Cipher Mining Inc. (CIFRW) on Q2 2024 Results - Earnings Call Transcript
Operator: Good day and welcome to the Cipher Mining second quarter business update conference call. At this time, all participants are in a listen-only mode. After the speakers’ presentation, there will be a question and answer session. Instructions will be given at that time. As a reminder, this call may be recorded. I would now like to turn the call over to Joshua Kane, Head of Investor Relations. Please go ahead.
Joshua Kane: Good morning and thank you for joining us on this conference call to address Cipher Mining’s second quarter 2024 business update. Joining me on the call today are Tyler Page, Chief Executive Officer, and Edward Farrell, Chief Financial Officer. Please note that you may also review our press release and presentation, which can be found on the Investor Relations section of the company’s website. Please note that this call will also be simultaneously webcast on the Investor Relations section of the company’s website, and this conference call is the property of Cipher Mining and any taping or other reproduction is expressly prohibited without prior consent. Before we start, I’d like to remind you that the following discussion, as well as our press release and presentation contain forward-looking statements, including but not limited to Cipher’s financial outlook, business plans and objectives, and other future events and developments including statements about the market potential of our business operations, potential competition and our goals and strategies. Forward-looking statements and risks in this conference call, including responses to your questions, are based on current expectations as of today, and Cipher assumes no obligation to update or revise them whether as a result of new developments or otherwise, except as required by law. Additionally, the following discussion may contain non-GAAP financial measures. We may use non-GAAP measures to describe the way in which we manage and operate our business. We reconcile non-GAAP measures to the mostly directly comparable GAAP measures, and you are encouraged to examine those reconciliations which are filed at the end of our earnings release issued earlier this morning. I will now turn the call over to Tyler Page. Tyler?
Tyler Page: Thanks Josh. Hello, this is Tyler Page, CEO of Cipher Mining. Thank you very much for joining our second quarter 2024 business update call. Let me begin the call with a few key metrics for Cipher as of the end of July 2024. The upcoming growth of our business is the major theme of the call today, and Cipher’s growth and expansion continued throughout the second quarter. We have now grown to 8.7 exahash per second of self mining capacity with a current fleet efficiency of 28.7 joules per terahash. As we look to the rest of 2024, we plan a significant upgrade of mining rigs at our largest data center at Odessa that will bring our self mining capacity to 13.5 exahash per second and improve our fleet-wide efficiency to 18.6 joules per terahash by year end. Those upgraded rigs are scheduled to ship in the third quarter, and we will install them as they arrive because the site is already prepared to take them. Our growth is then expected to accelerate considerably in 2025 with the addition of our new 300 megawatt Black Pearl data center reaching completion. We expect that the data center will expand our self mining capacity to 35 exahash per second and improve our fleet-wide efficiency to 15.3 joules per terahash. Cipher has continued to build its Bitcoin inventory, and as of the end of July held 2,270 Bitcoin. As a reminder for those that are newer to the story, Cipher is probably best known in the industry for its very competitive all-in weighted average power price of 2.7 cents per kilowatt hour. Electricity represents the large majority of our operating costs and our low price is a key driver of our best-in-class unit economics I mentioned that our focus will largely be on Cipher’s upcoming growth, and today I am very excited to provide early details of the extensive work we have been doing to build our pipeline of attractive new data center sites for future development. We have consistently said that owning our own infrastructure is vital to our success. Cipher has always sourced, owned and operated its own data centers. Historically, we have acquired sites that have already received interconnection approvals, but we recently expanded our scope and signed a letter of intent for an option to acquire three new sites in North America that are adjacent to transmission assets and in the final stages of approval for interconnection with a 500 megawatt targeted capacity per site. By getting involved earlier in the development timeline, we can avoid broader bidding competitions and source valuable sites that most of our competitors cannot, while improving the long term visibility for our supply chain and construction functions. Our ultimate purchase price for the sites under our agreement will be very attractive and based on the number of megawatts approved for interconnection, limiting our downside risk. In addition to this purchase option, we are moving forward with a front-of-the-meter site in Texas called Reveille that has already been approved for 70 megawatts, and we believe can be expanded to 200 megawatts by the time it energizes in Q1 2027. For future development after Black Pearl, we now have a potential pipeline of up to 1.7 gigawatts of capacity across four new sites. Notably all of these future data center sites sit at the center of the major trends we see impacting the data center space in the coming years: the continued adoption of the Bitcoin network and related value of Bitcoin mining as a flexible load, as well as the meteoric growth of AI-related HPC data centers. We believe large scale interconnections that can be used in a variety of ways will become more valuable over time. The four sites we intend to develop all have the necessary characteristics for development of HPC data centers, but also sit in locations with demand response programs that would allow us to monetize the flexibility of curtailment used in Bitcoin mining operations. With these sites, we have a lot of optionality, which is exactly where we like to be positioned, in front of trends with the potential for massive growth. In connection with the updates on our pipeline of new sites, I am also pleased to announce the launch of our HPC infrastructure business. Given the requirements for success in the HPC infrastructure business and the relative strength of Cipher Mining, we believe we will be a market leader in this space. In recent months, we have devoted considerable focus to the evolving HPC data center marketplace and have identified three specific advantages we have over competitors. A successful provider of HPC infrastructure needs to have access to the right data center sites, an experienced construction and operations team, and the capital to finance the necessary build-out. Cipher is well positioned in all three areas. Our pipelines sites all have access to adequate land and fiber necessary to service HPC customers. The existing construction and operations team at Cipher has extensive experience building and operating Tier 3 data centers at firms like Google, Vantage and Meta. As an example of the team’s excellence being recognized, the Uptime Institute recently awarded our Odessa data center the management and operations stamp of approval. The Uptime Institute sets industry standards in the data center industry and has historically focused exclusively on traditional data centers. This is their first instance of branching out to the Bitcoin mining industry. Our team literally sets the standard for operational excellence in our industry. When it comes to financing expansion, our management team has deep Wall Street experience with a proven track record of raising capital. Over the last few months, we have been inundated with requests from HPC lenders and investors managing billions of dollars dedicated to the space, asking for information about our data center sites, our development pipeline, and our level of interest in being a provider of HPC infrastructure. We have also had deep technical discussions with potential tenants and we are confident in our ability to build powered shelves that will be very attractive to hyperscalers and other large tenants. Given our site portfolio, our unique strengths in construction and operations, and the level of investment capital available, we are excited to embark on a new major line of business for the company. We are also still very excited about Bitcoin mining and the potential for managing the associated curtailment. While it is too early to predict the exact mix of our business lines over time, Bitcoin mining and HPC infrastructure are complementary lines of business with different risk and payoff profiles, and even have the potential to converge. We think Cipher is uniquely positioned to be best in class in both verticals and our strategy will be guided by our intent to maximize shareholder value over time as we develop our future data centers. Now let’s turn to the next data centers we are building. Slide 6 and 7 show a rendering of the completed data center at Black Pearl and photos from the current site work. We are scheduled to energize the site in the second quarter of 2025. Our O&M building is taking shape at the site and steel erection, concrete foundations and underground electrical work is progressing on schedule. Our design envisions 250 megawatts of air-cooled and 50 megawatts of liquid cooled Bitcoin mining. We have had hyperscalers inquire about our willingness to repurpose a portion of the data center for HPC infrastructure, and while we haven’t completely ruled it out, our current intent is to dedicate the full 300 megawatts to Bitcoin mining. At full capacity, the site is anticipated to produce roughly 21.5 exahash per second of hash rate. Slide 8 shows an overview of the Reveille data center site. The site is located in Cotulla, Texas in Load Zone South, which is a different area of ERCOT from Odessa and Black Pearl, which are located in Load Zone West. It has been approved for 70 megawatts, but based on early discussions with the transmission and distribution service provider, we believe we can expand the site to 200 megawatts in 2027. Given that the timeline to energize this site aligns with the necessary timeline to manage the supply chain and build a Tier 3 data center, we have focused our initial planning and discussions for Reveille on HPC infrastructure. It is still too early to determine the exact plan for the site, but we have had a high level of interest from capital providers and potential tenants, and our baseline plan now is to proceed with building a powered shell data center for HPC and securing a long term lease from a high quality tenant. Now let’s move to a review of our current operations. On Slide 10, we give a portfolio overview of our existing data centers and a near term timeline for expected scaling of our data centers and expansion in our self mining hash rate. Year to date, we’ve paid an average all-in electricity cost of $15,004 per Bitcoin produced at our data centers. We are very proud of this number, and it drives our best-in-class unit economics. Please note that when some of our competitors talk about these costs, they only include electricity and not transmission and other charges. In contrast, when we talk about all-in electricity costs, we mean the total cost to deliver electricity to our mining rigs, so our numbers include all transmission and other charges, and our low numbers dramatically demonstrate our competitive advantage. On the left side of this slide, you have an overview of our four current data centers along with our all-in electricity cost per Bitcoin at the sites year to date. The charts on the right side of the slide give you a graphic illustration of the number of megawatts we manage related to our self mining operations and the hash rate produced by those operations, as well as the additional growth opportunities in the coming year and a half. As you can see, we expect to manage 566 megawatts of self mining across our five data centers in 2025, and we expect those data centers to produce 35 exahash per second of hash rate At this point, we will turn to production by site. On Slide 11, you can see a picture of our Odessa facility. Odessa is the most significant part of our portfolio as it represents approximately 86% of our Bitcoin production in July. Recently Odessa became the first Bitcoin mining data center to be awarded the Uptime Institute’s stamp of approval for management and operations. Odessa is a wholly owned facility with a five-year fixed price power purchase agreement and some of the lowest cost power in the industry. We currently generate approximately 6.9 exahash per second at the site utilizing approximately 207 megawatts. Those same 207 megawatts will generate roughly 11.3 exahash per second with the pending rig upgrade expected in the coming months. We have mined roughly 1,622 Bitcoin at the site year-to-date through the end of July. On this page, we also provide the observed all-in electricity cost per Bitcoin at the site post halving, which was $23,563. Even after the recent halving reduced the number of new Bitcoin paid to miners, you can see how valuable it is for Cipher to have a cheap fixed price of power available on such a large portion of our portfolio. On Slide 12, we highlight our joint venture data centers of Alborz, Bear and Chief. With the recent expansions at each of Bear and Chief, the sites now have a total power capacity of 120 megawatts and currently generate approximately 3.7 exahash per second. We own 49% of the JV sites, and they now generate roughly 14% of our overall Bitcoin production. On this page, we also provide the observed all-in electricity cost per Bitcoin at the sites post halving, which was $28,784. As a reminder, both Bear and Chief operate as front of the meter sites, so there will be some expected seasonal fluctuations with their electricity costs and summer months tend to be higher. As we turn towards the rest of 2024, we look forward to our continued growth in both our Bitcoin mining business and our new HPC infrastructure vertical. At this point, I’ll turn it over to our Chief Financial Officer, Ed Farrell.
Ed Farrell: Thank you Tyler, and hello to everyone on the call. As I’ve done on previous calls, I’d like to start by providing some high level observations on the quarter, and then we’ll go through the key line items in detail. For anyone tracking Bitcoin and the miners, it should come as no surprise that revenues were down in the first full quarter after halving. From the outset, Tyler and I have emphasized designing a business that we drive through the entire cycle. Despite the anticipated drop in revenues, we are very encouraged by the business performance and the company’s growth profile as we move past the recent halving. Low cost fixed price power and a strong balance sheet remain key strengths of our financial position. Slides 14 and 15 give a snapshot which we provide every quarter on some of our financial metrics on both a sequential and year-over-year basis. Let’s move onto Slide 16 and delve into the numbers in more detail. In the second quarter, we faced significant industry-wide headwinds, including a drop in revenues driven by the halving as well as a drop of more than 50% in hash price over the cost of the quarter. For the quarter, we had a GAAP net loss of $15 million, a sequential decrease of 138% and a 16% decrease from the prior quarter, when we reported a net loss of $13 million. In the current quarter, we mined 563 Bitcoin, generating revenues of $37 million at an average price per Bitcoin of $65,000 compared to 924 Bitcoin in the first quarter of ’24 at an average price of $52,000 for $48 million in revenues, a sequential decrease of 24%. Year over year, our revenues increased 18% primarily driven by the increase in Bitcoin price partially offset by the halving in April. As I mentioned at the outset, our fixed price power is a critical contributor to our attractive unit economics. In the current quarter, the cost of revenues declined 4% sequentially. When comparing revenues in the current quarter versus the same quarter in the prior year, you can see that the cost of power on a percentage basis was down significantly by our growth over the past year. This is primarily attributable to our fixed price PPA at Odessa, which we’ve talked about in previous quarters. The value of that contract increased by $22 million this quarter alone, underscoring the inherent value of the power arrangement. I should point out that as the value of the PPA is in large part driven by the time remaining on the contract and the expected future energy prices, which are seasonal and volatile, it would not be surprising to see a decrease from quarter to quarter in periods going forward. Moving on, as you recall, we adopted the new crypto fair value accounting standard in 2023, and with the drop in Bitcoin price in the quarter, we recorded a loss of $21 million of the fair value of our Bitcoin holdings; however, this mark to market loss was offset by $5 million of realized gains from the sale of Bitcoin in the period. Our philosophy towards the growth of our Bitcoin inventory and approach to treasury management has not changed. We remain optimistic about the long term outlook for Bitcoin and believe there is significant advantages to growing our Bitcoin inventory beyond mere price appreciation. We’ve discussed this in previous quarters, but it’s worth reiterating: we maintain an opportunistic approach, continually assessing various funding avenues for our growth initiatives. While we generally aim to increase the size of our Bitcoin inventory over time, our decisions are guided by the markets and our over-arching capital allocation strategy. We constantly assess the markets to find the most attractive forms of capital available, weighing the pros and cons of different funding methods to support our business and expansion plans efficiently. This might include using our cash reserves, Bitcoin holdings, or issuing equity. Through this ongoing evaluation process, we determine our estimated cost of capital and manage our treasury dynamically. At June 30, we held 2,200 Bitcoin in treasury. As in previous quarters, I’d like to spend a minute on our G&A expenses and our philosophy for managing these costs. Last quarter, we began reporting compensation and benefits as a separate line item on the face of the income statement. Compensation and benefits increased $3 million sequentially to $16 million, primarily driven by share-based compensation. Current quarter versus prior year quarter increased by 29% primarily due to an increase in headcount and share-based compensation. Now onto general administrative expenses, which include IT, corporate insurance, professional fees, occupancy, and other public company expenses. These costs increased by $2 million driven by professional fees and public company expenses primarily related to Sarbanes-Oxley compliance. Current year quarter versus prior year quarter, these expenses were down 3%. As we have stated previously, building our team and technology stack are pivotal to maintaining the competitive edge, and as such, we have significant investments in both personnel and technology, anticipating that our model will scale effectively as we increase the total megawatts under management. We expect these investments to drive future top line growth, thereby positively impacting our bottom line. Depreciation and amortization expense totaled $20 million, an increase of $3 million or 17% from the prior quarter and a 41% rise compared to the second quarter of 2023. The sequential increase was driven by our change in depreciation schedule for our miners. Previously, we accounted for the depreciation of miners over a five-year period, however given our recent fleet upgrade and the rapid efficiency gains with next-generation miners, we now believe that a three-year depreciation schedule is more appropriate. Our expectations for upgrade cycles and our ability to purchase and install much more efficient machines have evolved, and we believe this should be reflected in our accounting treatment for the entire fleet. We made this change effective June 1 and accounted for it prospectively. When comparing current quarter versus prior year quarter, the increase primarily relates to additional assets placed into service in late 2023 to complete Odessa. Now let’s turn to our non-GAAP measures slide we use to reconcile adjusted earnings. As always, I will remind you that adjusted earnings exclude the impact of depreciation and amortization, the non-cash change in the fair value of our derivative assets, deferred income tax expense, the non-cash change in fair value of the warrant liability, and share-based compensation. These supplemental financial measures are not measurements of financial performance in accordance with U.S. GAAP, however we believe that these non-GAAP measures may be useful to investors for comparing our performance across reporting periods consistently. Internally, management uses these non-GAAP financial measures to better understand, manage and evaluate our business performance and to facilitate our operational decisions. When adjusting our second quarter GAAP net loss, we add $12 million for the items I just listed. This brings us to an adjusted net loss of $3 million for the quarter compared to an adjusted net income of $63 million in the prior quarter and $8 million in the second quarter of last year. Now let’s turn our attention to the balance sheet at June 30. Our total current assets amounted to $309 million, an increase of $154 million from the end of 2023. Our cash position grew to $123 million, an increase of $36 million at the close of 2023. Our current liquidity position was $211 million, comprised of $71 million in cash and $140 million worth of Bitcoin. The decrease in our cash position subsequent to June 30 was primarily driven by deposits related to miners we contracted to purchase and capital expenditures related to the build-out of our Black Pearl data center. I’ll now cover quickly some of our balance sheet line items at June 30. Prepaid expenses amounted to $4 million, flat from year end. This balance is primarily related to corporate insurance. We recorded a Bitcoin balance of $138 million, reflecting the 2,200 Bitcoin held in treasury. This figure marked an increase from the 780 Bitcoin held at year end valued at $33 million. In the second quarter, we liquidated 153 Bitcoin worth $10 million. Now I’d like to shift our focus to the value of our Odessa power contract, which we record as a derivative asset. We’ve discussed in the past the significant competitive advantage provided by this contract, enabling us to be a low cost producer of Bitcoin. As a reminder, we began reporting third party mark for this agreement in the third quarter of 2022. This mark is reflected as a derivative asset on our balance sheet and is subject to revaluation each reporting period. Essentially, it represents the in-the-money value of the contract relative to the time value and prevailing forward power prices at our Odessa facility. As of June 30, this asset was valued at $123 million, reflecting a $22 million increase in the second quarter and an increase of $29 million from year end. This change is recorded as a gain on our statement of operations. As always, fluctuations in the fair value of this contract will impact our GAAP earnings but we exclude it from adjusted earnings. Other significant assets include property and equipment totaling $239 million, primarily attributed to our Odessa facility. Within this category, mining rigs and related equipment accounted for $177 million, leasehold improvements are valued at $137 million and construction in progress at $20 million. These figures are net of $95 million in accumulated depreciation. Deposits on equipment of $58 million primarily consist of progress payments we’ve made in accordance with previously announced miner purchases. Additionally, we hold intangible assets totaling $9 million with $7 million attributed to the Black Pearl site and its associated ERCOT approval, and the remaining $2 million related to capitalized software. At the end of the first quarter, our equity investee interest in Alborz, Bear and Chief JVs stand at $50 million, and we had operating lease obligations of $10 million. We had security deposits totaling $22 million, which includes the $13 million of collateral posted at our Odessa power provider and $6 million deposit to Encore related to the construction of our new Black Pearl data center. There were no significant changes to the liability side of the balance sheet from year end, and as we reported in the past, we have no debt that hinders our capital structure. As always, we look forward to updating you in greater detail on our growth plans over the coming quarters. I will pause now, and Tyler and I are happy to answer your questions.
Operator: Thank you. [Operator instructions] Our first question comes from Brett Knoblauch with Cantor Fitzgerald. Your line is open.
Brett Knoblauch: Hi guys, thanks for taking my questions, and congrats on building up the infra pipeline, it’s nice to see. Maybe this one’s for Tyler. First, I think there’s been some speculation about Cipher being an acquisition target out there. Maybe you could provide any color or comment on that, and then secondly to that, could you provide some additional color on the M&A landscape and how you view maybe organic site acquisition versus inorganic opportunities out there, especially post halving? Thank you.
Tyler Page: Sure Brett, thank you very much for the question. Yes, this has been a popular topic. I think as far as market rumors about Cipher, I will stay consistent with what the company has been saying all along, which is we have no comment. It shouldn’t be surprising to anyone that the press would speculate that we’re a popular target, given all the growth and opportunity we just talked about on the call. We’re very bullish on our prospects, but the company has not put out anything and we will continue to have no comment on market rumors. That’s for the press to handle. I guess speaking then more broadly about--because I’ve talked about M&A on past calls, I think what we’re seeing now is there is a fair amount of M&A in the industry. It’s really the expected chaos that’s been unleashed by the halving, and I think it forces miners to sort of question what they are, right? If we see all-time low hash costs over the last quarter, a miner has to decide to do one of two things--excuse me, an all-time low hash price over the last quarter, so a miner has to decide what they’re going to do with their own hash costs, so that the cost that they spend to produce a unit of compute, you can kind of go in two directions. One is to basically strip out all the overhead you possibly can to take it down to really just your power and basic operations cost, so that you can continue to make a profit margin, or alternatively you lean into growth, you get a more efficient fleet, you add top lien hash rate and drive down your hash cost because you’re spreading it across more terahash and you’re pumping that electricity through more efficient machines. That is definitely where we’re going. To give a sense, our recent hash cost was about $43 per petahash per day - that breaks down roughly into $19 or so of power cost and $24 of ops and G&A. I view that as--like, the $24 of overhead there is because we’re a development company. We’re adding these new sites, we’ve got all these expansions coming in the near term, and that will drive down our costs significantly. To give you a sense, just looking at hash costs just post our Odessa upgrade, which is just a matter of rigs being delivered between now and year-end and us swapping out older rigs, just doing that will likely drive our hash costs to below $30 per petahash per day, and so if you have the opportunity with low power costs, now is a good time to buy equipment, invest in scale, and I think there’s a lot of growth opportunity. I think the other thing you could do is try to strip out all the costs, and then you’re really just a site that’s trying to maximize your economics at a site, you’re not a development company. I think in general, that’s what is driving a lot of the M&A, is folks trying to get to scale and maybe address their own shortcomings and their ability to do some part of the entire value chain. We do that all in-house, so we source our own sites, we take it from dirt to an operating data center, and then we trade and manage curtailment and energy around the operations once it’s done. What has happened is--I’ve talked about M&A, we’ve been very busy this year. We’ve been in a lot of data rooms looking at opportunities for sites, companies to acquire, etc. What has generally happened is we get outbid for built sites. Frankly, we think people overpay for built sites, and that may be rational for other companies because they don’t have the capabilities to build--source their own sites and build their own sites. We do, so I think we tend not to pay a premium, and often we don’t make it to the final bidding round. What that’s driven us towards is finding more sites earlier in the development chain. That’s where we see the highest return on our investment, and so from an M&A perspective, that’s where we’ve ended up allocating our time and efforts, and that’s why you see the new sites that are coming online.
Brett Knoblauch: Perfect, and then maybe just following up on the timing of new sites coming online, so Black Pearl, call it middle of next year. Can you provide an update on Cotulla and the new 1.5 gigs of sites that, I guess, you guys have kind of secured power for, or have options on, what’s the timing of those potentially being energized and how should we think about maybe financial outlays for that and when that would occur?
Tyler Page: Yes, so look - we’re really excited about it, but I think everyone needs to understand that this is a longer term opportunity. These sites are two and a half years away basically from energizing. Now, the good news about that, and I think as we think about the HPC business, we’ve spent a lot of time talking to potential tenants and thinking about how we would build a data center with three 9s or five 9s of uptime, and the supply chains you have to manage for that. Those supply chains go out years, so I think--I know there’s a lot of buzz in the industry about people throwing in HPC revenue. All I’d say is if you’re going to do it by providing infrastructure to a large tenant, it takes a long time anyway before that revenue is going to materialize, and so the opportunity set here for us is those are well beyond Black Pearl. We have really a best-of option. We like Bitcoin mining, these sites would be great for Bitcoin mining. They are also very large scale and have the necessary components for HPC. We’ve had so much interest in that space, we’ve spent a lot of time mapping out what it would take to build a data center with five 9s of uptime from scratch, and it lines up with that timeline. These are longer term projects, but I think where we see the industry going in the coming years is this value for interconnect, value for data centers that are in desired locations and have the desired qualities and we’re like a best-of option between the two businesses, but that’s part of the long term positioning for Cipher.
Brett Knoblauch: Perfect, appreciate it. Thanks for all the color, guys.
Operator: Thank you. Our next question comes from Bill Papanastasiou with Stifel. Your line is open.
Bill Papanastasiou: Hey, good morning gentlemen. Thank you for taking my questions. For the first one here, I was just hoping, Tyler, you’d be able to share your philosophy on the HPC AI opportunity and how you’re weighing the capital allocation strategy going forward. How might it look relative to Bitcoin mining now that you have these options in hand?
Tyler Page: Sure, thanks Bill. I guess first of all, since there’s--I guess some of our competitors are taking different approaches in this space, so let me distinguish what we’re interested in. We do not plan to buy our own GPUs and be a seller of compute - that is not a business we think is very complementary to our skill sets, and it has its own risks and enormous capital outlays associated with it, so that is not our plan. Our plan is to be effectively a host for HPC infrastructure, and so what we hope to do is build out sites and have hyperscaler or other large high quality tenants to sign a long term lease. Now, it’s still a little early. I have been very busy speaking to both the potential tenants and a lot of financiers. The market is extremely busy and interested in this. The thing to keep in mind that’s very different about the two businesses is there’s different markets and different times where we may make more money Bitcoin mining or we make more money in selling HPC infrastructure. The attractiveness of the HPC infrastructure opportunity is that those tend to be long term leases and obviously are not correlated with the ups and downs of Bitcoin price. The other thing is even though that’s a much more capital intensive business, there is quite a developed financing market with some of the biggest lenders and investors in the world, and if you’re focusing on the piece of the business that we are, there’s a lot of appetite to finance debt, and even equity if you set up a development company, to do very high LTC loans - you know, up to 90% of the cost might be loaned if you have a hyperscaler tenant at a very reasonable interest rate. Then as you develop more and more sites, you have the opportunity--you know, as they mature and have active tenants, there is a whole asset-based securities financing market that’s available, so it is a very different business in how it operates compared to Bitcoin mining, where lenders have shied away from Bitcoin mining for the past two-plus years, and maybe that doesn’t stay that way forever, but the successful Bitcoin miners, including Cipher, have pretty much entirely funded the build-outs with equity. Now, I think we can see really good ROI potentials on sites that you own and operate - you know, if you look at Black Pearl, of course everything is going to depend on what happens to hash price and Bitcoin price going forward, but we look at ROIs on Black Pearl of a year and a half to three years, depending on what happens on the market, and that’s great. I think going forward, we do have time. We’re going to launch the HPC business. I think it starts with getting a really high quality tenant that wants to sign up for a lease at one of these sites. We have had a lot of time with some of those big names. Our construction and ops team, and I know you know this, Bill, because we went over it recently, but we’ve got a lot of people that are ex-Google, Vantage, Meta, have operated hyperscaler data centers. We’ve had all-day meetings with technical teams, senior management with hyperscaler or potential tenants. I think our team is very well situated. This is a big competitive advantage for us, so if that helps us secure a lease in the near term, where a lot of design input can come from the potential tenant, pretty soon we would be dedicating those sites, whether it’s the site in Cotulla or the future sites under option, and then seeing where that business goes. Again, Black Pearl, we expect to have a very, very robust Bitcoin mining business next year. I think in a perfect world, we will offset it with a very nice long term HPC tenant contract, and then we’ll see as it goes. The great thing about those sites that we’ve got in the pipeline is they’d also be wonderful for Bitcoin mining. I think the long term prospects for an HPC tenant are kind of nice. If you look at the longer term leases, they probably in the back years of those contracts will exceed what Bitcoin mining data centers are likely to produce, because hash price over long time periods tends to go down. There’s a cyclical nature around halving, we’re very bullish over the next year and a half for hash price expansion, but the long term trend is that you’ve got to get more efficient, and so the great thing about HPC revenue would be the sort of back end of a potential lease, where the numbers are likely to be higher in dollar terms that you’re being paid for producing that--providing that data center.
Bill Papanastasiou: Awesome, I appreciate that color. I recall visiting the sites over a year ago and was really impressed with the team there in Odessa. Just shifting gears here, Tyler, Cipher has been able to leapfrog a number of peers and is now a top five miner on account of scale with 35 exahash, so just focusing on the Bitcoin mining aspect of the business, I’m curious to hear how important it is for Cipher to maintain this level of network market share relative to peers going forward. How should we think about that?
Tyler Page: Yes, I think that’s a good proxy for what we want to do, but we don’t really do capital planning in terms of what’s our market share. I guess we do as kind of a derivative of the way we think about it, but we’re really focused on what’s likely to happy to hash price and what will be our hash costs to produce the hash, and that margin, that spread between those two numbers is really the business. Listen - the foundation of that, as I harp on, on every one of our earnings calls, is the low cost power. That drives--you know, that’s the number you cannot change or tweak, and so once you lock that in, you can then be opportunistic through the cycle to do things like improve the efficiency of your miners, so I would say versus the last couple years, we’re still in a relatively cheap pricing period to acquire new rigs, and so the pick-up in efficiency can be very large for what is, at least compared to historical numbers, a smaller capital outlay. We tend to look at projections on--you know, can we lower--how much do we lower our hash cost, what’s the capital outlay to produce that, what’s the payback period? Now, I guess the knock-on effect of that is it implies we will take market share, and certainly thinking about what happens to network hash rate will determine the market share, but that’s all kind of accounted in the forecasting for hash price, so if Bitcoin prices go up, there will be more network hash rate, you’ll be fighting for market share and you might have less of it, but of course Bitcoin prices will be higher. Really, the dynamic we look at is forecasted hash price versus what our hash costs will be.
Bill Papanastasiou: Appreciate the color, Tyler, and congrats again on those HPC AI announcements.
Tyler Page: Thanks Bill.
Operator: Thank you. Our next question comes from John Todaro with Needham & Company. Your line is open.
John Todaro: Great, thanks for taking my questions. I have two of them, I guess first on the Bitcoin business and then the down the road HPC business. Tyler, how do you think about allocating costs and investing in the business, and I guess another way to put it is how should investors be thinking about? Is it on more so a Bitcoin cost basis in this post-halving environment, or that cost on a hash basis as you’ve mentioned a few times here?
Tyler Page: That’s a good question, John. I think this is--as investors get more savvy in the sector, I generally think the better way to think about things is hash price versus hash cost, because it accounts for the multiple variables, especially if you’re projecting forward. I don’t think the right thing to do is project forward on a cost per Bitcoin basis. That said, we get this question, and almost big institutional investor that’s maybe newer to the space, they always start with a framework of what’s your cost per Bitcoin, right, because everyone follows the Bitcoin price. It’s easier to think of in that sense. I guess on a backward looking basis, I’m comfortable doing that because you don’t have to think about, well, what happens to network hash rates? If we just look post-halving, if I translate our hash cost and hash price--hash cost numbers, rather, and I’m backwards looking just post-halving because that’s what’s most relevant, our power cost to produce a Bitcoin is about $24,000 a Bitcoin, and for our all-in cash costs, it’s sub-$50,000, and that’s post-halving in what we expect is--you know, the expected bad environment post-halving, this is what it’s supposed to be, the market that squeezes miners. This is a known-unknown when you go through the halving, that you’re going to have a rough period. I think backwards looking, I can tell you cash cost to produce a Bitcoin was sub-$50,000, and that’s with this idea that, again, we are completely focused on growth. We think these trends around data centers and Bitcoin mining--or rather Bitcoin network adoption are likely to be really big, and I know you know this and most of our investors know this, but as hash price goes up or as Bitcoin price goes up, in the moment those additional dollars per Bitcoin drop to our bottom line, so we’re very focused on expansion but half the cash costs we’re spending, I view as an investment in scale and expansion, and so those numbers will get better and better by year end when we have a fleet that is much more efficient and has a bigger top line hash rate, because of the upgrade at Odessa. Then in 2025, we’re projecting about a 15.5 joules per terahash efficiency across the entire fleet and potentially nearly tripling the top line, so that investment today that we sort of make in the people that originate these sites and build these high quality data centers, that’s now defrayed over a smaller, less efficient fleet. We are very much on the growth side of--growth philosophy of driving down that hash cost, and then that translates to Bitcoin. But for a metric in the ugly months post-halving, which are expected to be ugly, we’re spending less than $50,000 of cash, and that includes what I’ll call this investment in the forward nature of the business and expansion, as well as the cash costs to actually pay for electricity and mine the Bitcoin.
John Todaro: Got it. Thanks for that. Then just my other one, real quickly on HPC, we spent a lot of time due diligencing different sites and locations. In Texas, there is a little bit of a concern around curtailment. Do you guys envision this site, if it is for HPC, say the full 200 megawatts for HPC which it sounds like is what you guys are leaning towards, do you envision curtailment at that site, and especially as you think about the broader 1.5 gigs, does that actually impact ERCOT approval? Have there been questions about that piece, any kind of color on that?
Tyler Page: Yes, I mean at a high level, I’d separate it even within ERCOT. You apply to be a flexible load or an inflexible load. Part of the approval to do what we do at our Bitcoin mining data centers is getting what’s called the LFL approval, the large flexible load approval, where mapping out curtailment is a big part of how they think about balancing the grid down there and the particular areas of transmission and distribution. The HPC opportunity, as you know, these have availability to power 100% of the time, and even the site at Reveille, the site in Cotulla, Texas, that set-up is a front of the meter site where, again, for Bitcoin mining, if we were to do Bitcoin mining there, we would absolutely manage curtailment because you’re--you know, when we manage curtailment, we’re stripping out the highest prices we would be paying for power and cutting our overall average power cost nearly in half by stripping out nearly 5% of the most expensive hours. HPC is a very different business - you’re being paid for that 100% uptime, and really most of the capex goes into that additional reliability of uptime. I mean, to give you a sense, when we say three 9s of uptime, of course I’m sure everyone knows, but when you’re thinking about 99.9% uptime and then going out to five 9s, which is more typical for, say, a hyperscaler, so 99.999% of uptime, the cost goes up dramatically, right? You might be spending $3 million to $4 million per megawatt for infrastructure to get three 9s of uptime, whereas you could be up to $10 million or even more, depending on the real estate location, for five 9s of uptime. But the point is, that is generally all built around the idea that you are not managing curtailment. Now, what I’ll say, I did allude to this earlier on the call, this is a very dynamic market. A year ago, I think hyperscalers would only maybe look at real estate in a place like the Dulles corridor and want five 9s of uptime, and if you talked to them, they may have 300 line items that they require for redundancy and very specific technical design requirements. What we’ve seen is there’s now such a drive for megawatts, and it’s kind of dynamic, megawatts that are available sooner are even more valuable. You’ve seen a lot more flexibility on the side of the potential tenants, and so I think playing that out, what that means and could be interesting over time is maybe that trend continues, and in the future curtailment becomes part of HPC. Maybe they say, hey, if you could drive down our costs, maybe we’d be willing to give up a couple percent of the uptime - it all remains to be seen, that’s pure speculation. I think we’re really well positioned if that is what happens to the market, given just how much of that is our core business. We’ll see, but based on let’s call it the other compromises on location and some of the other technical requirements, it wouldn’t surprise me if we see a market that continues to evolve on the HPC side in terms of what’s required.
John Todaro: Appreciate that, thanks Tyler.
Tyler Page: Thank you.
Operator: Thank you. Our next question comes from Mike Colonnese with HC Wainwright. Your line is open.
Mike Colonnese: Hi, good morning guys, and congrats on all the recent deals. Definitely a lot in the works right now for Cipher. I know it’s still early days, but how should investors think about the revenue model and the economics for this upcoming HPC business, based on your planned go-to-market strategy which sounds like you’re going to take more of a co-location approach here?
Tyler Page: Yes, thanks Mike. I’d say, listen, first off I’ll try to give you a sense for what we’ve seen in preliminary discussions, but it’s very clear to us that deals we would strike are pretty bespoke, so I don’t know that the deals we’ve seen would necessarily be indicative. But I think what we’re seeing is gross margins in that business of 80%, and then financing quite a bit of that capex, depending on the stage you are, who the lender would be. Anecdotally, I can tell you there’s quite a few investors that are out there, that are interested in earlier stage financing, maybe when you’re just building a data center or you have, say, a letter of intent from a large high quality tenant. There is a much more developed market if you’ve got a lease from a high quality tenant that is, like, bank financing, and much more reasonably priced. These are all kind of dynamics we’re looking at. Suffice to say investors are very interested in just the overall breadth of our portfolio and the potential capacity to do 1.7 gigawatts in the coming years has a lot of people excited. I think it’s really hard for us to size specifics because we’re just doing lots and lots of meetings on this right now, trying to move quickly. For what it’s worth, that’s what we’ve seen on the type of business we’re looking at, and we’ll probably have more details as specific conversations advance.
Mike Colonnese: Got it. Thanks for that. Just curious what made you guys decide to go with the liquid cooled infrastructure for 50 of the total megawatts over at Black Pearl versus going air cooled for the full site, and how are you guys estimating out the capex-related costs for liquid cooled versus what you would otherwise see for the air cooled? Thanks.
Tyler Page: Yes, great question. Look, we’ve been modeling a lot of different operating strategies in the test kitchen for a while now. Obviously Texas is hot, and so managing that heat and efficiency and all the various ways you can under-clock and try to optimize your performance are all part of the analysis. At a high level, air cooled tends to be the cheapest capex, and then the question is--you know, also the opex is low but you’re probably going to have worse efficiency and production. The question is where do you cross over to make a bigger upfront investment worth it. For us to do 50 of the 300 megawatts in hydro, you’re talking about an incremental spend of probably close to $20 million, using round numbers. I think what’s important to us is, number one, we think we’ve arrived at a set-up and a system that’s going to be best-in-class and will easily justify the incremental spend when we operate there. I think the other thing is it’s important for us to have a data center that, frankly, we can show off where we’re doing hydro, because there’s implications of that for HPC, obviously. I think as part of a large scale first step, we wanted to do 50. Now that said, we know air cooling pretty well. We know with our curtailment and energy trading strategy, we can operate well in Texas with air cooled, and so 50 megawatts ended up being the Goldilocks just right amount, that it’s large scale but it’s not increasing capex in a crazy way with new way of operating. I’m sure as good as we are, we’ll certainly stub our toe and learn some lessons along the way, but overall that’s kind of the incremental pick-up for the 50 megawatts of hydro that we’re going to do there.
Mike Colonnese: Great, thanks for taking my questions, Tyler.
Tyler Page: Thanks Mike.
Operator: Thank you. Our next question comes from Tyler DiMatteo with BTIG. Your line is open.
Tyler DiMatteo: Yes, hi guys, good morning. Thanks for taking the question here. Tyler, I’m curious - on the HPC opportunity and the sites, I guess at a high level, can you just add some color in terms of the actual sourcing process and your strategy for that, maybe who you’re bidding against, how early does the process have to start? I guess what I’m wondering, you hear a lot about HPC, I’m curious is there a sweet spot for megawatts? I know it’s early days for customers, but is there a sweet spot as well? Any thoughts there broadly?
Tyler Page: Sure, so I’ll go in reverse order on those questions, Tyler, thanks. The largest tenants, let’s call them hyperscalers, are probably not looking sub-100, and really would prefer 200-plus megawatts. I did mention earlier on the call, we’ve had a lot of people ring us up and ask about Black Pearl - hey, could we take that for HPC and ditch your Bitcoin mining plans? The answer has been no. I think from a scale perspective, that’s the most attractive thing about this portfolio of sites we’ve sourced, is that we’re hoping they are all approved for 500 megawatts of interconnect. That process is not done, but we’re very constructive on it, and I think the good news is we came up with a pricing framework that we’re only paying for the amount of megawatts that get approved. Our hope is that even if they don’t get to 500 each, they are several hundred and in the ballpark, and they are all going to be very attractive. I expect to have more color on those in the coming months. I expect--well, I hesitate to give exact timing because sometimes approvals processes take a while, but in the coming months, I think we will have more color on some of that as far as what gets approved. To take your first question on sourcing overall, I think this is something we take a great deal of pride in and I’m very proud of our team. It’s no secret to everyone on the call, it’s an extraordinarily hot market to source opportunities. You add on top of that that in Bitcoin mining specifically, it’s a very active M&A environment, lots of assets are for sale, companies are for sale, etc. I think anyone can go buy a finished product, and what you see is, like I mentioned, people overspend in our opinion. I think Bitcoin miners are optimists, and so when you’re looking at an asset, you can price it with rosier assumptions about what’s going to happen to Bitcoin and justify a higher cost. We tend to be a little bit more bearish in our forecasting just to protect the downside. It’s not what we actually think is going to happen in the real world. What we’ve observed is there are not very many of our competitors, and even this is true on the HPC side, like the hyperscalers have departments of people sourcing opportunities, but they’re not necessarily experts in all the local rules for getting a site approved for interconnect, and so what we’ve found is there’s just not as many people fishing in that pond and we’ve been pretty consistent on that. All of our sites were sourced as un-built dirt patches, and that’s continued, and you’ve seen that with Black Pearl. We got a great 300 megawatt opportunity at Black Pearl for $7 million. Now, the downside to that is they’re not ready as soon. Sometimes I know our Twitter antagonists say we move too slowly. I think we’re just moving very deliberately to build what is ultimately the best company in this space. It’s just that good things take time.
Tyler DiMatteo: Okay, great. Thanks for answering the questions there, Tyler, really appreciate it. I’ll turn it back to the queue here.
Operator: Thank you. Our next question comes from Reggie Smith with JP Morgan. Your line is open.
Reggie Smith: Hey Tyler. Thanks for taking the questions. First, I wanted to give you guys credit for managing operating expenses. It’s not lost on us that you guys are holding the line there pretty well. I guess just based on the last answer, it sounds like your M&A team is the real MVP, and you guys need to make sure that they don’t leave and go to a hyperscaler. But my question I wanted to ask you, thinking about what you’ve announced today, and I’m trying to, I guess frame the capex requirements for that, I think if I looked at the 1.5, 1.8 gigawatts you guys have announced today, if you were to build that as a miner, my math is that it could cost roughly $3 billion to build out. Thinking about HPC, rough math, I think for every 100 megawatts, you’re looking at maybe a billion or $2 billion in just infrastructure build out. One, I wanted to see if those numbers are correct; and then two, with that as kind of a backdrop, how do you think about balancing capex and returning capital to investors, or holding cash or whatever it is? It seems like there’s a heavy period of investment that we’re kind of looking at, and I’m curious how you balance that, so the question is are those numbers accurate, and then two, how do you balance the need for capex with the need for returning cash or generating free cash? Thanks.
Tyler Page: Sure, thanks Reggie. Let me say before I answer your questions that I’m very blessed - I have a lot of MVPs here. I have about the easiest job in the world because I have really good people in sourcing sites, but also we’ve got MVPs in building the sites, operating the sites, trading and hedging, etc. All I have to do is watch our great people do great work, so we’re sort of an all-star team here that I get to just coach. That said, quick math on your numbers, yes, I’d say on the HPC side, that sounds ballpark correct. It is a very capital intensive business, and that’s why when it comes to thinking about investment returns, it’s really driven by the funding there. That’s why we mentioned this is one of the legs of the stool - you’ve got to have the sites, you’ve got to have the team that can do it. Again, we’ve already got the sites, we’ve got the team already hired, we don’t need to go out and hire people that know how to build and develop HPC data centers. The third piece is the capital, and it’s early on that. I’d say we’re comfortable talking about this business at such an early stage because of just the amount of interest from very high quality lenders and investors that have been all over us for this business. So yes, the capital outlay is enormous if you think about a 1.7 gigawatt-type portfolio, but the only way that will be built by Cipher is if lots of it is financed, debt financed, and that could include--you know, we’ve got multiple proposals for setting up devcos that would finance separately from the Bitcoin mining business. There is a lot of structuring analysis and optimization that needs to be done, and none of that is finalized, so you’re highlighting one of the main challenges, is just like wow, that’s a lot of capital if you’re going to do that all for HPC, and you would definitely not finance that business the way we finance Bitcoin mining. Now separately, running the numbers on--without thinking about--I don’t know that I can do the math on the fly on what Bitcoin mining would likely to be on 1.7 gigawatts. I think--let me give you some framework for how we think about return on Black Pearl. We look at the build costs, and that includes obviously where you’re locking in rig prices as a massive part of the cost there. We look at ROIs under various forward hash price scenarios, and again we look at ROIs that we’re targeting a year and a half to three years payback period on Black Pearl, so that’s going to be a similar framework for how we decide what do with any data center for Bitcoin mining. Now, what becomes different is if we’re thinking about these sites that are maybe two, three years off, how we finance them, I hope that someday reasonably priced debt financing becomes available for Bitcoin mining. I think it will - you need Bitcoin to be more generally accepted, maybe we’ve got broader acceptance across the investing universe here with continued growth of the ETFs and embraced by politicians, etc., and then you can debt finance more of it and it’s less of just an equity finance game. That said, we’ll have to see what the market looks like. I know there’s lots of avenues available to us to finance Bitcoin mining, but it’s hard to project on the entire portfolio - that’s a lot of capital. That’s why I’m pretty comfortable saying a decent portion of that is highly likely to be HPC.
Reggie Smith: Perfect, thank you.
Operator: Thank you. Our next question comes from Mike Grondahl with Northland. Your line is open.
Mike Grondahl: Hey Tyler and Ed, thank you. Over the last 90 days, it sounds like you had a lot of discussions with hyperscalers and finance partners. What was the biggest learning for you, and maybe what do you see as sort of the biggest challenge going forward with that HPC effort?
Tyler Page: It’s a good question, Mike, and I think you actually alluded to at least what I think the biggest challenge is, is that you’re pulling together two pieces to make it work. You need to figure out a path for financing for any of these large scale builds, and you need to figure out a high credit quality tenant, so we’re running both in parallel. We’ve learned quite a bit. I’d say on the tenant side, a lot of it is around the technical requirements, so our team, as I often highlight has spent a lot of time at places like Google, they’ve built hyperscaler data centers, the market a few years ago before they were at Cipher was a little bit different in that I think a hyperscaler tenant might be a lot more set in their ways about what their technical requirements might be, and to be specific, some of these potential tenants, imagine hundreds of requirements on an endless spread sheet for all the technical details of the build. It’s not to say they’re compromising any of what they want; it’s more that there’s some flexibility because I think there’s such a desire for large scale sites, and so I think what’s been interesting is learning just how flexible some tenants have become in what they’re willing to accept. The most obvious first step of that is, like sure, we’d love a data center somewhere in West Texas, like what about--I told you we got some inquiries about Black Pearl, but that’s because fiber networks continue to get built out, it starts to become questions of what’s nice to have versus must have as folks build out this massive need for capacity. I think on the financing side, still a learning process. I think it’s learning about, I think both how mature the financing market is for a full-on lease from a high quality tenant. I’m not sure I realized how quite fully developed and broadly acceptable by the biggest banks in the world that marketplace is. I think we’re still learning kind of then if you go earlier in the financing and construction financing stages, like what the marketplace is like and what people are willing to loan for in the pre-leased stages, and that’s also highly variable, so it’s an ongoing learning experience. I mean, I think my calendar is chock-a-block full of meetings with people, and there’s a pretty wide variety, actually, on what people will finance and how anxious they are to get in the business. I think the number one takeaway, I think the hardest thing is lining up the two pieces simultaneously because they’re very much connected, right? You want to get to a place where you have a full-on lease because that’s going to give you the broadest, cheapest financing options. On the financing side, it’s also just how bespoke it is, and I think it’s indicative of a market that’s just frenzied to find large scale sites that work.
Mike Grondahl: Got it, that’s helpful. Best of luck.
Tyler Page: Thank you.
Operator: Thank you. That’s all the time we have for questions. I’d like to turn the call back over to Tyler Page for any closing remarks.
Tyler Page: Well, thank you everyone for taking time this morning to join us. We’re at the cusp of a lot of growth and scale coming to Cipher, and we’re really excited about where we’re positioned, so look forward to giving future updates as everything builds out. Thank you again.
Operator: Thank you for your participation. This does conclude the program and you may now disconnect. Everyone have a great day.
