RIOT Platforms Q3 Earnings | Top Bitcoin Mining Stocks to Watch Now | CEO Jason Les | RIOT

Thank you for having me on, Bryce and Anthony. Always good to catch up with you guys. Fresh off earnings, it certainly is.

Yeah, one of the themes that we've been talking a lot about recently is operational excellence, and I'm not saying we're there right now. You know, I think clearly we're not, but that is what goal everyone around in this company is focused around. So what that means is getting higher and higher amounts of our hash rate that's deployed being utilized on a regular basis in production. And I think we continue to be able to show substantial improvements month over month here with the new deployments at Corsicana and improvements that are being made at Rockdale. That's what's driving that big increase that you talked about, Bryce. And in fact, it would be even better, but I think I touched on the last time we were on the podcast, the first couple days at Corsicana, the substation was down for maintenance, so that entire site was offline for about two days. And if it wasn't for that, that's a substantial amount of our hash right now, that's our biggest site now, that would have made the results even better. So we're really pleased with what we're seeing at Corsicana, the immersion systems, the MicroBT miners, these are all performing very well. We're outside of the forced PPA months right now, there's less curtailment. Rockdale, you saw an improvement in hash rate as well. It's still not where it needs to be, but we have some very big projects underway we touched on at our earnings call last week, and those projects we believe are going to further push that site up to operating, you know, as close as possible to that deployed hash rate level. So with everything that's going on, I think we're in a really good position to continue this positive momentum month over month, and I think we're going to be going into 2025 very strong. It was awesome to cross that 500 Bitcoin kind of psychological barrier. You can imagine we were sweating it very closely here down the end. We had some increases in difficulty along the way, but one of the things that we talked about at the start of this year is as a result of our growth plans underway, we were positioned to be mining more Bitcoin per day post-halving than we were pre-halving. And in October, this past October, we mined more Bitcoin than we did March, the first full month before the halving this year. So I think we're going to continue to see momentum. Unfortunately, that difficulty keeps going up. We're slated for once again another decrease, but we're racing against that, getting more and more hash rate online.

Yeah, so that is primarily being driven by part of the Kentucky expansion being pushed out a little longer than we thought. You know, the great thing about our strategy, something that we've been pursuing for multiple years now, is through this vertical integration strategy, we have pipelines of different opportunities all the time, some organic, some inorganic. We're working on ways to grow our portfolio of power capacity that we can then monetize with Bitcoin mining. So we're always prioritizing and deprioritizing things based on a number of factors. With specifically Kentucky, we had an opportunity to do it as fast as possible, or we had the opportunity to take a little more time and maximize what we could get out of that. I mean, both in terms of the total amount of power and what the cost of that power would be. So with all that other growth that we have going on, we made the decision to think longer term. Let's push this out. Let's get a better deal for ourselves as opposed to rushing, getting it as quick as possible. And you know, that has a contributing factor to some of the reasons that our 2025 target was pushed out as well. So for 2024 though, our previous target was 36 exahash. The revised target is around 35 exahash, so just about one exahash. And we're charging forward through that with the deployments like Corsicana, that's continuing to go very well, and some big projects we have underway at Rockdale that are going to increase that hash rate by year end.

Yeah, first, it's been remarkable to witness the progress underway at Corsicana. You see that picture in our monthly update, that's a picture very recently. That was all dirt at the start of this year. There were no buildings up, the substation was not up. So such an incredible push has been done by our team on the ground there to get that site up. That fourth building, building B2, it's technically combined with B1, but you know, they're delineated B1 and B2. So all four of those buildings are up now and for all intents and purposes complete. In fact, a good portion of building B2, the fourth building, is in operation now. So the remaining steps is just commissioning the final electrical systems. I believe the miners are already in the tanks, or if not, imminently going to be completed, and then getting those miners turned on and getting them set up correctly. So the big components, the substation, the electrical infrastructure, that is all complete now, and it's just the final steps to getting that fourth building, B2, 100% online, but a good portion of it in operation already.

Yeah, so you can see in our earnings call we also announced the initiation of the second phase, 600 megawatts, taking that to one gigawatt of capacity. You can already see on the ground some kind of outlines have been laid out of where those buildings will go. It'll be C1 and C2, then D1 and then D2, as similar structures to what you've seen. And then buildings E and F will be each 100 megawatt buildings that are on the other, I guess, adjacent corner of the lake there, Lake Riot, if you will. So we're moving forward on that. The substation, like all these projects, is the long lead time part of that. I think you'll see buildings coming up in 2025, but way before that substation is ultimately getting there. And thankfully we've already ordered that equipment in the past, and we're looking to get that in and just continuing to maximize the amount of capacity we have there. But very excited to get the rest of the buildings up and complete that site and get that more and more water inside that retention pond.

Yeah, that was nothing there. We created that lake, and that was a lot of work to do that. And that increased part of the cost on phase one. And I know I talked about in the past, hey, phase one was a little bit more expensive because it included these things that service the entire one gigawatt. That's a good example of that. So we've created that lake. It holds, I'm blanking on the number, but you know, tens of millions of gallons of water. And it does two things. One, of course, it collects water that falls from the sky, but we've designed the site where every building, every portion collects storm water runoff and then flows it back into that lake. So we're taking as much water as we can that falls all over the entire Corsicana facility and putting that into that retention pond. And then we use that to help mist over and cool the dry coolers when the temperatures get higher. So not wanting to waste any drop of water, trying to send it all to that retention pond to make sure operations can run smooth when the temperatures increase outside.

So somewhat to be determined. You know, part of it depends on how this immersion technology evolves over time, how much heat future models of miners produce, with the performance of those dry coolers, the future models that we use, how are they working. That will kind of guide how much water use we eventually need because of the advancements in all of those things. We're already estimating to use way less water than we thought when we started this project. The estimates have gone way down. So will that retention be enough to cover it? It's possible, but we do have a few backups, you know, just in case. So it would be great to be 100% sustainable. I think that would be a home run achievement, but we've got some other cool things in the works to make sure we get the water no matter what.

Yeah, sure. So we have actually a slide in our deck on this that kind of goes over our current fleet efficiency and what the forward-looking one would be. So right now, as you noted, 23, and then by the end of the year, with our planned deployments, we would improve that to 21.4 joules per terahash. We're deploying all the latest generation equipment. However, you know, Riot has been mining for a while. We have lots of 30 joules per terahash miners that are still in operation right now, many that have been, you know, fully depreciated, some still being depreciated. I remember we got our first S9s in, I believe, June of 2020. So a long time running with that series. So that is something that will make that average efficiency a little bit higher despite the fact that we have all these new deployments coming in. But as the new deployments become a more significant makeup of the overall fleet, as equipment that's under 20 joules per terahash starts to become a bigger composition of our fleet, we'll get to, expected to get to 20.3 joules per terahash at the end of 2025. And then if you assume we fully execute and deploy our MicroBT purchase order, all the ops in that order, then we would get to 18.5 joules per terahash by the end of 2025. A majority of our fleet, about 65%, would be under 20 joules per terahash. Now remember, the efficiency of the machine, our objective here is to pay as little for power as possible and achieve that unit of hash rate. The other component of there is the cost of power. So Riot has been very good at demonstrating its ability to achieve a very low cost of power. So that makes us a little bit less sensitive to needing to have the absolute latest generation efficient equipment in order to be competitive, in order to be that lower quartile in direct cost to mine Bitcoin. So something that we're still focused on, but the power strategy is something that allows us to, you know, not always need the latest generation stuff in order to be operating at a below break-even price.

We're super pleased with the performance. You know, we didn't go into this decision blind. We tested MicroBT equipment for a period of time before entering into these longer-term purchase orders. But you know, first off, at Corsicana, these machines are purpose-built for immersion, which is different than how we did immersion in the past. We had to get air-cooled miners and we had to convert them to work for immersion, which took time, it took money, and had some complications. It's not a perfect system, how we were doing it. MicroBT making the M56 and now M66 series miners designed specifically for immersion, that has reduced a lot of complexity in deployments, and we're very pleased with the resilience of these machines. That's what's driving a lot of the improved operating performance you're seeing. Of course, Corsicana is running very solid. Our average operating hash rate was 11 exahash over October. Now, however, we're making deployments over the course of October, so you're really seeing a very much more narrow gap between operating hash rate and deployed hash rate at Corsicana than you've seen previously. And it's a result of, I think, the improved miners that were being used there, the improved immersion systems being used there, etc. Over at Rockdale, where we're primarily air-cooled, we're seeing great performance with the MicroBT machines there as well. In fact, we have one building that is all MicroBT M60 series miners, and that building is running at over 95% uptime, even during the hottest periods of the summer, and excluding for the periods of power curtailment, of course, but showing super strong operational resilience in lots of conditions. So that has been great for us to see. We've learned a lot through that process, and that has led us to take those lessons, expand on them, and two more air-cooled buildings at Rockdale, we are putting in all of the MicroBT air-cooled machines, and we think that's going to even further improve the operating performance of the Rockdale facility as a whole and increase the hash rate there and improve the overall efficiency fleet-wide as well.

It's a combination of everything. First, I'll say MicroBT has provided excellent support. There's lots of little small things that, you know, sometimes we need tweaked, and they'll help change around for us. They have technicians, they come out to our facility and offer advice, help with problems if they occur. But you know, with brand new immersion equipment, as you noted, Anthony, there's not too much going on yet. But the broader answer to your question, at both of our sites, we have an on-the-ground miner maintenance team. They're in charge of repairing miners. If we need help beyond that, we use external vendors as well to assist with kind of the overflows and repairs that have occurred. That has been a little bit of a bigger expense for us in the first part of this year, but I think now going to the end of the year, that's going to be a lot less significant, and that's a result of, I think, the better performing machines and more capability we're further in-housing and getting better at.

Yeah, we're seeing less repairs on the immersion fleet. You know, once again, it's still fairly early months here at Corsicana, so we need more data, more time to go on to say for sure. But the stable operating environment, the absence of any type of dust or particulates coming through the air because the machines are submerged, that really helps eliminate a lot of potential problems. So that's one of the main reasons that we're excited about immersion cooling. We think it will prolong the useful operating life of these machines. You know, some people ask, well, do you have data to back it up? Well, I don't think immersion's been around long enough. The technology's changed. There hasn't been enough time for us doing this at scale to be able to back that up over many, many years of data yet. But that's what at least our short-term data is indicating to us is how things will play out in the future. So we are pleased with what we're seeing with the immersion results there. You know, that maybe ties into another thing that contributed to an adjustment of our forward-looking hash rate targets. One of the things you might notice in both our 2025 hash rate target and for the hash rate if Corsicana was fully deployed, is we decreased the amount of hash rate at Corsicana. Now, why is that? Well, it's not because there's less power there, but we are putting less miners into those future buildings to give ourselves the ability, the option to overclock. So the numbers that you see there in that deck reflect just the default settings of miners, and we're deploying less because, like I said, we believe we're getting a lot of improved performance, and we want to leave that additional power capacity open so we have the ability to overclock. If that's successful, that goes a long way in improving capital efficiency. You're buying less miners to get the same hash rate, which is another one of the long-term thesis around immersion and the advantages to it. Now, if that doesn't play out, then we'll just end up buying more miners and we'll get more hash rate anyways. But we're really encouraged by what we're seeing, so we're setting ourselves up to try and be successful there.

Yeah, so a couple big key drivers to that loss, and it's primarily non-cash expenses. As you noted, it's about $60 million in depreciation amortization, about $30 million in stock-based compensation. There was approximately $38 million loss on our position in Bitfarms, and then we had about a $25 million loss on our power derivative. That's our power contract, or it's a combination of all of the power hedges that we have, the PPAs, you know, depending on what term you like to use, and the value of those changes quarter to quarter based on what the power market is looking like. I've seen some people before think it's based on how much time is left on the contract. That's part of it, but it's more so driven by what the market value of those contracts would be based on the forward curve in ERCOT. So all of those non-cash expenses ended up being big components of what that loss was. Then if you go into the SG&A side, we had a few elevated, I would call them one-time expenses, expenses that we would not expect to occur on an ongoing basis running our business, but ones that we had occurred recently. We had acquisition cost related to the closing of the Block Mining acquisition that was around $3 million. For all of the work that went around our initiative with Bitfarms over the summer, we incurred a cost of around $4 million. And then we have some litigation, and so there was some litigation work that increased during that period that was cost for about $5 million. So those are a bunch of things that, you know, we incurred, those impact our financials, but those are not cash costs that are necessary to continue running our business. The costs that occurred now as a result of M&A and litigation, and, you know, hopefully we can get that behind us as soon as possible. The good news is our all-in cost for Q3 2024, it was about $35,370 per Bitcoin, and so that includes both power and non-power related direct costs. So that's the main thing that we are trying to scale up over a wider and wider scale as we get more and more hash rate online. We have significantly more hash rate online now than we did over the average in Q3, so I think that's a real positive catalyst we have for improved results going forward. One thing that's important to note is if you look, and we have a slide for this in our deck, our direct non-power costs quarter over quarter are improving, which means we're achieving the economies of scale at these sites. Power is the one linear cost that scales with more and more hash rate. All the other costs are nonlinear, or I'm sorry, sublinear. So for Q2, our non-power cost per Bitcoin was about $10,000, and then as a result of increased hash rate, that reduced to $8,000 in Q3. So we would like to see that metric continue to improve. And then overall, with more and more hash rate deployed, meaning more and more Bitcoin being mined, as long as difficulty doesn't go too crazy, that will increase the revenue base and make all of these numbers proportionately improve over time. So a lot we're working at. It's a matter of getting more hash rate online, and it's a matter of tying up these other non-recurring items so they don't continue to impact the financials.