stack and all the folks there the best. Um, but we are not in ERCOT. You know, I'm not saying don't never say never, um, but uh, we don't like those dynamics. And again, as you guys know, we know what we know and we know what we don't know. We know power. That's what we're really good at. Um, and you know, we don't haven't found the right opportunity in ERCOT.

SPEAKER_00 32:53

Patrick, I'm gonna cover a couple of questions on on cost. And um, this deal we've seen now is I mean, that the the total length, including the options, is very similar to what others are signing on the market, you know, maybe 15 plus three times five years, total 30 years. This seems like the longest at 20 years for that initial contract. Um, when you have your discussions about uh um, you know, um what happens when uh the advancement of technology changes during that 20-year period? So we know that GPUs tend to have a life maybe five, six, or seven years. You may have to change uh at least twice, maybe three times in that period in terms of the the fit outs. What what's the what's the discussion like in terms of as technology changes, the racks that you're uh you're you're putting in in those buildings um may need to be um you know uh tailored to to fit future uh technology. How how does that conversation go with with your with your partner? Because at the end of the day, you've already said you both want this to succeed. Is that more of an amicable discussion going forward? Um, you know, uh it's this is that this is sort of like you know, a future challenge that will affect everybody, but how you how you deal with it in those conversations now probably helps you.

Tech Refresh, Maintenance Capex, And Mispricing

SPEAKER_02 34:13

Yeah, for sure. You know, and it's funny, Anthony, you just you you triggered something, so I'm I'm gonna go there on this because I I just saw Jim Chanos, who who sadly like when I was growing up in finance, he was like this he was a king. I remember like, you know, again, I started in the business in 2000 and he was around and and um you know one of the early short sellers, and he just put out some nonsense on us last night. I mean, the guy, like literally, I read it, and I'm like, you're an idiot. And by the way, if you want to be short or stock, God bless you, but at least call me up so I can tell you what you're doing wrong in your model, because it literally is nonsense. And so one of the things he did in the model that just triggered this, you know, me reminding it, is he's modeling maintenance cap acts of like a million to a million and a half a megawatt. Like, I like I think on an annual basis. So, because I looked at his model and he's got 8 billion of DNA on a $4.4 billion built. And I'm like, hey, genius. The last time I checked the tax code, you can only depreciate, like, not even the full amount, like it cost me $4.4 billion. About 80% of that is depreciable, like roughly. I'm like, so where the hell are you coming up with $8 billion of DNA? And I think it's because he's assuming you know, another million to a million and a half of capex, maintenance capex, which by the way is ridiculous. And so let's just level set on that point. Yeah, I build the shell and I build the power and water to the rack. I don't do anything else. That's it. So, to your question, what happens if technology changes or the rack changes? That is my tenant. Um that's them. Like whether it's a fluid stack, Amazon, Anthropic, like whoever it is, it's their technology, it's their silicon, it's their rack that they've put together with their integrator, whoever the integrator is. It might be Dell or Microsoft or Supermicro, I don't know, or it might be Google. But all I do is make sure they have power and water and a temperature and humidity controlled building, and that they have access to that power 99.999% of the time. That's it. So people ask me, well, what about maintenance capex? I'm like, look, if they want to change, if the technology changes in three to five years, that's on the tenant. Like, I'm still providing electricity and water. So you're telling me on a $10 million per megawatt data center build, I am gonna spend a million to a million and a half dollars on maintenance cap. I mean, it's like, and by the way, you know, Jim, I hope he listens to this. Pick up the phone and call me, bro. Like my phone number is out there online, or you can, you know, email me. But like stuff like this, honestly, and I'll give you another example, and you guys will love this one. But like back in the Bitcoin mining days, I had a New York Times reporter reach out to me and they wanted to do a story on Bitcoin mining up at Lake Mariner. And I said, Yeah, I'm happy to get on the phone. And I got on the phone, and what I found was an ignorant person that did not want to understand that they were wrong and how they were reporting. So I ended up ending the interview because they literally didn't care that what they were writing was wrong. They just wanted to sell eyeballs, which basically is what Jim Chanos just did. And I lost all, like again, I this was a guy who I held in high regard when I was coming up in the business. Like, but but like it's it literally you're wasting my time and you're just trying to sell eyeballs, and and that kind of stuff really bothers me. So sorry to get triggered on it, but it's like, you know what? If you want to learn something, pick up the phone and call me. Because that's what I do. Like, if I don't understand something, I pick up the phone and call somebody and say, hey, I don't get it. What am I doing wrong?

SPEAKER_00 38:05

Yeah, yeah. That question is me as an accountant, not not me listening to Jim Chanos and what he's saying there. So there's no there's no connection there

Labor Constraints And Grid Versus Behind The Meter

SPEAKER_00 38:13

in that question. So anyway, I it it was No, the the the second part is really about um, you know, you've highlighted the sort of the capex cost per megawatt from project to project, and you've seen that sort of grow there. What are you seeing in terms of the fact that we are seeing an industry that's growing now, that's going to require, I'm gonna say probably a lot of electricians out there um to deliver these uh you know these big sites. Are you seeing now the increased costs because of the um the supply and demand of specific types of of uh of the of building of the building trade uh to develop these massive sites for you? Um and and that's something that you you know have to have a as a cautious eye because you know you're not talking about 26, 27, 28, you're already mentioning sites now that 28, 29, and 30 coming, you know, coming online as well. So how do you sort of like how do you sort of like keep that in in sort of like um in your sort of cost cost assumptions moving forward?

SPEAKER_02 39:15

Yeah, 100% impacting, you know, folks ask this all the time about supply chain impact stuff. Like I would say it's less like equipment, less supply chain. I can see how that would be an issue if you're building like a two gigawatt site, yeah. I mean, because you need literally like hundreds of coolers from the same comp. Like, so yes, I can see how if you have a massive site that you're trying to literally deliver all at once, might you have supply chain issues? Yes. But to answer your question, principally, like where we are seeing it definitely is in electricians and labor and that specific market. To just anecdotally, it it is more acute in Texas because it's not just data centers, right? You have LNG facilities, like uh, there's other big projects going on there. Less so for us in Lake Mariner, because again, there's just not as many projects. So we had, you know, Buffalo Bill Stadium was the other big project up there. So we pulled um as that kind of ramped down, we pulled a lot of electricians and labor from there. We're pulling from as far as Detroit on um electricians uh in upstate New York. But again, um regionally, like it matters, right? Because there we can pull, and it's not as acute, you know, because you have folks that are willing to travel, but they don't want to travel to Texas from Detroit. So I think regionally it's very dependent. Um, in Kentucky, this is one of the reasons why we pulled in Floor as our EPC. You know, they were actually winding down a couple big projects in that region and were able to kind of bring folks from those jobs like in. So that stuff does really matter. Um, and the ability of like the EPC to execute there matters. I think the other thing that you know we've talked about is all of our sites that we have added in recently all have the ability to expand with generation. And it's not behind the meter generation. Like the approach that we take is we want to work with the local utility to add power gen that sells onto the grid that allows our hyperscaler to connect to the grid. Because again, if you're connected to the grid, there's a thousand points of failure. If you are connected to behind the meter, there's a few, and and stuff goes wrong behind the meter. And then in particular, right again, is one of the reasons why we're not we haven't done one of these big behind the meter deals in West Texas. It really depends on what you are running. Like, meaning, are you running combined cycle facilities? Like those are called intermediate and baseload facilities and power parlance for a reason, because they run you know 80 to 90 percent of the time. Reciprocating generators, right, in power parlance are called peakers because they're supposed to run during times of peak stress, like five to 10% of the hours in the year, not 99.999% of the time. And so if if you look at you know why GE, Siemens, like all these companies are so profitable, you know, selling the equipment is a piece of it. The big piece of it is the long-term service agreements where they come in and maintain your equipment. So if you think about like that's easier to do on like a combined cycle because you have probably two gas turbines and a steam boiler. And a site where you have 80 reciprocating generators behind the meter, and by the way, you maybe have two or three different OEMs, meaning you've got cat, Wartzilla, GDE, that is a dog's breakfast and a recipe for disaster because the maintenance that is going to be required for that, again, running peakers that like are now trying to run 99.99% of the time, like, good luck. Uh, that's not something you know that we have any interest in getting involved in. Again, because we are power people, like that is very hard. Now, I'm not saying it's impossible, but it's very hard to do. And then when you add in the SLAs and the penalties you you have from your lease, you know, counterparty if you're not meeting their 99.99% of the time, like stuff can go wrong really quickly. And so um, I think our focus again at having Floor come in. Floor historically was a big EBC builder of L and G facilities, power plants, refineries. They have that capability in-house, and that's one of the reasons why we're partnered with them.

SPEAKER_00 43:47

I I've worked with uh Floor before when I was in uh Kazakhstan, so I know their operations in the oil and gas uh industry. Um, just one final follow up before I pass it back to Bryce. Um, in terms of the location. and and and and that you're you're going to be developing now um and it's linked to the lay the labour source for those areas um if you're having to sort of like um attract people from further distance away is it is it is a is sort of like temporary accommodation or something like the company's thinking about in terms of you know removing out having to travel hundreds of miles backwards and forwards on a daily basis we we did it we did a Texas tour last year and um you know it's some of the sites are quite remote and you know to get people there we know that some of um some of the other um companies in the in the business space or moving into this business space now are building sort of temporary accommodation which is then more attractive for for workers to come and stay during the week maybe go home at the weekends and back again is that something that you need to do or do you feel that you've got sufficient quantity of that right skill set in a in a location closer to your facilities I think it's more the latter you know um Anthony you've been watching land man watching the the the series landman with the man episode yeah every episode yeah yeah it's a good series but uh no we're we we haven't we haven't done that yet uh Anthony and um haven't seen the need to yet um certainly a possibility in the future but again I think what we're doing is unique up at Lake Mariner.

SPEAKER_02 45:28

I don't think we'll ever do that again um where we are like the EPC general contractor all of it that was a really unique circumstance for us um for a couple reasons. One we had to move extremely fast right to meet our customers' objectives so no EPC was really willing to come in and do that. And two, we had all the contractor relationships because we've been building up there for five years right so that I I think was incredibly unique to that region. I I don't think we'll do that which is why we're bringing floor in in in Kentucky gotcha.

What To Watch Next And Closing

SPEAKER_02 46:00

Hey I just wanted to give you a quick opportunity to close out I know we've covered a lot of ground here Patrick you've alluded to a few additional sites what should we be watching for next and what are your final words to investors yeah sure look I think we've been pretty vocal with you guys that there were lots of different sites that we've looked at and that we were going to add them in uh to the company which we did with you know the two in February both Hawtsville and Morgantown we are waiting uh for FERC approval to close on Morgantown you know we are expecting that by early August so stay tuned on that one that one in and of itself is a company maker I mean we're talking about you know over a thousand megawatts of power and data center just at that site alone. Eastern Kentucky was another one we added in in the interim right that site is up to a thousand megawatts as well with a defined power schedule with our power partner AEP so those two sites alone literally double and triple the size of the company so if you think about I just want to frame for folks like on cocktail napkin math how I think about our company if you take what we've leased up at Lake Mariner and what we've now leased at Hawesville to me and and and you apply a 15 to 20 times NOI multiple on the on those businesses and again use roughly 600 uh million fully diluted shares outstanding in my opinion that's 20 to 30 dollars a share okay so literally just what we have signed and leased today you know the stock I think today was trading below that for a moment so 20 to 30 bucks a share and then I have ready right to go eastern Kentucky which literally is a thousand megawatts and doubles the size of the company in one fell swoop that doesn't include Morgantown which is coming too and then as we've mentioned to you all there are other sites that you know we are we have diligence that we are very deep in that are of the size and the scale again of eastern Kentucky and Morgantown. And I would say um it's you know we have been primarily North America focused we now have some really exciting opportunities um in Europe that are coming into the fray. So I think you know as you guys know customer diversification geographic diversification these have been themes we talk about a lot and that is because we the board management team insiders own about 25 to 30% of the equity of this company. Okay so we constantly talk about terminal value terminal value of the site terminal value of the customer we're signing like we are shoulder to shoulder with every single shareholder in this company and you know folks ask me all the time again like hey why are you diluting why aren't you doing 95 100% equity believe me we don't want to dilute just as much as all the shareholders in the company don't want to dilute but I am going to have a comfortable equity layer in this company I'm not gonna lever it to the gills. It was only as you guys know three years ago where I had to issue $32 million at 68 cents. That is not lost on me and I had to dig we collectively me, Paul, Nazareth Stephanie we had to dig this company out of being over levered. I won't let that happen again. So um super excited again I think we now have like when I look at it really concrete $2030 you know per share in the stock and I think we can double and triple the size of this company again in the very short term. So that's what I'm excited about. I think you know I'm just trying to take a very measured approach because as you guys know you know our guidance of 250 to 500 megawatts a year that's two and a half to five billion dollars of capital. And so you know I want to make sure we bite it off in bite-sized chunks so that we can execute on it and we don't choke on it. Right. And so that's why doing things like that preemptive equity offering in May sets us up now so we can just cruise. And because I have seen a lot of folks have reached out to me and they're like hey are you going to dilute us based on this lease no I already did it. So and again I don't want to dilute if we don't have to the only time I'm gonna go to the equity markets or the convert markets is when we have accretive growth to execute upon that's it. So that's the message I'm really excited about it.

SPEAKER_00 50:52

Sorry I kind of beat up uh Jim Chanos he used to be a uh a hero of mine but not anymore you'll be pleased to know Cancer give you a $37 um uh price target yesterday after your announcement so uh brent brett at uh at cancer there was obviously going through those numbers and using his formulas to come up with a with a target there they do the good amount of countries they do break it down but obviously they don't have all the numbers they're just making some assumptions in terms of on terms of cost and you've already mentioned about the weighted average cost capital and I'm sure if they had all the information they could do their own net present values and and and come up with a with with figures that would support close to you but your you're being a little bit more sort of like you know $20 to $30 they come out with $37. We even did our own model yesterday which was a bit closer to to Cantor's model actually but at the end of the day it's it's it's just a it's just a model it's it's no it's it's I haven't got all the facts we only have the information that was presented so yeah look like you know as you guys know I tend to I I tend to be a bit more conservative right because I come from like credit distress land.

SPEAKER_02 51:55

So you know I'm sure some of my shareholders are going to be upset that I wasn't telling you oh it's 50 to 70 I mean are we going to get there 100% I mean look at Steven Bird. He's got a price to upside price target I think of like 70 bucks. So yeah am I going to get there yes but I got to tell you right now I'm focused on the next 24 to 48 hours and putting one foot after another to kind of get there than I am you know I'm trying to you know as they say enjoy the journey not the destination because that is what makes it fun. But we got an awesome team our company's growing by the way you guys we we I think exited you know last year with around 100 employees. I think we're you know approaching 200 at this point and we're probably going to exit this year somewhere around like 400.

SPEAKER_01 52:39

But yeah so couldn't be prouder of the team the execution our partners I mean it's a privilege to be doing what we're doing with the partners we are um so yeah super excited so thank you guys guys have been huge supporters of us for a long time so I really appreciate it hey we appreciate uh watching you on the journey here we'll let you back to it Patrick I know you're busy you got a million things going on we appreciate the update here you guys if you have any additional questions leave them in the comment section below thanks so much for watching we'll see you back here tomorrow