Riot Blockchain, Inc. (NASDAQ:RIOT) Q4 2024 Earnings Call Transcript February 24, 2025
Riot Blockchain, Inc. beats earnings expectations. Reported EPS is $1.95, expectations were $-0.27.
Operator: Good day, and thank you for standing by. Welcome to Riot Platforms Fourth Quarter 2024 and Full Year 2024 Earnings Conference Call. Please note that all participants have been placed in listen-only mode until the question-and-answer session begins following the company’s presentation of its prepared remarks. Please also be advised that today’s call is being recorded. I would now like to hand the conference over to Phil McPherson, Vice President of Capital Markets and Investor Relations at Riot Platforms. Please go ahead.
Phil McPherson: Thank you, Michelle. Good afternoon, and welcome to Riot Platforms fourth quarter and full year 2024 earnings conference call. My name is Phil McPherson, Vice President of Capital Markets and Investor Relations. And joining me on today’s call from Riot are Jason Les, CEO; Colin Yee, CFO; and Jason Chung, Executive Vice President, Head of Corporate Development and Strategy. On the right, Investor Relations website, you can find our fourth quarter and full year 2024 earnings press release and accompanying earnings presentation, which are intended to supplement today’s prepared remarks and which includes a discussion of certain non-GAAP items. Non-GAAP financial measures provided should not be considered as a substitute for or superior to the measures of financial performance prepared in accordance with GAAP and are included as additional clarifying items to aid investors in further understanding the company’s fourth quarter and full-year 2024 performance.
During today’s call, we will be making forward-looking statements regarding potential future events. These statements are based on management’s current expectations and assumptions and are subject to risks and uncertainties. Actual results could materially differ due to factors discussed in today’s earnings press release, in comments and responses made during today’s call and in the Risk Factors section of our Form 10-K and Form 10-Q, including for the year ended December 31st, 2024, which will be filed later today as well as other filings with the Securities and Exchange Commission. With that, I will turn the call over to Jason Les, CEO of Riot Platforms.
Jason Les: Thank you, Phil, and good afternoon, everyone. Before we dive into Riot’s 2024 results, I would like to reflect on the industry as a whole. When Riot began mining Bitcoin in 2017, the entire network hash rate was just 3 exahash. By 2024, Riot alone grew our self-mining hash rate by 19.1 exahash to a total of 31.5 exahash, while the global network reached an all-time high of over 750 exahash. In January of 2024, the approval of the Bitcoin ETF brought further institutional adoption, resulting in the most successful launch of ETFs ever as measured by AUM. By July of 2024, Bitcoin had become a key topic during the presidential election. And post-election, Bitcoin hit new all-time highs of 75,000 and then 100,000. Riot has been mining Bitcoin since 2017 with conviction in its long-term value proposition.
In January 2024, Riot stopped selling Bitcoin produced through our mining activities in order to retain more Bitcoin. Additionally, we purchased Bitcoin for the first time in December 2024, following our inaugural successful convertible debt offering, ultimately adding over 10,000 Bitcoin to our balance sheet over the course of the year, ending 2024 with 17,722 Bitcoin worth $1.7 billion based on the Bitcoin price of $93,354 as of December 31st, 2024. Our efforts to maximize the value of our unique power assets does not mean we have changed our long-term view on Bitcoin. Riot currently maintains approximately 4% of the global network and we will continue to pursue a compounding increase in the value of Bitcoin held per share as a key operating metric while seeking to maximize the value of all of our assets for our shareholders.
In 2024, AI/HPC demand for access to power resulted in Bitcoin mining power assets being converted into AI/HPC use. This has continued into 2025, increasing the value of large-scale sites with secured access to power near major metropolitan areas, sites such as our Corsicana and Rockdale facilities. There is a unique timely opportunity for Riot shareholders to benefit from the value of our assets and to maximize shareholder value. And I am excited to share further updates on our aggressive pursuit of this opportunity. I would like to start by reviewing Riot’s key accomplishments during 2024. We successfully energized the Corsicana Facility. Following approximately 24 months of development, Riot successfully energized the 400-megawatt substation, developed four buildings, and deployed immersion systems in miners, ending the year with approximately 14 exahash of self-mining hash rate at the Corsicana Facility alone.
Our Bitcoin yield strategy. In January of 2024, with ETF approvals imminent and the 2024 having approaching, Riot’s decision to retain all mine Bitcoin and focus on increasing Bitcoin per share, i.e., Bitcoin yield has resulted in a nearly 40% Bitcoin yield in 2024 and more than $100 million in incremental asset value appreciation in 2024 alone. Our convertible senior notes offering. In line with Riot’s Bitcoin yield strategy and target to continue to increase Bitcoin held per share, Riot was able to take advantage of favorable market conditions and issue our inaugural convertible senior notes offering. This issuance lowered our cost of capital, expanded our institutional investor base and proceeds were utilized to acquire Bitcoin and further enhance Bitcoin yield.
The Block Mining acquisition. This was Riot’s first acquisition in more than two years and our first business expansion outside of Texas into an attractive new jurisdiction, offering 60 megawatts of current mining operations and expansion opportunities to 305 megawatts and beyond. Finally, the E4A acquisition, our most recent acquisition, which supplements our engineering business, while also enhancing our core operating capabilities. For 2025, we expect to grow our Bitcoin mining hash rate by approximately 22%, while aggressively pursuing AI/HPC opportunities to maximize the value of our unique assets. We believe that this approach will bring the most value to our shareholders, while allowing us to remain flexible as market opportunities continue to present themselves.
With a strong balance sheet, a solid operating team and a management team focused on value-creation, I am truly excited at the path we have ahead of ourselves. With that, I would like to now turn the call over to Colin Yee, CFO of Riot Platforms.
Colin Yee: Thank you, Jason. I’m excited to present Riot’s financial results for the fourth quarter and fiscal year ending 2024. For ease of reference, we have highlighted key metrics on Slide 6, which presents a snapshot of key financial and operating metrics for the full year 2024. However, let’s just jump into the details on the following slides. During 2024, Riot increased its self-mining hash rate from 12.4 exahash to 31.5 exahash, representing a 154% increase over the course of the year and outpacing the increase in the global hash rate, which rose by 67% in 2024. Furthermore, in 2024, Riot produced 4,828 Bitcoin, 27% lower than the Bitcoin produced in 2023, primarily as a result of the 2024 having event in April. However, in spite of this April’s halving, Riot ended the year producing approximately 16.6 Bitcoin per day as compared to 16.8 Bitcoin per day at the start of the year.
In January of this past year, Riot stopped selling Bitcoin produced in order to increase our Bitcoin holdings ahead of the halving. In conjunction with December’s convertible notes offering, proceeds from which were used to acquire Bitcoin in the spot market, we have increased Bitcoin holdings per million fully-diluted shares from 31.8 to 44.3, representing an increasing Bitcoin yield of 39% for the year. Going forward, we will continue to focus on increasing Bitcoin yield in order to ensure that our shareholders are able to participate in the long-term value-creation opportunity that we believe Bitcoin represents. Riot ended 2024 holding 17,722 Bitcoin, an increase of 141% relative to the 7,362 Bitcoin that we held at the end of 2023. And Riot continues to retain 100% of all self-mined Bitcoin.
For the full year 2024, Riot reported total revenue of $376.7 million as compared to $280.7 million for 2023, a 34% increase year-over-year. This increase was primarily driven by higher Bitcoin prices. Gross profit for the full year 2024 was $147.6 million as compared to gross profit of $97.6 million for the full year 2023. Non-GAAP adjusted EBITDA for the full year 2024 was $463.2 million as compared to non-GAAP adjusted EBITDA of $214 million for the full year 2023. Riot adopted FASB’s final standard on crypto assets issued in December 2023 under which Riot now recognizes its Bitcoin held at fair value and with it, changes in fair value are now recognized in income. As a reference, the Bitcoin price at the end of 2023 was $42,265, and the price at the end of 2024 was $93,354.
This resulted in a mark-to-market upward adjustment of $458.7 million for [Technical Difficulty]. Net income for the full year 2024 was $109.4 million or $0.40 per share compared to a net loss of $49.5 million or $0.28 per share for the full year 2023. Full-year 2024 net income includes non-GAAP stock-based compensation expense of $125.2 million. Unrealized loss on marketable equity securities of $69.5 million and depreciation and amortization of $212 million. As a reminder, beginning in the first quarter of 2024, we adjusted our depreciation schedule for mining hardware from a two-year to a three-year schedule based on our evaluation of our own operational history. For the full year 2024, Bitcoin mining revenue totaled $321 million, a 70% increase relative to the full year 2023 Bitcoin mining revenue of $189 million.
Bitcoin mining cost of revenue primarily consists of direct production costs of Bitcoin mining operations, including electricity, labor, insurance and other expenses, but excluding depreciation and amortization. Bitcoin mining gross profit for the quarter was $165.5 million, representing a margin of 52% as compared to $163.6 million or a margin of 87% for the full year 2023. Cost to mine, excluding depreciation in the fourth quarter totaled $42,011 per Bitcoin, of which power costs amounted to $33,281 per Bitcoin or 79% of total cost per Bitcoin. Quarter-over-quarter, power costs increased from $0.031 a kilowatt hour to $0.038 a kilowatt-hour as a greater proportion of our power was procured at the Corsicana Facility at spot market rates, which at the times purchased were on average, elevated relative to the fixed rate under PPAs at the Rockdale Facility with proportionally fewer offsetting power credits than in the previous quarter.
For the full year 2024, we achieved an all-in cost of power of $0.034 a kilowatt-hour, which remains one of the lowest in the industry. Direct non-power costs, which include direct labor, minor insurance, minor and minor related equipment repairs, land lease and related property taxes, network costs, and other utility expenses totaled $8,730 or 21% per Bitcoin, down from the third quarter of 2024 when direct non-power costs accounted for 25% of total costs. This drop is a result of Riot continuing to achieve economies of scale at our operating facilities. As Riot continues to grow hash rate, we expect to see non-power direct costs as a percentage of total cost per Bitcoin to trend lower. I’d now like to turn the call over to Jason Chung, Riot’s EVP and Head of Oper — and Head of Corporate Development and Strategy to discuss our most recent acquisition.
Jason Chung Thanks, Colin. I’m excited to share information on Riot’s latest acquisition. This past December, we closed the acquisition of E4A Solutions for $52 million in cash consideration at closing. E4A solutions based in Houston, Texas is a leading provider of consulting, commissioning, and procurement services designed to support electrical project needs for substations, from low voltage to high voltage. E4A employs approximately 50 employees and services power plants, transformers, substation equipment and switchgears for a blue-chip roster of clients, which include FirstEnergy, Hitachi, Hyundai and Toshiba among many others. In 2024, E4A generated $28.2 million in revenue, a 31% increase over 2023 revenue of $21.5 million and $4.6 million in EBITDA, representing margins north of 16%.
The acquisition of E4A Solutions offers a number of strategic benefits for Riot and significantly bolsters our engineering business. More specifically, our engineering business, ESS Metron will now have the ability to add higher-margin recurring services revenues to customers. E4A services expertise in maintenance, servicing, training, and repair can now be offered on equipment manufactured by ESS Metron, while the countercyclical profile of the services business relative to ESS Metron’s business is also expected to reduce overall revenue volatility in our engineering business going forward. With this acquisition, Riot has now brought in-house expertise in developing, maintaining and servicing medium- and high-voltage substations, which will further derisk infrastructure development and save costs, while also improving our ability to perform on-site maintenance on switchgears and, most importantly, leading to further improvements in operating uptime for our Bitcoin mining business.
In 2024, Riot’s engineering business generated $38.5 million in revenue, down from $64.3 million in 2023. This reduction in top-line revenue was largely driven by one large manufacturing contract for a governmental entity, which took longer than anticipated to complete due to supply chain constraints during the year, which resulted in decreased receipts of materials and delayed recognition of revenue. The custom electrical products manufactured Riot’s engineering business, such as switchgear and power distribution centers are particularly important components in data center development and in power generation and distribution facilities. And there has been a significant increase in demand for these products due to the continued increase in AI/HPC data center development and the continuing — continually increasing demand for power.
With our now expanded engineering business products and services capabilities, Riot is well-positioned to capitalize on this growing demand and alongside the recent acquisition of E4A Solutions has a clear path towards reaching $100 million in revenue in 2025, while also increasing [Technical Difficulty] With that, I’d now like to turn the call back over to Jason Les.
Jason Les: Thank you, Jason. Now, I’d like to turn the presentation to Riot’s treasury strategy. Riot continues to hold all of its self-mined Bitcoin. In 2024, holding all of our Bitcoin versus selling at the time it was mined resulted in an increase of $120 million in value. Our bullish belief in Bitcoin has not wavered and we believe that holding Bitcoin, increasing Bitcoin yield will result in greater shareholder value over time. This slide demonstrates the delta we achieve on holding Bitcoin production over the year versus selling it at various end-of-year Bitcoin prices. One of the tenets of Riot strategy has been to maintain a strong balance sheet, underpinned by our growing Bitcoin balance since 2018. This has allowed Riot to act opportunistically and grow our portfolio of assets, which in turn has created the new opportunity set that we are executing on today.
While Riot has always been conservative with regards to long-term debt, the increased value of our Bitcoin and the attractive terms in the convertible debt market led to our inaugural debt issuance in December 2024. We view this as a method to utilize our strong balance sheet to acquire more Bitcoin at an attractive cost of capital. Our internal policy is to not take on additional debt such that doing so would exceed a 40% debt-to-Bitcoin value ratio. We feel this limit on leverage is prudent and allows us to create additional value for our shareholders. Further, as we continue to increase our Bitcoin balance every month by retaining production from our mining operations, we are organically deleveraging over time. For 2025, Riot anticipates capital expenditures totaling $198 million.
While we have halted the Bitcoin mining expansion at Corsicana, existing expenses primarily consisting of the 600-megawatt substation development. Additionally, we continue developing the various facilities that comprise our Kentucky operations including $23 million on infrastructure expense and $110 million on miners. However, it’s important to note that $35 million of that is for deposits for hash rate coming online in the first half of 2026. As a result of these investments and our efforts on the ground, we expect to end 2025 with 38.4 exahash in deployed hash rate. A 22% increase over the 31.5 exahash we ended the year 2024 with. We expect this to approximately match the pace of global network hash rate growth this year, maintaining our approximate 4% share of the overall network.
In 2024, the rapid growth of AI applications and the resulting increase in demand for large-scale access to power needed for large language models and inference compute became a dominant market theme. We have long viewed our Bitcoin mining business as being at the forefront of the convergence of energy and sound money. With the emergence of AI/HPC, we now find ourselves at the center of the convergence of energy and high-performance compute. As such, Riot’s uniquely positioned portfolio of high-quality energy assets can now be viewed through a wider lens. The core of Riot’s strategy has been to take a vertically integrated approach to securing, developing and operating energy infrastructure for Bitcoin mining. This focus has led to Riot owning and operating two premier power assets, which now form the basis for the expanded opportunity set uniquely available today to explore all opportunities available to us to maximize the value of energy assets across our portfolio.
The market has already taken notice of the value of AI/HPC contracts and has rewarded those companies, which have made this pivot with elevated valuation multiples. The chart on page 21 demonstrates both the valuation multiple expansion some of our public peers have experienced as they have pivoted parts of the power assets to also include AI/HPC as well as the lagging valuation multiple seen by the remaining Bitcoin mining pure-play players. While revenues from Bitcoin mining can exhibit volatility in the near term, AI/HPC contracts offer long-term predictable cash flows with credible counterparties. Which can reduce the overall volatility in our financial performance and to which the market assigns higher valuation multiples. One of the most active themes we see in the market today relates to the significant need for power faced by the major hyperscalers in their arms race to deploy more high-performance compute capacity and the associated capital expenditure plans they committed to in order to achieve their goals.
Microsoft, Meta, and Alphabet alone have collectively committed to more than $200 billion in spending plans this year with the rest of the industry closely following suit. In order to meet their aggressive growth targets, hyperscalers face immediate need for large-scale access to power with industry forecasts calling for nearly 30 gigawatts of additional power demand over the next five years. As I will get into, Riot has a clear and unique opportunity to capitalize on this tremendous demand imbalance. As hyperscalers continue to search for large-scale access to power, they particularly prioritize three key attributes. One, proximity to major markets; two, speed-to-market to deploy compute quickly; and three, large-scale access to power in the hundreds of megawatts.
Riot’s primary assets in Rockdale and Corsicana are particularly well-suited to satisfy each of these key requirements, which I will get into shortly. We believe the Corsicana Facility offers the most compelling opportunity available today for hyperscaler to get immediate access to large-scale power, all in close proximity to Dallas, one of the most sought-after data center markets in the country. This gives hyperscalers access to existing connectivity hubs and lower latency. Most importantly, Corsicana has 400 megawatts of power that can be utilized today and we’re actively developing an additional 600 megawatt substation to be ready in early 2026. Combined, this is one gigawatt of power that can be available for a hyperscaler to deploy almost immediately, which when combined with its location, we believe makes Corsicana an extremely rare offering.
In addition, we own the land at Corsicana. We have access to water with that access growing further and multiple redundant sources of fiber available today. These are the key criteria hyperscalers are first looking at when evaluating sites and we think Corsicana is extremely well-suited. For many of the same reasons Corsicana is uniquely well-positioned to attract hyperscalers, we believe that Rockdale represents an exciting opportunity as well. At Rockdale, we have a 700-megawatt substation already developed and capacity that can be made immediately available. The current Rockdale site spans more than 200 acres, most of which is cleared open land, which a hyperscaler could begin immediate development on in order to take advantage of this large-scale access to power.
Similar to Corsicana, Rockdale is located just 45 miles from Austin, which also represents a major data center market, which hyperscalers are seeking proximity to. We also believe the relatively short distance between the Rockdale and Corsicana facilities at about 100 miles creates another compelling opportunity for one hyperscaler to deploy up to 1.7 gigawatts of power across the two sites that are not only in close proximity to each other, but the two major data center markets. Late last year, Riot began to receive inbound inquiries from hyperscalers interested in exploring opportunities related to our power assets. And as these discussions progressed, it became clear to us that there was a significant opportunity at hand. As a result and since then, we have taken a number of steps to capitalize on this tremendous opportunity presented to us.
First, this past January, we engaged Altman Solon, a leading telecommunication strategy consultancy to help us conduct a study to evaluate the feasibility of utilizing our Corsicana site for AI/HPC uses. This study has now been underway for a number of weeks and is expected to conclude next month. Second, earlier this month, we made meaningful changes to our Board of Directors with the addition of three new independent directors and the expansion of our Board to include six directors in total. The new independent directors who have joined our Board come with a unique set of skills, including significant data center development expertise, deep real asset development, capital allocation experience and broad corporate governance experience. Our newly constituted Board will continue to oversee and advise the management team on this and other key initiatives at Riot.
Third, we have engaged Evercore and Northland as our financial advisors to advise us on engaging with potential AI/HPC partners in a value-maximizing approach. Both firms have significant experience and strong relationships with hyperscalers, data center developers and financing partners and in advising companies like Riot. Fourth, although we previously announced the halt to the Phase 2 expansion of Bitcoin mining at our Corsicana Facility, it is important to highlight that we continue to develop the additional 600-megawatt substation on-site to ensure that we enhance the attractiveness of the site to hyperscalers and data center operators. We are taking concrete steps to ensure that we are maximizing the value of the opportunity available to us today and we’ll provide further updates on our progress as they occur.
As a result of the foundation we have built over the previous years, we believe that 2025 will be an exciting and rewarding year for Riot shareholders. We are dedicating a tremendous amount of resources towards executing on AI/HPC opportunities. We also continue to remain focused on increasing Bitcoin yield, driven by our Bitcoin mining operations and prudent capital strategy. Underpinning our mining operations is our industry-leading low cost of power, which Riot continues to demonstrate exceptional results over our now multiyear track record. For the past few quarters, we have talked about our focus on operational efficiency and the results of these efforts have taken shape over the past few months as demonstrated in our monthly production results.
Operational excellence continues to be one of our priorities and our teams are hard at work to drive top-tier results. In closing, myself and the rest of the Riot team could not be more excited about the direction the company is heading. We have the assets, the capital and the right team to execute on significant growth of shareholder value. Thank you all for listening to our presentation. We would now like to open the call to questions. Operator?
Q&A Session
Follow Riot Platforms Inc. (NASDAQ:RIOT)
Follow Riot Platforms Inc. (NASDAQ:RIOT)
Operator: Thank you. [Operator Instructions] Our first question comes from Greg Lewis with BTIG. Your line is open.
Greg Lewis: Yes, thank you and good afternoon and thanks for taking my questions. Jason, thank you for the presentation. Definitely a little more insightful than in previous quarters. I guess my first question is around the HPC opportunity. And I know we’ve been talking about this on and off for a while now. But as you — as you look at the landscape, I think from going back to last year, there was always this expectation of a requirement of prompt power where there was a window. As you’ve — as you’ve engaged and had increasing conversations with more hyperscalers, are you seeing that window potentially widening, i.e., you know, it’s great to have front power in 2025, but having power in further out years is almost just as valuable?
Jason Les: Yes, Greg. I think the power is extremely valuable in 2025, but any power available, including in 2026 and in 2027 is extremely valuable as well. As we highlighted in the presentation, you’re looking at about 30 gigawatts of AI/HPC demand — that’s just AI/HPC’s demand alone over the next five years. Transmission, new interconnects take four, five years to get approved. Long-lead substation items, those are taking four to five years to get shipped now and transmission cannot keep up with the growth. So capacity in 2025 commands a real premium, which is why we think our assets are so valuable, both with hundreds of megawatts available today and continuing [Technical Difficulty] looking for a solution to and Riot is very well-positioned to be that solution for hyperscalers.
Greg Lewis: Okay, great. Thank you for that. And then the other question is, obviously, we’re engaging with hyperscalers in signing-up — potentially signing-up HPC transactions. What — is there kind of a dual process, i.e., is Riots kind of talking with EPCs and some of the company — and some of the companies or suppliers that would help build these facilities? Or is it kind of let’s get the hyperscaler in first and then figure out kind of let them lead the build-out of the conversion or the build-out of the greenfield site?
Jason Les: Greg, I’d say we have multiple tracks going on at once. There’s a number of ways that these opportunities can materialize and we’re keeping our options wide open in order to make sure we get the value-maximizing approach. We received interest from hyperscalers and large AI companies. These discussions have been ongoing with our new financial partners. We’re now reaching out to additional infrastructure partners and financing partners as well. So a lot of going on in the mix and what we’re going to end up to is the path that maximizes value after surveying kind of the whole market here. But as I highlighted, in parallel, we’re continuing to build out the capacity at the site. We’re pursuing on what would be the first stage of the development anyways. So I would say multiple tracks being run down at once.
Greg Lewis: Okay. Super helpful. Thank you, guys. Have a great day.
Jason Les: Thank you, Greg.
Operator: Thank you. Our next question comes from Nick Giles with B. Riley Securities. Your line is open.
Nick Giles: Thanks, operator. Good afternoon, everyone. My first question is, you know, as hyperscalers race for readily available capacity at scale, there’s a clear desire to build large-scale campuses and this does require significant amounts of land. So I was curious to what extent Corsicana could accommodate additional capacity beyond the targeted 1 gigawatt or based on your earlier comments, could Rockdale be better positioned from this perspective? Thanks very much.
Jason Les: So we think the Corsicana is an amazing site because it has the 1 gigawatt interconnect approved. It has the power capacity available today, it’s expanding. We have dual sources of fiber, water close to major data center market. We’re really excited about what Corsicana can mean in these discussions. So when you think about land for AI/HPC developments, there’s kind of a spectrum of which these things can take and that impacts design decisions. The more land you have generally, the cheaper it is to build, the easier it is from a layout and cooling perspective. The less land you have, the more next-generation kind of rack and cooling technologies that you end up using that can kind of be more expensive. But nonetheless, we have — you noted we have a lot of land at Rockdale.
We have a good amount of land at Corsicana and we are expanding that land portfolio at Corsicana right now as well. And the final point I’ll make is oftentimes when hyperscalers are talking about campuses, they don’t always generally mean they are directly adjacent to each other. Sometimes they are many — sometimes up to 100 miles apart. So we are increasing our land portfolio in order to increase the optionality of what we can do here in this endeavor.
Nick Giles: Jason, that’s super helpful. I appreciate those comments. My next question was, we’ve seen economics vary to some degree in the deals that have been announced. So what’s your main priority as you think about terms of any potential HPC deal, is it margins, capital contribution, duration, size or kind of all of the above?
Jason Les: Well, Nick, I’d say a key part to getting a project of this scale done in the hundreds of megawatts — gigawatts that we’re talking about, financing is a key thing. So to be successful at financing, you really need a blue-chip counterparty, someone who’s very well — very well-capitalized that we would be able to securitize the cash flows on in the best possible terms. And further, we’re talking about hundreds of megawatts that’s really only attracting the largest player in this space. So in addition to driving the financing terms, the type of counterparties that we would enter into is really going to have a big impact in driving valuations. So as you can see, one of the key points I’m hammering on that we think about is the type of counterparty is necessary in order to maximize shareholder value here.
As far as some of the other terms, there’s a whole mix of how all of these things can be played together. I think it’s kind of too early to say which one may be more or less important to another. At the end of the day, what’s important to note is Corsicana and Rockdale, but specifically Corsicana is ideally well-positioned. We think this is a premier asset that commands a premium for its quality and its timing to getting power online and that’s something obviously that hyperscalers are prioritizing. And that — these attributes will ultimately lead to better deal terms that reflect the value of those assets.
Nick Giles: Jason, I really appreciate all the color. Continue. Best of luck.
Jason Les: Thank you, Nick.
Operator: Thank you. Our next question comes from Darren Aftahi with Roth. Your line is open.
Darren Aftahi: Yes. Hey guys, thanks for taking my questions. First if I may, Jason, on your comments about Corsicana, you talked about procuring additional land, any kind of indulgence on how large of an additional plot that would be? And is that something that the counterparties you’re potentially talking to have kind of asked for?
Jason Chung: We have spoken to counterparties, to some it matters, to some it doesn’t. No, like I touched on earlier, it ultimately depends on the type of build-up that you’re looking to do. The technology with data center development is becoming very efficient with density, are on to 100 to 200-kilowatt racks now, direct chip liquid cooling, all of these components — all of these technologies are improving density. So it kind of depends — different types of customers have different needs for this land. Nonetheless, our focus is on increasing, maximizing the optionality that we have. So I don’t have a specific number for you, but I can tell you we’re procuring in the hundreds of acres in order to — just like I said, make sure we can accommodate as wide of a market as possible.
Darren Aftahi: Great. And then one more if I may. Just in the deck, you kind of advertising Corsicana, advertising Rockdale and then you potentially talking about maybe doing a fiber loop and advertising whole campus. So I guess as it stands today, Rockdale has always been a Bitcoin mining facility. Is there a scenario analysis economically where you would flip the switch and completely make that an HPC campus? Thanks.
Jason Chung: Yes. If the — if the opportunity was there to — if there was interest in that site and the economics made sense to take over the capacity that we had there, we would certainly be interested in that. I think when you think about a 30 gigawatts of demand, 1 gigawatt at Corsicana and then 700 megawatts of Rockdale represents a really meaningful portion of something that could fulfill that demand. 1.7 gigawatts would be about 5%. If someone had both of those sites, 5% of hyperscaler AI demand over the next five years, I think that would give those players a real strategic advantage. So if there is that interest and the economics are there to convert that site, we will certainly be looking and evaluating that closely.
Darren Aftahi: Great. Thanks, Jason.
Jason Les: Thanks, Darren.
Operator: Thank you. Our next question comes from Brett Knoblauch with Cantor Fitzgerald. Your line is open.
Brett Knoblauch: Hi, guys. Maybe my first question similar to the last, is all of Corsicana on the table if a hyperscaler wanted the full gigawatt there?
Jason Les: Well, so we have the 600 megawatts that’s being unutilized being developed right now. That would be the least friction way to accomplish a deal, Brett. But if there was someone that was interested in the full 1 gigawatt taking over the entire site as it is and had the economics to make that accretive to right shareholders, yes, we would definitely look at that. And that it’s always possible that presents itself because that’s a lot of power coming online, 400 megawatts now and then the remaining 600 megawatts 12 months from now or approximately 12 months from now. So we’re assessing demand and we are going to maximize the value of these assets.
Brett Knoblauch: Understood. And then maybe just on the power agreements place at these sites, do they need to be authored to support AI/HPC and maybe in particular, with Rockdale, does like your participation, a lot of demand response and counter programs, does that need to switch from maybe a flex load to firm load?
Jason Les: No, they do not need to be altered. Our power supply agreements do not require any participation in demand response. Everything we do there is voluntary. And that’s a part of our power strategy that works with Bitcoin mining because it’s interruptible load. Obviously, that doesn’t work with AI/HPC. So we do that with Bitcoin mining voluntarily to drive the best economics, but not a requirement as a part of our power supply agreements.
Brett Knoblauch: Perfect. Thanks, guys. Really appreciate it.
Jason Les: Thanks, Brett.
Operator: Thank you. Our next question comes from Reggie Smith with JPM. Your line is open.
Reggie Smith: Hey, guys. Thanks for taking the question. Congrats on the quarter and the year. I had a question. I was looking at your OpEx trends and I don’t want to rain or any parade or anything like that, but noticed, we looked at SG&A minus stock comp and we’ve seen that number grow sequentially each quarter. It sounds like you’ve done some acquisitions. There may be some deal costs in there, but like what’s driving that? How are you benchmarking OpEx growth and things of that nature? And I was curious if you had an employee count because I was trying to calculate stock comp per employee. And I got a follow-up. Thank you.
Jason Les: Okay. So working backwards, Reggie, I don’t have exact employee count for you. It will be in the 10-K, which should be filed very shortly here. As far as G&A goes, so our G&A has been elevated both in 2023 and 2024 — I’m sorry, Q3 and Q4 2024 due to one-time M&A and legal litigation expenses. They are not recurring expenses for running our business, they’re not a part of running a Bitcoin mining operation or AI/HPC operation. So that was pretty meaningful. And especially Q4, we had approximately $22 million in one-off costs. Like I said, mainly litigation expense. We have some stock-based advisory services that was closing out a three-year engagement. We had some M&A expenses as well. So obviously, we would like our litigation tied up in a timely manner. So we’re not up for incurring these expenses anymore and we can get cash SG&A down. Okay. And I have an employee count for you. Thanks to help in the room here, that’s 783 full-time employees.
Reggie Smith: Got it. Okay. That makes sense. All right. And so it sounds like there is some one-timers. Is it possible to handicap like where you would like that to be kind of on a — on a run-rate or go-forward basis like stripping out all the noise? Like what do you think you can run your Corsicana facility and your Rockdale and everything else you get going on a quarterly basis.
Jason Les: So first to note how the expenses are allocated, there is a split between our direct operating costs. Those are in cost of goods in addition to power, although power is the big expense. And then we have some other kind of more corporate operating expenses, facility-wide expenses that fall under SG&A. But to directly get to your question, we think our 2024 cash run-rate quarter is going to be in line to modestly above what our previous Q4 2024 run-rate guidance was, which is approximately $30 million to $33 million. It’s very difficult to anticipate what litigation or one-time maybe M&A items might be or advisory fees that may pop up. But on a run-rate basis, that’s what we are looking at per quarter, $30 million to $33 million cash. And we’re focused on optimizing that, especially relative to our overall hash rate portfolio.
Reggie Smith: No, that makes sense. Okay. Great. And then if I can get one more in. I was curious, the consultants that you’ve spoken with, have they given you any guidance on like how long it may take to finalize a deal? Just kind of speaking through like how quickly something could get done or how long might it drag on. And it sounds like and maybe I’m reading too much into this that if something isn’t done by late 2025, then you would then consider going back to kind of Bitcoin mining — expanding your Bitcoin mining operation. It sounds like you’ve given — you feel comfortable with where your hash rate will be relative to the network hash rate. Am I thinking about that right?
Jason Les: Well, I don’t think we’re thinking of — we need to get — if there wasn’t the capacity fully spoken for by the end of this year, we would resume Bitcoin mining. I think what — based on the demand and the quality of assets that we have, we’re pretty optimistic about getting something done. How long that takes is probably unpredictable from a soup to nuts to completion, but there are going to be steps along this way. Depending on the ultimate deal structure that comes through, multiple steps could include signing a lease; having a development partner, for example; securing financing, for example, and those could occur in a number of orders. So we’re really aggressively pursuing this. We want to communicate updates as they occur.
But we aren’t thinking about this as we need to get this done by ex-date or to go back to Bitcoin mining. We think these assets are really exciting for AI/HPC data center development and we want to get something meaningful done there.
Reggie Smith: Got it. Great. And I guess, again, I just want to congratulate you guys. I know we had talked previous quarters about improving uptime and operations. You guys have done a good job of that. So I wanted to acknowledge that. So a great job on that. And I look forward to — for more updates on the — on the HPC side. Thank you.
Jason Les: Thank you, Reggie. Appreciate that. All the credit goes to the guys in the field, guys and gals in the field working hard to improve operations every day.
Operator: Thank you. Our next question comes from Mike Grondahl with Northland. Your line is open.
Mike Grondahl: Hey guys, thank you. Two questions. The first one, can you provide a little bit more color? I guess my specific question is on the 400 megawatts at Corsicana and the 700 at Rockdale, have you gotten specific interest for that power yet? Or would you say it all kind of so far relates to the 600 megawatts at Corsicana? And then secondly, any operational goals for 2025 on the mining side?
Jason Les: So starting backwards with your question there, Mike, operational goals in the Bitcoin mining side, we’re looking to increase our hash rate by approximately 22% this year. That’s really driven by expansion at our Kentucky facilities. We’re focused on operational excellence. We want to continue executing on our power strategy and improving uptime as months goes on. So we have a good amount of work ahead of us to improve that business line as well. And I think we’re making all the right moves to show better results there and hit those growth targets in Bitcoin mining ending the year with 38.4 exahash. As far as the interest in the power, I would say that hyperscalers, large companies have seen our access to power and they’re interested in getting some piece of it.
It hasn’t — the inquiry has not necessarily been specific as we want your next 600 or we want the 400 here or anything like that. It’s more of, hey, we need power fast, we want to have a conversation with you of how might that work with your assets.
Mike Grondahl: Got it. Okay. Hey, thanks a lot.
Operator: Thank you. Our next question comes from John Todaro with Needham. Your line is open.
John Todaro: Hey, guys. Thanks for taking my question. I have two. So I know it’s early days in the convos, and I hear you on all the demand out there and what everyone’s expecting. But I think the market is starting to question some of that demand, especially on days like today. So first question is just as those conversations with hyperscalers have gone, have you noticed any change recently versus maybe a few weeks ago or a month or so ago? Just kind of any key leads there? That’s my first question.
Jason Les: We have not seen any change yet. We’re explaining the scope of the counterparties we are talking to now in conjunction with our financial advisors, Evercore and Northland in order to make sure we get the fullest view of the market here. But, I think there’s been lots of new stories on AI demand and how that’s fluctuated. The store — the news about Microsoft today, it seems very quickly that a lot of that was disputed and was a bit misleading or misunderstood information to start. But I’d also point out to developments like Claude 3 is now the number-one graded AI. And how did they achieve the results in that AI is they did a significant amount of pre-training because xAI got a 100,000 going on 200,000 video GPUs up fast.
So even though there’s efficiency gains being had, or being studied, it still remains that large-scale access to power and then by extension compute gives hyperscalers, gives AI developers an edge in these products. So as long as there’s demand for AI, which I only believe is going to continue, I think we’re going to continue to see increased demand for power to build data centers to get a competitive edge in that sector.
John Todaro: That’s great. And then my next question, so just try and frame it again for us because I know you guys have made the argument before how the Bitcoin business complements HPC because I see you’re basically doubling down in Bitcoin, especially on Bitcoin on the balance sheet, but didn’t want to aggressively go at HPC. And it seems like at least from the investors we talked to, like the guys who like HPC don’t really like Bitcoin. So do you spin off one of them or how do they complement each other, I guess?
Jason Les: Well, I’d say today, Riot’s basically positioned itself at the center of two rapidly growing industries. We have Bitcoin and we have AI/HPC data center demand, which is a very common theme. Bitcoin’s rapid rise to institutional adoption and the price appreciation is what has positioned Riot as uniquely one of the handful of companies that has direct exposure to Bitcoin. This — our Bitcoin balance in addition to our Bitcoin mining business has provided Riot with the framework to raise capital at attractive levels and invest in projects like Corsicana, like Rockdale and that has resulted in the opportunity set today. We think we can create a lot of value from that. So with the AI/HPC opportunity, this gives us another way, seemingly better way to use energy assets and generate substantial economic returns.
And I think being flexible and opportunistic is what is going to give our shareholders a diversified path to greater returns. So, John, I would say that Bitcoin is the reactor that has grown our business and grown our assets and got us to this point where we could do this and we believe it’s going to continue generating opportunities for us. And having the ability to develop AI HPC data centers or do deals in that sector is just another super exciting way in which we can increase value.
John Todaro: Great. Thanks for that and congrats on the quarter, guys.
Jason Les: Thank you.
Operator: Thank you. Our next question comes from Martin Toner with ATB Capital Markets. Your line is open.
Martin Toner: Thank you so much for taking the question. Congrats on the deal and the improved efficiency this quarter.
Jason Les: Thank you, Martin.
Martin Toner: A quick question on the deal and the $100 million run-rate and the margins. Just wondering at what — when might you expect this business to produce like industry — margins more in line with the industry and more in line with the acquired company?
Jason Les: Thanks, Martin. I’m going to turn that question over to Jason Chung, our EVP and Head of Corporate Development.
Jason Chung: Sure. Hi, Martin. Interesting question. We definitely see — if we think about what the engineering business is today, following the acquisition of E4A Solutions, the services business by nature tends to have higher margins. And so we definitely see the higher-margin profile demonstrated by their financial performance. They’re at the center of the — this incredible growth we see in demand for power generation and data center development. So we’ve seen — historically, their top line continues to grow while they’ve been able to maintain healthy margins and we definitely expect to see that continue going forward. And I think that really forms the basis for the sort of high-level view we have around 2025 performance on the engineering side.
And the other part of that — the other component to that is the Metron business where now that we’ve gotten this one particular contract behind us, we expect that business to recover meaningfully. And so if you look at where they were, call it, in 2023, top-line there, similar margins, but now we’re combining the E4A business on the — with the top-line that’s expected to continue to grow strongly this year and maintain healthy margins. Then on a sort of combined pro forma basis, we think that the engineering business will show significant improvement in margins. And then as you — then if you kind of look out further, we continue to see some really interesting synergy opportunities in combining these two businesses. Both on the cost-savings side and cross-selling and even just given the E4A solutions based in Houston where we think some of the capacity issues that have currently constrained the ESS Metron business can potentially be alleviated.
So I think there’s some really interesting near-to-mid-term opportunities to increase that margin profile even further now that we’ve got the two businesses together.
Martin Toner: That’s great. Thanks so much. That’s it from me.
Jason Les: Thank you.
Operator: Thank you. Our next question comes from Patrick Moley with Piper Sandler. Your line is open.
Patrick Moley: Yes, good evening. Thanks for taking the question. So I have one on the Rockdale site. If I’m not mistaken, I think that you currently lease that site, you don’t own it outright like you do with Corsicana. So I’m just — can you maybe just talk about the specifics of that lease and whether the fact that you lease that land makes it any less attractive to hyperscalers when they’re thinking about choosing a site for the data center? Thanks.
Jason Les: Yes, Patrick. So you are correct, whereas at Corsicana, we own the land; at Rockdale, we are on a land lease. That is on a pretty long-term though. The initial term is 10 years and then there are multiple 10-year renewals at our option. I think a total of 40 or 50 years in total on that. So it’s a pretty long-term ground lease. I think what would impact deal there is the type of deal that’s getting done. I think it depends on what the level of involvement from the customer would be. I think ultimately drives whether or not Rockdale and the fact that the Rockdale land is leased impacts if it’s viable for them. But at the end of the day, the major theme here is constraint on power. And we believe that as that constraint becomes more and more real and these hyperscalers and AI companies are competing with each other for capacity, factors like that may start to mean less, if it means getting capacity on sooner.
Patrick Moley: Okay, got it. And then just a follow-up. It’s obviously going to vary by the needs of the tenant, but some of the similar deals that we’ve seen get done in this space have been north of 10-year lease terms — sorry, 10-year terms on the AI/HPC deals. So based on your conversations, can you talk about how long you think a contract if you were to sign one could potentially be? Has that come in at all as some of these questions have arisen about longer-term demand? Anything on kind of like the length of some — of some of these deal terms would be great. Thanks.
Jason Les: Yes. I think long-term demand is — I’m sorry, the long term of an agreement is a key point and that’s both for customers and for us. We’re going to be building infrastructure, we’re going to be spending capital on this. And there are good terms available today. We want a long-term agreement to lock that in and give ourselves the visibility of the financial return for the assets over a long term. So I think both sides are similarly interested in having long-term agreements or 10 years and beyond.
Patrick Moley: All right, great. That’s it from me. Thanks, guys.
Jason Les: Thank you.
Operator: Thank you. Our next question comes from Bill Papanastasiou with KBW. Your line is open.
Bill Papanastasiou: Good evening, gentlemen. Congrats on the deal and thanks for taking my questions. My first question is with respect to driving higher Bitcoin yields. Curious to hear how the company is thinking about dilution going forward to drive a higher Bitcoin per share. And could we see Riot issuing other forms of structured products down the road?
Jason Les: So minimizing dilution is a key part of being successful at Bitcoin yield. The core part that drives it for us as a Bitcoin miner is our mining operations, we’re achieving a very low direct cost of production. But that is what is, yes, the basis that forms how we organically generate Bitcoin yield. And of course, we’re always going to be looking to raise capital in the least dilutive, lowest cost of capital manner in order to enhance Bitcoin yield. And it’s a reason why this metric is powerful, we think, it helps whole teams such as ourselves accountable to making dilution as minimal as possible. So that leads to how we think about going into the market with equity. You’ll notice over the few months or so, we’ve been very light with our ATM program.
It also is what drives us to find additional forms of financing. We think with our assets and the things that we’ve — what we built up here at Riot, we are well-positioned to find alternative forms of financing. All of this contributes to driving a Bitcoin yield. As far as other kind of structured products, I assume you’re alluding to the things that like what a strategy has done. That’s not something that we’ve looked too closely at yet. Our main focus has been the AI/HPC opportunity right now, but I think we’ll closely monitor how these products come about and see what investor or market demand looks like for them.
Bill Papanastasiou: Appreciate that color. And then secondly, and apologies if I missed this earlier in the call, but can you share some more color on how the design engineering aspects of the AI/HPC data center could look like? Will you need to hire or partner with more external parties to get it up and running? Curious to hear what the plan is there. Thanks.
Jason Les: So I think it will ultimately depend Bill on the type of deal that comes together. Like we talked about earlier, we want to pursue what maximizes the most value on a risk-adjusted basis. And there’s multiple different avenues in which we can execute there. In some of these avenues, we already have a team in place to do that. We’re building out our power capacity very well. We’ve been exceptionally good at that and the acquisition of E4A makes us even better at building out what you would say is the power shell of an infrastructure — of a data center development. Further, Riot and Bitcoin mining, we’ve developed our infrastructure there with state-of-the-art cooling techniques and that experience translates into supporting data center development, especially with the type of state-of-the-art cooling techniques we’re seeing there.
So I think we have a good amount of talent. Nonetheless, we are focused on bringing on additional help to support execution there and we’re working on that from both an internal and third-party perspective.
Bill Papanastasiou: Awesome. Thanks for the color. Looking forward to the feasibility study next month.
Jason Les: Thank you, Bill.
Operator: Thank you. Our next question comes from Joe Flynn with Compass Point Research and Trading. Your line is open.
Joe Flynn: Hi, thanks for the question. I was wondering maybe based on early conversations with your consultants, if there’s any operational metrics you can share such as, you know, ultimately what POE you’d be able to deliver at and your build cost per megawatt and how that could be different between the 600 megawatts at Corsicana and the Rockdale site, given there’s already-existing power infrastructure on in the ground? Thanks.
Jason Les: Yes. I think it’s probably too early to comment on what POEs could look like, especially since the technology is changing so rapidly and it’s going to depend a lot on what the type of customer would be. I think both Rockdale and Corsicana are in very similar climates only being 100 miles from each other. So they would be similar on both sites. At both sites though, we have very sufficient access to water, especially at Rockdale. So we have the ability to support a wide variety of cooling technologies there. So, Joe, I couldn’t tell you what it would be like just yet. I think the conversations would need to advance further. We’d have to be kind of in the design phase to know what that particular customer is looking to accomplish.
Joe Flynn: And then just going back to just the maximizing the value of the sites, does that include potential outright sales of capacity and like on that front, have you been approached or even gotten offers for an outright asset sale?
Jason Les: There’s a whole spectrum of potential deal structure here. It could be on one end, it could be lease land, it could be a powered shell, it could be building something to suit of basically a large complete data center. And then the biggest end-of-the-spectrum would be an outright sale of the asset. So we’re open to all of these as we progress in discussions here. And we’re driving — we want to drive a strong competitive process here. It’s why we’ve engaged Evercore and Northland to give us the most broader — most broad penetration to the market we can.
Joe Flynn: Great. Thanks.
Jason Les: Thanks, John.
Operator: Thank you. That’s all the time we have for questions. I’d like to turn the call back over to Phil McPherson for closing remarks.
Phil McPherson: I’d like to thank everybody for tuning in today to our earnings conference call. We look forward to updating you on future events for Riot platforms. As always, you can contact us at ir@riot.ink for any other future questions, and we’ll talk soon. Thank you.
Operator: Thank you for your participation. This does conclude the program and you may now disconnect. Good day.