MARA Holdings, Inc. Q3 FY2022 Earnings Call

Good day, ladies and gentlemen. Welcome to Marathon Digital Holdings Third Quarter 2022 Earnings Webcast and Conference Call. I would now like to turn the call over to your host, Charlie Schumacher, Vice President of Corporate Communications. Please go ahead.

Thank you, Diego. Hello, everyone, and welcome to Marathon Digital Holdings Third Quarter 2022 Earnings Call. Joining me on today's call are our Chairman and CEO, Fred Thiel; and our CFO, Hugh Gallagher. Before we get started, I'd like to remind everyone that our prepared remarks may contain forward-looking statements, which are subject to risks and uncertainties, and that we may make additional forward-looking statements during the question-and-answer session. These forward-looking statements are subject to risks and uncertainties, and actual results may differ materially. When used in this call, the words anticipate, could, enable, estimate, intend, expect, believe, potential, will, should, project and similar expressions as they relate to Marathon Digital Holdings, Inc. are as such a forward-looking statement. Please refer to our earnings release for a full recitation of our forward-looking statements. Investors are cautioned that all forward-looking statements involve risks and uncertainties, which may cause actual results to differ from those anticipated by Marathon at this time. In addition, other risks are more fully described in Marathon's public filings with the U.S. Securities and Exchange Commission, which can be reviewed at www.sec.gov. Finally, please note that on today's call, we will refer to certain non-GAAP financial measures in which Marathon excludes certain expenses from its GAAP financial results. Please refer to our company's periodic reports on Form 10-K and 10-Q for a full reconciliation of its non-GAAP performance measures to the most comparable GAAP financial measures. We'll begin today's call with prepared remarks from Fred and Hugh. After their comments, we will be going through some of the more popular questions from our investors before transferring to a live Q&A with our covering analysts. And with that covered, I'm going to turn it over to Fred to kick things off. Fred?

Thank you, Charlie, and thank you all for joining us today for our Q3 2022 earnings call. The third quarter was a transition and rebuilding period at Marathon. With the facility in Hardin, Montana offline and energization of miners in Texas delayed, we entered the third quarter with only 6,000 miners operational, producing approximately 0.7 exahashes per second. Unsurprisingly, the operational transition that occurred during the third quarter caused our financial results to dip both quarter-over-quarter and year-over-year. However, as our consistently improving Bitcoin production substantiates, our confidence in our ability to rebuild our hash rate while maintaining a healthy balance sheet was well founded. Today, we believe Marathon is a strong foundation on which we can continue to efficiently grow towards our goal of 23 exahashes per second by mid-2023, and expand our position as the leader in securing and supporting the Bitcoin ecosystem. Since the end of the second quarter, we have sequentially improved our Bitcoin production from 72 Bitcoin in July to 184 in August, then 360 in September and then to a record 615 in October. October was the most productive month in our company's history, during which our production was almost equivalent to the entire third quarter. We now hold approximately 11,300 Bitcoins, making Marathon the second largest holder of Bitcoin among publicly traded companies worldwide, according to various third-party sites tracking this data. The consistent improvement in our Bitcoin production is the direct result of increasing our hash rate by bringing more Bitcoin servers online and improving those servers uptime. During the third quarter, we transitioned out of the data center in Montana, where our servers were drawing their electricity from a coal plant that had consistently incurred maintenance issues that negatively impacted our time. This scheduled move was part of our broader strategy to become both more operationally effective and more environmentally sustainable. By transitioning away from the Hardin facility, our renewable energy mix has increased, our uptime has improved, and our fleet has become more efficient on a joule per terahash basis. While we were in the process of exiting Hardin, new facilities, most notably the data center in McCamey, Texas, that is co-located behind the meter of a large wind farm called King Mountain, began coming online. The energization of this new site during the third quarter enabled us to increase our hash rate from approximately 0.7 exahashes per second on July 1 to approximately 3.8 exahashes per second on September 30, as over 30,000 miners were brought online during the quarter. And since the quarter's end, we have increased our hash rate an additional 84% to approximately 7 exahashes, or 10 times greater from where we entered Q3 with approximately 69,000 of our Bitcoin miners operational as of November 1. Not only have we been able to execute in line with our recent expectations, but we've been growing consistently at a time when many in our industry have been struggling. However, before discussing the current state of the industry and why we believe it creates uniquely beneficial opportunities for Marathon, I'm going to turn the call over to Hugh to discuss our financial results for the third quarter. Hugh?

Thanks, Fred. This was a particularly eventful quarter as we energized our miners in King Mountain and completed the exit from the Hardin, Montana site, marking a significant strategic pivot for the company. We also announced additional hosting capacity with Supply Digital, and we secured new term loan and revolving credit facilities. In September, we learned about the ongoing Compute North bankruptcy, which we will discuss further later. Now, turning to the numbers. We recorded a net loss of $75.4 million during the quarter, compared to a net loss of $22.2 million in the same period last year. This reflects a $53.2 million increase in our net loss, and I will break down the components of this increase. As noted in our earnings release, production was very low in July due to our decision to exit Hardin and the initial stages of energizing King Mountain. These factors, along with lower Bitcoin prices, led to a $39 million decrease in revenues compared to the prior year. Decreased production accounted for $26.3 million of this change, while lower Bitcoin prices accounted for $12.7 million. Cost of revenues increased by a total of $29.8 million, primarily due to $20.8 million in accelerated cost recognition linked to the early exit from Hardin. This affected cost of revenues, energy hosting, and other by about $5.7 million and cost of revenues, depreciation, and amortization by around $15.1 million during the quarter. Fortunately, we have moved past recognizing these costs. Regarding the value of our digital currencies, we saw a $5.9 million impairment in the carrying value of digital currencies during the quarter compared to a $6.7 million impairment in the same period last year. However, in last year’s quarter, we also noted an increase in the fair value of digital assets in our investment fund amounting to $42.1 million. As you may remember, we eliminated that investment fund in the second quarter of this year, and we now hold all of our Bitcoin as intangible assets subject to impairment. Therefore, the overall negative impact from last year to this year was $41.3 million. As I mentioned, it was indeed an eventful quarter. Next, I’ll highlight two significant and unusual items recorded during the quarter. In October, we finalized our previously disclosed legal settlement of $25 million, which incurred an expense recorded during this quarter. We also previously noted our exposure to Compute North, which included investments in preferred stock, loans, and operating deposits, totaling approximately $81 million. During the quarter, we assessed our position with Compute North and recorded a $39 million impairment related to the preferred stock, loan, and certain deposits. The bankruptcy process is ongoing, and we cannot comment further except to note these financial impacts. Additionally, we experienced a $3.8 million increase in interest expense, largely due to the convertible notes issued late in 2021. Partially offsetting these unfavorable variances were several positive items. We recorded gains on asset sales amounting to $31.9 million during the quarter. This included the previously disclosed sale of equipment linked to King Mountain development and other asset sales connected to our exit from Hardin. We sold about 22,000 miners from Hardin for proceeds of around $46.5 million and recorded a gain of roughly $4 million from this transaction. Operating expenses decreased significantly, mainly due to a reduction in non-cash compensation expenses, although this was partially offset by increased costs associated with our business growth. We also recorded an income tax benefit of $5.8 million this year compared to a benefit of $0.1 million last year. Adjusted EBITDA was a loss of $8.7 million, in contrast to adjusted EBITDA of $78.8 million in the prior year. This decline in adjusted EBITDA was primarily due to lower total margin. To clarify, this is EBITDA, excluding depreciation and amortization of $46.9 million, the $41.3 million that affected the carrying value of digital assets, the legal reserve, and an increase in cash operating expenses of about $6.3 million. These negative variances were partially mitigated by the previously mentioned gain on asset sales. Now, turning briefly to our Bitcoin holdings and liquidity. At the end of September, we had $55.3 million in unrestricted cash, which decreased slightly to $52.1 million by the end of October. Our total Bitcoin holdings were 10,670, valued at $207.3 million as of September 30, and increased to 11,285, worth $231.3 million, by October 31. Unrestricted Bitcoin holdings were 6,842, approximately $133 million on September 30, and decreased to 3,464, about $71 million, by October 31. In October, we borrowed $50 million under our revolving credit line, which is why our unrestricted Bitcoin decreased. We plan to repay this by making the final draw on our term loan facility in November without anticipating significant additional collateral requirements. We expect to increase our Bitcoin holdings over time mainly through our mining activities. As our mining operations grow, we may sell a portion of our produced Bitcoin in future periods for managing monthly operations or for general corporate purposes. We also plan to fund our operations judiciously via our ATM facility. Looking ahead for the rest of the year, we anticipate modest cash needs for investments, including shipping costs for previously ordered miners. With that summary, I will turn it back to Fred for the remainder of the call. Fred?

Thank you. There's no doubt that the macro environment presents challenges for Bitcoin miners. Bitcoin's price has been relatively flat for the past four months. Power prices have increased, and the global hash rate has climbed to over 260 exahashes per second. These forces have depressed margins across the industry. And so far, hosting providers seem to have felt the largest impact. While these variables present challenges for some, they also present opportunities for others, including Marathon. We believe there's no better time to be scaling our Bitcoin production and to be commencing the installations of our previously purchased S19 XPs, which are the most energy-efficient machines available and approximately 30% more energy efficient than the prior generation S19 J PROs, predominantly being installed today by many of our competitors. Today, it is unprofitable to mine with S9s and S17s unless your energy pricing is below $0.03 and $0.065 per kilowatt hour, respectively. Even with S19 J PROs, the breakeven cost to mine is approximately $0.085 per kilowatt hour given the current global hash rate. For these reasons, we believe one of Marathon's unique strategic advantages is that over 60% of our hash rate is expected to be generated by S19 XPs by the time we achieve our primary target of 23 exahashes per second in mid-2023. Given the large mix of XPs in our mining fleet, we believe the efficiency across our fleet will measure approximately 24.2 joules per terahash once fully operational. To put this in context, it's been estimated that global Bitcoin mining fleets are currently operating at an average of 45.9 joules per terahash. Said another way, by mid-next year, we expect Marathon's mining operations will be consuming 47% less energy than the Bitcoin network is today on a per terahash basis. Why does this matter? It means we're positioned to keep the lights on when others cannot. And since Bitcoin mining is a zero-sum game in which the difficulty of mining is dynamic, being able to survive the winter, while others are out in the cold, provides us with excellent downside protection, and it also provides us with more leverage should Bitcoin begin to turn in a positive direction. Installing and energizing our miners and achieving 23 exahashes per second remains our primary goal for the upcoming quarters. Based on conversations we've had with Applied Digital and our other hosting providers, we believe we are still on pace to achieve this target near the middle of 2023 as previously stated. The most recent construction and deployment schedules indicate that miners should start to be energized at Applied Digital facility in Texas during the fourth quarter of this year, while our deployments at their facility in North Dakota should mostly occur during the second quarter of next year. King Mountain is now nearly fully operational. And depending on the outcome of the ongoing Compute North bankruptcy, the second phase energization that will follow may occur in Q4 with final energization potentially in Q1. These timelines are always subject to change, and we continue to provide detailed updates on this progress in press releases, quarterly filings, and elsewhere. With production scaling and our hash rate becoming more consistent, the obvious question to ask is, what's next? In Marathon, we don't speak in detail about our growth plans until we have contracts in place. However, today, I do want to share a part of our philosophy to give you all a sense of how we see the company evolving over the coming quarters. To drive value, we believe it's imperative to become effective and then more efficient over time to maximize optionality and to be proactive rather than reactive. And as part of that strategy, we're constantly evaluating new technologies and searching for new hosting arrangements. This past year, we significantly reduced our reliance on fossil fuels as we broke the mold on deploying behind the meter of renewable energy sites. While the majority of our hash rate will be located in near-sustainable power sites by this time next year, Marathon is always striving to set the pace to make Bitcoin mining more energy-efficient and environmentally sustainable. While we believe there are several innovative opportunities in North America to increase our geographic diversity, decrease our power costs and drive further towards carbon neutrality, we are also investigating international markets which are becoming increasingly more interesting, especially as we look to deploy emerging technology. Bear markets are a great time to build, if you can do so, and we believe Marathon remains well positioned for growth. We look forward to building on our current momentum through the end of the year and beyond to become the largest self-mining Bitcoin miner and to improve our position as a leader in supporting and securing the Bitcoin ecosystem. With that, I'll turn it back to Charlie so we can begin taking questions. Charlie?

Thanks, Fred. At this time, we're going to commence the Q&A section of today's call. We'll start by answering some of the questions submitted by investors on our Q&A platform. The first question comes from Eugene who asks, continued sideways between $18,000 and $21,000 for the next 6 to 12 months, is Marathon currently in a financial position to weather the storm and come out on top if the value of Bitcoin begins to increase in the future? Fred, would you like to take that one?

Sure. As you can see from the Q and our prior production reports, our cost to produce Bitcoin remains very competitive in the marketplace. We think, as we've been saying, we expect Bitcoin to trade in this kind of $18,000 to $22,000 range for some time. We think we're very well positioned to weather that storm and come out the other side very attractively as Bitcoin goes up in price. To date, we haven't yet sold any of our Bitcoin. We will hold that Bitcoin unless we deem it's necessary to cover operating expenses or other expenses. At the same time, as we're going to continue to look at Bitcoin's price behavior. Today was a unique day in the market; the news with the little battle between CC and Sam Bankman-Fried is obviously causing turmoil in the price of Bitcoin, but we think that Bitcoin will likely come back within the range that we've spoken about, and that's a range we feel very comfortable with.

Great. Aisan H. and Mohamed M. were both curious about some of the potential risks to the business. So I guess the questions essentially are, does Marathon currently face any threat of bankruptcy or other major risks associated with it? And what are the company's plans to avoid those risks?

Well, there are lots of risks that we can't control, like the price of Bitcoin, global hash rate, things like that. But we think that barring those things, we're very well positioned to weather through this winter and come out the other side and definitely harvest as the environment improves.

Our next question is about the future of Marathon in 2023. Additionally, will Marathon consider diversifying its business beyond Bitcoin mining in the foreseeable future?

So 2023 is going to be an important year. I mean today, we're operating at a little over 7 exahashes per second. And by mid-2023, we'll be at 23 exahashes. That's pretty substantial growth in a very short period of time. If you look at the balance of the year and if you remember the comments I just recently made earlier on this call, we're beginning to look at international opportunities and other opportunities, which we believe will provide us with very attractive energy pricing, very consistent energy sources, and very well-capitalized hosting partners and energy partners as we continue to evolve kind of our strategy around renewables and driving that renewable mix as high as 100%.

Our next question is from Jacob B., who asks, is Marathon considering an acquisition of another mining operation or data center? So Fred, maybe you can talk about your ideas around potential consolidation that may occur within the mining space.

One of the big challenges, and I've said this at many conferences, is my kind of stock answer to this question. In this industry, when times get tough, the cost to replace assets goes down as well. And so in traditional industries, if you're going to build a factory, whether the industry is doing well or not, that cost differential is marginal. In our industry, when the price of Bitcoin drops, the price of Bitcoin miners drops. And then you have the technology cycles. We've just gone through a technology upgrade cycle. If we were to acquire another miner, they likely will have machines that are S19 J PROs or older, and they may have hosting agreements that don't necessarily fit our mix, which is why we prefer to buy the latest state-of-the-art miners to deploy them so we can have an energy consumption advantage or an efficiency advantage, if you will, in the industry. While we don’t foresee us consolidating the industry necessarily, there may be unique opportunities, and we'll obviously be open to looking at things.

And our next question comes from Brian D. who asks, what is your energy diversification strategy to avoid overexposure to any one provider like Compute North and/or avoid another Hardin incident? Is there any consideration of hosting some of your own miners?

So if by hosting some of our own miners, it means owning hosting infrastructure sites ourselves and contracting for power directly, our model currently is asset-light, where we don't like to invest in the infrastructure. That being said, there are potentially new solutions for energy that could cause Marathon to potentially invest in developing those further, which could drive our energy costs down substantially in the future. I think one of the most important things for miners to have control over is not their infrastructure, but their power price. If you can control your cost of power and drive that down, then you'll have a better competitive position than those miners.

At this point in the interest of time, I think we'll wrap up the section of Q&A. Again, we really appreciate the questions and interest from all the investors who added to the Q&A platform. I'm now going to turn the call back to Diego, our operator, to open the line for questions from our covering analysts. Diego, the mic is yours.

Thank you. We will now open the call to questions from analysts covering Marathon. Our first question comes from Jon Petersen with Jefferies.

Maybe to pick up on a comment, actually, you just made, Fred, on the importance of controlling your power. I wonder if you could give us some more details on the hosting agreements you have with Applied Digital and this next wave with Compute North? How are those power agreements structured? And are they structured in a way that maybe a continued low price of Bitcoin or rising network hash rate coupled with rising energy costs protects you from these hosting providers having future issues such as Compute North is having right now with their bankruptcy?

So the Applied Digital hosting agreements are at a fixed price. Rising costs of energy don't impact our hosting costs in those circumstances. In the case of the King Mountain site, as we've talked about before, we have a PPA on the wind side that is extremely attractive energy price and also gives us the ability to sell that energy to the grid whenever ERCOT needs it, and we get 100% of that upside. For the balance of the energy there, we're buying from the market and following traditional principles of dealing with energy cost risk mitigation. If you look at the future, what's important is being really at the point of power generation rather than being on the grid at all. I don't want to open too much on this because it's a strategy we haven't talked publicly about a lot. But suffice it to say, we believe that Bitcoin miners in the future will be very intertwined with the energy industry and more directly involved in energy generation than before. But this is kind of an extension of the thoughts that we've had regarding partnering directly with the power industry. In those circumstances, hosting is only just an outsourced service you procure from somebody. As long as you control your input costs, the rest is really just a question of deciding what you want to own and what you don't want to own.

Right. Okay. And then maybe a question for Hugh. You used some of your Bitcoin as collateral for your line of credit, and I believe you mentioned taking out a term loan. The bigger question is why, given the current debt costs and uncertain markets, would you take on an additional term loan and more debt? Why not simply pay down Bitcoin and use that to fund your business?

It's just a treasury management decision that we've made keeping our flexibility for the price appreciation of Bitcoin. There's really nothing more to it than that. The term loan we put in place last quarter is a $100 million term loan for two years. We consider that part of our longer-term cap structure evaluation. We try to balance the short-term needs and the ultimate long-term needs of the company when we come up with those decisions, and that's where we ended up.

Our next question comes from Chase White with Compass Research and Trading.

Could you just give us a bit more color on the regulatory issues with ERCOT at the Wolf Hollow side? And any idea when those could be resolved? And what options do you have if there continue to be delays?

Our understanding is that ERCOT is going to give a decision on energization imminently. I don't believe it will be a negative. The number of miners we have installed there in the scope of our overall miner fleet is quite small. In today's marketplace, there are a lot of opportunities to plug holes 5,000 miners here, 10,000 miners there very cost efficiently. There are many people whose miners have been shut off and can't run. They have PPAs to cover. We are seeing ample opportunities to essentially cover that mix if we need to, and we don't believe at all it is going to impact our delivery of our 23 exahash.

Great. That's helpful. And then how much in prepayments do you guys have left for the year? I might have missed that, but for the year-end, just in total, on the main orders that you have for the 23 exahash?

We mentioned that the spending is quite modest. Reviewing our investments so far this year, we have allocated about $200 million in the first quarter, $207 million in the second quarter, and approximately $96 million in the third quarter. The fourth quarter is expected to be relatively light, likely around $20 million to $30 million.

Our next question comes from Stephen Glagola with Cowen.

Just first, Fred, can you clarify for me, did you say that the PPA with Applied is a fixed PPA?

Yes.

Okay. I'm just curious; in light of that comment, we saw from a document in July that had some verbiage around the ability for Applied to pass through some higher power costs up to a certain threshold, and I just was curious if that document should be disregarded or anything you can say around there?

Let me go back and look at the document, the fully executed MSA before I respond to that.

I appreciate that. My other question is about your guidance. You mentioned the 23.3 again, but I understand you don't provide explicit near-term guidance. However, I noticed that you now anticipate reaching 9 exahash by the end of the year, which is a reduction from your October 2019 presentation of 11.5 exahash, and that was also a previous downward revision. Can you explain what is causing the lower near-term guidance on the company’s hash rate?

Sure. It's driven by a couple of things. One is when you contract for the hosting and when they actually have it ready and live to go, there can sometimes be 30- or 60-day lags in that. What you're seeing is us being a little bit more prudent regarding the full energization of the first Applied Digital site in Texas and how that flips over into the new year. Again, we're not changing our end number for midyear at all.

Our next question comes from Lucas Pipes with B. Riley Securities.

My first question is about modeling uptime in Q4 and in 2023. When you reach the 9 exahash threshold and then the 23.3 exahash threshold, what is a good range to estimate for equipment miner uptime? That's my first question.

With Wolf Hollow, we have a very short period of experience so far. We're seeing very good uptime though at that site, actually exceptional uptime. It's a little too early to tell, most probably on that, but we've not had a lot of curtailment there at all in comparison to other sites in Texas that we've been monitoring. I think you're going to see that remaining quite high. Let us get back to you with a better number on that. The other Applied Digital sites, we expect to have very good uptime as well, especially the North Dakota site just because the climatology is much better there, and there's an opportunity for us to overclock miners at that site during the right times of the year. But again, for the Texas sites, you are going to have to look at some seasonality because of temperature. We're still very early on in that. More than happy to work with you on kind of coming up with a model and a way to look at that.

I appreciate the color. And then second question is more strategic. Obviously, there’s a ton of distress in the space miner. Prices are way down on a dollar per terahash basis. Fred, how are you looking at the opportunity set here? Is this the time to raise more capital, be more aggressive? Or would you say, look, you have 23.3 as a target for the middle of next year, let's kind of stick to that. I'm just kind of trying to get a sense for your strategic priorities given the distress in the industry.

Our primary priority remains getting the 23.3 exahashes deployed and operational. Beyond that, as you look at the back half of next year, we're obviously continuing to evaluate opportunities for expansion. Over the next couple of months, you'll see us talk more broadly about that. This is a time period where you can acquire miners at very low cost. There are more and more hosting opportunities becoming available each day. We're also thinking more about the long term. It's not a question of just grow for growth's sake, but rather how do we grow strategically so that we're getting better leverage on energy costs, and we're getting better uptime and availability as we transition to things like immersion and operating in locations where we get very good opportunities for energy arbitrage.

And then I'll squeeze in one last one, if I may. What is the current mix between XPs and J PROs? And if you want to pick the 9 exahash figure for the end of the year, that's fine too. Just trying to get a sense of what the current breakdown is between XPs and J PROs. And then secondly, how are the XPs performing versus the S19 J PROs? Are there any notable performance differences other than, of course, the specified performance metrics, just in terms of uptime and things like that?

Today, I would say that the King Mountain site is entirely made up of S19 J PROs, which constitutes most of our capacity at this time. We also have a few smaller sites operating with S19 J PROs. The first XPs will go into production soon, with Applied Digital being fully equipped with XPs. You can expect some XP units to come online in Q4, but the majority will be up and running in the first half of next year. Once fully deployed, 66% of our hash rate will be from XP, which might actually be a bit lower. In general, about two-thirds of our hash rate is expected to be XP by mid-next year. In terms of performance, the XP machines offer better quality compared to the S19 J PROs, operating within a cooler temperature range. When comparing the input and output air temperatures, we notice significant advantages with the XPs. This capability allows them to operate in warmer climates with less risk of shutdown and suggests good potential for overclocking. We are very satisfied with the performance of the XPs we have been testing in our labs as we start to deploy them.

Our next question comes from Greg Lewis with BTIG.

This is Tyler on for Greg. Most of mine have been answered, but I just wanted to follow up on the merge comments, Fred. Could you provide a little bit more color on how you're viewing immersion today realizing that the XPs are going to be rolled out and then you alluded to the 60% of the fleet, I think, by the middle of next year? Just trying to get a sense as to where the merge is really going to come into play? And as we start to get a real meaningful incremental benefit? Just any other color on that front would be appreciated.

Immersion is particularly beneficial in warm climates like Texas because it creates a more stable environment for miners, reducing the need to shut them down due to temperature concerns. While curtailing for power issues remains necessary, immersion technology allows miners to operate within a narrower temperature range, potentially extending their lifespan and decreasing maintenance frequency. The overall cost of ownership with immersion is expected to be lower, as air-cooled miners require monthly cleaning and upkeep, whereas with liquid immersion, that need is eliminated. Because operating temperatures are less extreme when equipment is immersed in fluid, we anticipate fewer component failures. We believe immersion offers advantages beyond overclocking capabilities and can lead to a longer lifespan and reduced maintenance costs.

Our next question comes from Brian Dobson with Chardan Capital.

You mentioned global geographies becoming more appealing for mining. Could you elaborate on which geographies you're seeing becoming more appealing? And is Africa an area that you're considering as an opportunity?

I'll answer it this way. There are certain geographies with very good energy infrastructure. You have no risk of regime shifts, good legal systems, but you may have energy asymmetry due to seasonality where you need a lot of energy in one season to provide cooling or heating. In the opposite season, you don't need as much energy. These locations are very interesting. There are areas like Latin America, where you have ample hydroelectricity that is not utilized perfectly and makes it available to Bitcoin mining. There are parts of the Middle East with temperature asymmetry in seasonality, and you have some opportunities potentially in Asia. Africa, unfortunately, today, outside of North Africa and the Middle East, faces issues around regime risk. While Kenya, for example, has very ample geothermal opportunities, they have issues with regime risk and developing infrastructure. As we look around the world, predominantly Latin America, the Middle East, and parts of Asia interest us.

That's interesting. Turning to global difficulty, where do you see it now? What are your thoughts on next year given the price of Bitcoin?

The beauty of the Bitcoin network is that it's constantly looking for stasis. When the price of Bitcoin drops, people have to unplug miners because they can't operate profitably. The cost to operate an S19 J PRO is around $0.085, and if your energy costs and hosting costs are above that, you're likely not operating profitably. If you look at the recent price drop of Bitcoin into the $18,000 range, many miners are now marginal operators. This is why we believe it’s important to push hard to use those S19 XPs because of the energy efficiency they offer.

Our next question comes from Kevin Dede with H.C. Wainwright.

Hugh, yes, you mentioned $20 million to $30 million, perhaps, as a last payment for the XP batch?

Yes, that's our forecast for the fourth quarter of this year, and there's not much after that. That's an all-in number, including payments, shipping, and everything.

Can you give us a balance sheet perspective on the two debt levers? Where were they at the end of September? What do you expect them to be at the end of the year?

We have $50 million in the term loan. Right now, we borrowed $50 million on the revolver. We're going to borrow $50 million on the term loan and pay off the revolver. So what we'll have is $100 million of debt outstanding.

All that will happen towards the end of the year?

That will happen in November. And my whole point to that is the collateral is already outstanding for the revolver. So there's no additional collateral needed that revolves into a term loan.

Our next question comes from Gus Gala with Truist Securities.

I would like to know what the main obstacles or potential challenges are regarding the 23 exahashes mid-year 2023 outlook. We are expecting to reach about 14 exahashes in six months. What contributes to that 14 exahashes, and what will be plugged in but not operational at the start of the year? Please provide some detail on the timing related to this.

You've got the Applied Digital site in Texas, which we're deploying now. Part of that may be energized this side of the year, with the balance really in January potentially. Could it energize all this side of the New Year? Possibly, but we prefer not to put that out there. We're just taking a prudent approach on that. North Dakota site construction is moving ahead very nicely; we don't foresee any delays there currently.

Got it. How do you think about running in a scenario where Bitcoin remains sideways, maybe through midway '24, and how do we think about running the business with depressed prices?

As long as we can mine profitably, we'll continue to mine profitably. There aren't many reasons to operate your miners if you're losing money with every Bitcoin you're mining. We have to keep an eye on that. Our margins are still strong. We think our margins would support Bitcoin dropping below this current range and still allow us to operate. The question is really more about how Bitcoin will perform as we get closer to the halving; nobody can project that.

Thank you. At this point, there are no further questions. I'm going to turn the call back to Charlie Schumacher for closing remarks.

Thank you all for your time today. If you have any questions that were not answered during today's call, please feel free to contact our Investor Relations team at ir@mara.com. Thank you, and enjoy the rest of the day.

Thank you. This concludes today's conference. All parties may disconnect. Have a great day.