Cipher Digital Inc. (CIFR) Earnings Call Transcript
May 5, 2026
Earnings Call Speaker Segments
Hello, everyone, and welcome to this conference call, where we'll discuss Cipher Digital's business update covering Q1 of 2026. On today's call with me are Tyler Page, our Chief Executive Officer, and Greg Mumford, our Chief Financial Officer. Our press release and presentation are available in the Investor Relations portion of the company's website, and this call is being webcast there at the same time. I'd also like to point out that this call belongs to Cipher Digital, and recording it or reproducing it in any way is strictly forbidden unless we've given our consent beforehand. Before we begin, let me remind everyone that what we discuss next, along with our press release and presentation, includes forward-looking statements. Such statements cover, among other things, Cipher's financial outlook, business plans and objectives, and other upcoming events and developments, including remarks regarding the market potential of our business operations, possible competition, and our goals and strategies. The forward-looking statements and risks discussed on this call, including answers to your questions, reflect expectations held as of today, and Cipher undertakes no duty to update or revise them, whether because of new developments or for any other reason, unless the law requires it. In addition, what follows may include non-GAAP financial measures. Non-GAAP measures may be used by us to describe how we manage and operate our business. We reconcile non-GAAP measures against the most directly comparable GAAP measures, and we encourage you to review those reconciliations, which were filed at the end of our earnings release issued earlier this morning. I'll now hand the call over to Tyler Page, our CEO. Tyler?
Appreciate that, Drew. Morning to all of you, and thanks for being here with us. My name is Tyler Page, and I serve as CEO of Cipher Digital — it's a pleasure to welcome everyone to this business update call for the first quarter of 2026. For Cipher, 2026 is the year we execute, and we opened it up with a solid first quarter. We put pen to paper on our third data center campus lease alongside an investment-grade hyperscale tenant, wrapped up a $2 billion high-yield bond offering for Black Pearl that funds that project all the way through to completion, finalized a $200 million revolving credit facility from a group of leading global financial institutions, and drove meaningful advancement on construction at our Barber Lake and Black Pearl HPC data centers. Execution — that's the word that sums up this quarter, and it's what will keep defining the remainder of the year. Once we adopted the Cipher Digital name, we put out a statement: "We are Built for Hyperscale." That phrase carries real meaning for us, and I want to spend a moment unpacking what it signifies in light of the results from this quarter. Built for Hyperscale is not merely a marketing slogan. It captures the core of who this company is and the outstanding group of people we've brought together — power origination, engineering, construction management, and operations, all handled internally and designed specifically to provide state-of-the-art data center infrastructure with the speed and precision that hyperscalers demand. With each quarter that passes, we add fresh evidence. The first lease we signed at Barber Lake demonstrated the concept. The second one at Black Pearl demonstrated that we can do it again. The third lease we signed during this quarter demonstrates that Cipher Digital is itself a top-tier development platform, one Built for Hyperscale. As 2026 unfolds, with 2 data centers already under construction, a third getting ready for mobilization, and a substantial pipeline sitting behind those, the market will increasingly see what sets us apart. On Slide 4, you'll find a snapshot of where Cipher Digital stands right now; we develop and operate industrial scale data centers as a vertically integrated business, and they're built to serve the world's leading companies. Thanks to our in-house capabilities, we can deliver power dense, large-scale facilities that match exact hyperscalers specifications, and we do it at speed. Right now, our operating and contracted capacity sits at 907 megawatts. The foundation for that is three signed leases for data center campuses with hyperscalers of the highest caliber. Across base lease terms running 10 to 15 years, that portfolio holds roughly $11.4 billion in contracted revenue, which gives Cipher cash flows that are durable, high-quality, and long-term. On top of that, we have a pipeline of about 3.3 gigawatts in grid capacity, which offers a substantial runway for growth ahead. Our pipeline's scale and quality amount to a competitive edge that's hard to copy, and it reflects years of disciplined work originating power. We've moved past being an aspirational HPC developer. Signed contracts are in hand, billions in capital have been raised, and several data center construction projects are moving toward completion. Stepping back to look at the bigger picture, you can see on Slide 5 how far our development platform reaches geographically and how large it has become. Roughly 4.2 gigawatts of grid power make up our portfolio, spanning sites that are operating, under contract, and in the pipeline. Pipeline HPC opportunities account for about 78% of that capacity, and the remaining portion is split between capacity already contracted, capacity brought under contract this quarter through the third lease, and the Odessa Bitcoin mining operation we run. The West Texas concentration is deliberate; for a long time, the traditional data center sector took it as a given that hyperscalers would stay within major metropolitan areas, and that our locations were situated — to borrow a phrase I used before — at the edge of the known world. We saw things differently, and events have proven that judgment correct. When it comes to building AI infrastructure at a massive scale, West Texas now ranks among the nation's most coveted areas, and Cipher is strongly positioned to capitalize on that rare opening. Adding the Ulysses property in Ohio was a purposeful move to spread our footprint across geographies. What big hyperscalers need is data center capacity spanning several markets and power grids. With locations in both ERCOT and PJM, our pitch as a development partner grows stronger for tenants who need capacity across multiple markets. This portfolio — its size, the caliber of the locations, and how far it reaches geographically — comes from years of sourcing work done on the ground. Building something like this takes time — you can't just throw it together in a day — and it will keep giving us an edge as more and more organizations need really big data center capacity. Now let's look at where Cipher's contracted cash flow picture is headed. Across the 3 data center campus leases we've signed, we anticipate roughly $787 million in average yearly net operating income between October 2026 and September 2036. Come 2035, the contracted net operating income we expect to have stands at about $892 million. By bringing in our third lease during this quarter, we've made that contracted cash flow profile even stronger and added a substantial amount of net operating income to what we're projecting. Just to remind everyone, what we're talking about here is contracted net operating income—it isn't a forecast built on leases we assume will happen later, and it isn't a model that relies on guesses about how things might turn out. What we have are long-term agreements that have been signed with counterparties carrying investment-grade ratings, and those agreements produce contractual growth that is visible and stable across the coming decade. On this slide, you can see how fundamentally the financial character of this company has changed. Ours is a business whose defining feature is long-term cash flows that are stable. The leases have been signed, the financing has been arranged, and construction is already in progress. The cash flows shown on this chart are not something we hope for. They are contracted. Now I'd like to go over the particular achievements that shaped our first quarter. In terms of the company overall, we reached 3 major execution milestones, and each one reflects how strong and how mature this platform has become. To start, we closed our third data center campus lease — an agreement with an investment-grade hyperscale tenant whose initial term runs 15 years. That makes 3 long-term leases in a row with world-class counterparties, all within roughly 8 months. On top of that, we finished a bond offering that raised $2 billion with a 6.125% coupon, which covers the entire build-out of Black Pearl through delivery. Demand for the offering far exceeded supply, and it featured roughly $233 million returned to Cipher as repayment for what we had previously put into the site as equity. Then, to wrap up, we finalized our first-ever $200 million revolving credit facility, backed by a group of top-tier worldwide financial institutions. Cipher hit a major milestone here. Never before have we held a committed credit facility at the corporate level — evidence of how our platform has matured, the strength of our contracts, and the trust that the planet's most robust institutional lenders place in Cipher. Turning to physical execution, the numbers we posted are just as strong. Over at Barber Lake, daily active workers on site surpassed 1,100 during April, and we anticipate more than 1,400 in May. Even more significant, cumulative labor hours on site have passed 1 million, all while maintaining 0 lost time incidents. For a project of this size and intricacy, that safety record stands out as remarkable, and it speaks to the operational discipline culture our construction team has cultivated. Over at Black Pearl, demolition of the pre-existing Bitcoin mining infrastructure needed for the Phase I retrofit was finished within 1 month of kickoff. Ground was broken on Phase II a mere 3 months following design kickoff. As both sites stay on track, we have moved into the second quarter carrying considerable momentum.
Alright, let's walk through our development portfolio in more depth now. Over at Barber Lake, the build is far along, and everything we're doing there centers on execution. I'm happy to say things are moving along nicely at Barber Lake. This past April, the structure reached its top, which means the main steel framework is finished. Starting with the initial column and ending with the final structural beam, we erected a building of approximately 800,000 square feet in 127 days. That result speaks volumes about how strong our construction management group is and how deep our ties run across the supply chain. Work streams for mechanical, electrical, and networking systems are moving forward simultaneously at this point, and the schedule remains on course to hit every contractual early access and substantial completion milestone date. On the procurement side, we've now locked in about 99% of the equipment this project needs to finish. Delivery timelines for that equipment line up with the dates we need to complete construction, so the chance that supply chain issues throw off our schedule is quite small. Design work stands at 100% finished. Every design milestone was met on time, removing yet another significant construction risk at this point in the project. On the next slide, investors can see a visual of everything that's been built at this location, and I'd ask you to spend a moment examining it carefully. Slide 10 contains an aerial photo of the Barber Lake campus that was taken 7 days ago. Study this picture closely, because in my view it reveals something that figures and bullet points simply can't get across. Eight months back, this location was nothing but open field in West Texas. As of today, it ranks among the biggest data center builds currently underway anywhere in the United States. What this photograph reveals is enormous in scale — a campus designed to provide 207 megawatts of critical IT load to a number of the most advanced technology firms on earth, emerging from the dirt with a pace and accuracy that, in our view, has no equal in this sector. When I reflect on the path that brought us here — all those years spent finding sites, negotiating leases, arranging financing, doing design and engineering work, handling procurement, managing construction — and then I look at what now stands on this land, I feel real pride in each person on this team; this is what Built for Hyperscale means when you see it happen. We're really looking forward to having our tenants on the campus in the latter part of 2025. Black Pearl is moving ahead with the very same discipline and speed that Barber Lake has shown. Every piece of Bitcoin mining infrastructure has been taken offline at this point. The site has fully transitioned into data center development mode. The retrofit's Phase I is progressing nicely, and mechanical, electrical, and networking work fronts are all fully underway. Crews are working hard to install the cooling systems and electrical gear so that the older building can handle the new HPC hardware. On the procurement side, we've secured equipment for about 93% of Phase I, and delivery schedules align with the completion dates we're aiming for. Turning to Phase II: in April we began work on site layout and earthwork — just three months after the design kickoff. In addition, roughly 80% of the equipment needed for Phase 2 has been secured, and the delivery schedules align with the completion dates we are aiming for. Taken together across both phases, the project is still positioned to meet the contractual early access and rack ready dates, and our confidence that we will deliver this campus on time has remained steady. On Slide 12 you can see an initial look at how construction is progressing at our Stingray site in Andrews County, Texas. As a reminder, that location carries 100 megawatts of gross capacity that is fully approved, with energization targeted for the fourth quarter of 2026. Work began at Stingray in Q1. At this moment, earthwork and pad preparation are in progress. Substation electrical work is now underway, which lines up with our timeline for energization in Q4 2026. We look forward to providing additional updates about this site as construction mobilization continues. Mining of Bitcoin continues at Odessa, and that location delivered a solid first quarter. A reminder that Odessa's fixed price power purchase agreement of roughly $0.028 per kilowatt hour maintains Cipher's position among the industry's lowest cost Bitcoin producers. Currently, we are operating 207 megawatts of capacity, generating approximately 11.6 exahash per second of total hash rate with a fleet efficiency of approximately 17.2 joules per terahash. In Q1, we produced roughly 346 Bitcoin at our Odessa facility. That mining segment funds itself entirely. Additional money for that side of operations isn't something we anticipate needing, because our focus continues shifting the platform toward HPC. As leases at our data centers progress toward the point where revenue begins later in the year, Odessa continues generating solid cash flow. Now let's turn to what's happening with our development pipeline. If you look at Slide 15, I'd like to go through the particular sites in our near-term pipeline and help investors understand the current status of each. Among our precontracting opportunities, Reveille and Ulysses are the furthest along, and both have received full interconnection approval. Our Cotulla, Texas site, Reveille, has gotten the green light from ERCOT for interconnection at 70 gross megawatts, and we've kicked off substation work there. Talks are underway and fairly far along with several prospective tenants about an HPC hosting lease at that location. Because Reveille's capacity sits under the level that would push it into ERCOT's batch process, and because its interconnection approval is already secured, the Q3 2027 energization target for this site won't be exposed to batch process risk. The load is firm, the approvals are secured, and we're working hard to turn this site into a contracted asset. Ulysses, which is our 200-megawatt location in Southeastern Ohio, holds every approval required to take part in the PJM market, and in a similar way we're in advanced talks with potential tenants regarding an HPC hosting lease at this location; this marks Cipher's first site within PJM, and it fits HPC data centers well. We're aiming for energization during Q4 2027 and stay extremely confident about that schedule. Beyond these near-term locations, we have McLennan, Mikeska and Colchis. Every one of them is moving along the ERCOT interconnection process and performing on track. We keep driving all required workflows ahead, paying all deposits when scheduled, and making sure these energization timelines hold. Each of the three sites is expected to come online at some point in 2028, and per ERCOT's new batch procedure, every one of them falls into batch 0. We have a great deal of confidence that our pipeline is extremely well positioned, and we continue to speak proactively with prospective tenants across all of these sites. Slide 16 brings the entire portfolio together in one picture and illustrates just how extensive the platform we have constructed truly is. Our current operating and contracted capacity appears on the left-hand side: 207 megawatts of Bitcoin mining at Odessa, 300 megawatts contracted at Barber Lake together with FluidStack and Google, 300 megawatts contracted at Black Pearl together with Amazon Web Services, and an additional 100 megawatts contracted under the third lease we recently executed. Altogether, as of right now, that adds up to 907 megawatts in gross capacity that's either operating or under contract. The capacity timeline on the right lays out what lies ahead: inside the 2027 timeframe, Reveille plus Ulysses together make up 270 megawatts of pipeline capacity for the near term, while from 2028 through 2029, Mikeska, McLennan, Milsing, and Colchis add another 2.5 gigawatts of pipeline capacity.
And as we think about 2030 and past that point, another 500 megawatts at Barber Lake is a bigger opportunity still, located at a site that's already contracted and running. When you put it all together, that adds up to as much as 4.2 gigawatts of total portfolio capacity coming from the grid. What we've seen this past quarter has only made us more certain; in our view, Cipher ranks among the best positioned companies anywhere in the world when it comes to turning this pipeline into contracted long-term cash flows, and we're eager to show that in the quarters to come. At this point I'll hand things over to Greg Mumford, our CFO, who'll take you through our financing activities, capital structure, and financial results for the first quarter. Greg?
Thanks, Tyler, and good morning to all of you. During the last 12 months, Cipher has made major moves to transform its financial shape, shifting from a fledgling Bitcoin miner into an institutionally backed digital infrastructure platform featuring long-term contracted cash flows and a capital structure built for that purpose. In Q1, we kept reinforcing that financial base in ways that substantially reduce execution risk and sharpen forward visibility; we came into 2026 centered on strategic capital allocation, ring-fencing construction risk through nonrecourse financing, and boosting our corporate liquidity. During the first quarter, we closed 2 more financings: the $2 billion Black Pearl project-level financing, which fully funds our second data center campus all the way through completion, plus a $200 million revolving credit facility; this revolver is the first among our peer group, locking in multi-year committed liquidity from top-tier financial institutions, among them Morgan Stanley, Goldman Sachs, JPMorgan, Wells Fargo, Santander, and SMBC. Every one of these deals underscores how Cipher's platform keeps maturing. The Black Pearl financing is our third successful project-level bond issuance, proving repeat access to capital markets and an expanding, diversified institutional investor base that follows our story. We secured highly competitive pricing while preserving structural flexibility via a callable format, which we think sets us up to actively manage and optimize our capital structure going forward. The revolving credit facility, backed by a wide syndicate of leading global banks, further highlights growing confidence among institutional lenders that Cipher is a scaled, creditworthy counterparty, and it gives us the liquidity foundation for continued growth. Moving to Slide 18. I'd like to take investors through our entire capital structure and liquidity position as of March 31, 2026. At the corporate level, we hold a 4-year committed revolver worth $200 million along with 2 unsecured convertible notes. The revolving facility carries interest at SOFR plus 125 to 175 basis points, tied to the company's total debt-to-market capitalization ratio. That facility sat undrawn when the quarter ended, yet it affords us flexibility to support working capital, issue LCs, and fund growth initiatives. On the project side, we keep pursuing nonrecourse financing through construction, which reflects our disciplined approach to capital allocation. Cipher Compute LLC, the entity holding the Barber Lake lease, carries roughly $1.7 billion of 7.125 senior secured notes maturing November 2030. Those notes amortize, matching debt service to the cash flows the lease generates. Black Pearl Compute LLC carries $2 billion of 6.125 senior secured notes maturing February 2031. Both bonds currently trade at a premium to par, which reflects investor confidence in our ability to execute on both projects. Altogether, total principal outstanding on our debt came to approximately $5.2 billion. Unrestricted cash and cash equivalents total $715 million, giving us substantial corporate liquidity on top of Bitcoin amounting to $76 million and our undrawn revolver availability. We had roughly $3.5 billion in restricted cash. That breaks down as about $1.8 billion sitting at Cipher Compute, which is around $1.5 billion once you exclude DSRA and interest-bearing construction accounts, plus about $1.8 billion at Black Pearl Compute, which comes to roughly $1.5 billion after netting out DSRA and interest-earning construction accounts. On the basis of what we currently estimate it will take to finish, both projects have enough capital to make it all the way through construction. We built this capital structure for a specific purpose, and our liquidity is adequate to carry our near-term development pipeline forward with no need to raise more equity, which gives us clear line of sight into execution. One thing worth stressing: most of our debt carries no recourse and is attached to contracted assets, which walls off risk during construction and keeps debt matched to cash flow. Now let's move on to walking through our financial results for the first quarter of 2026. We brought in $35 million of revenue in Q1, a drop from $60 million in Q4, which mirrors the deliberate retirement of mining operations at Black Pearl and the shift we're making toward contracted data center revenue. Mining at Black Pearl was completely decommissioned during February. On the quarter, our GAAP net loss came to $114 million, or $0.28 per diluted share, versus a GAAP net loss of $734 million, or $1.85 per diluted share, in the prior quarter. Several things drove the Q1 loss: lower revenue stemming from the deliberate retirement of mining operations at Black Pearl, a decline in the fair value of our PPA, and higher interest expense tied to our new debt facilities. As for the Q4 loss, noncash and onetime items were the main drivers, among them the embedded derivative revaluation on the 2031 convertible notes along with write-downs of mining assets. Cost of revenue in Q1 totaled $18 million, falling from $24 million in Q4, which mirrors the move to Odessa as the only site we now operate. Compensation and benefits came in at $35 million for the quarter, essentially flat versus the prior quarter. Looking year-over-year, compensation climbed from $14 million, largely because of headcount growth and equity-based compensation tied to scaling the platform; our headcount went from 50 people in Q4 to 70 in Q1, and we're beginning to normalize and ease off hiring at roughly 85 full-time employees. General and administrative expenses reached $12 million, an increase from $10 million in the prior quarter, mostly because of higher legal and professional fees related to our lease negotiations and financing transactions. Depreciation and amortization fell to $19 million from $52 million in the prior quarter, mainly as a result of mining asset sales and decommissioning tied to the Black Pearl retrofit. The change in fair value of our power purchase agreement was a $28 million decrease this quarter, versus a $12 million decrease in the prior quarter. As we've said all along, this item is noncash. What makes the Luminant contract valuable is its long-term fixed price power, which underpins industry-leading power costs of roughly $0.028 per kilowatt hour — among the lowest anywhere in the industry. Let's go further down to the operating line. Interest income came to $32 million for the quarter, rising from $19 million in the prior quarter, which reflects higher average cash balances after the Black Pearl financing. Interest expense came in at $59 million, which is up from $33 million during the prior quarter, and that reflects the financings we've done at the project level. On our warrant liability, the change in fair value produced a $44 million noncash gain in this quarter, versus a $13 million loss in the quarter before, and that mirrors movements in the value of the Google warrants tied to the Barber Lake lease. Now let's move to the balance sheet, dated March 31, 2026. Total assets reached $6.4 billion by the end of the quarter, climbing from $4.3 billion in the previous quarter, mostly because of the Black Pearl project financing, which shows up in growth across both property and equipment and restricted cash. Cash and cash equivalents stood at $715 million. Restricted cash that's ring-fenced at the project entities and earmarked for construction spending came to roughly $3.5 billion when you combine current and noncurrent portions, and that figure includes proceeds from the Black Pearl financing. Property and equipment, net of depreciation, rose to $1.3 billion, up from $633 million at year-end, driven by construction that's underway across several projects. Looking at liabilities, borrowings came to about $4.7 billion. Accounts payable increased to $198 million by quarter end, compared with $40 million at year-end, which lines up with the acceleration in construction activity and the usual cadence of vendor payments. The balance sheet shows a company in an active investment phase, putting capital to work across numerous large-scale construction projects, with contracted revenues that ramp as those assets come online. We're carrying out our plan as intended, and from both a liquidity and capital allocation standpoint, we're still well positioned to meet our commitments and grow the platform. Before we take questions, I'd like to spend a moment on the broader picture of where Cipher Digital stands now that the first quarter of 2026 is behind us. We've signed 3 long-term data center campus leases that will generate roughly $11.4 billion in contracted revenue across the base lease terms. We have 2 data centers actively under construction, and both are tracking well; there's a third site where mobilization of construction starts in Q2; our balance sheet is strong, and we've shown a repeatable capacity to finance construction projects at competitive terms; we've kept flexibility in our financings, and our capital structure will keep evolving as time goes on. And lastly, we hold approximately 3.3 gigawatts of additional capacity in our pipeline, which sets us up for continued growth well into the back half of this decade. We stay firmly committed to disciplined execution, capital efficiency, and delivering long-term value to our shareholders. Construction milestones, revenue commencement, and the ongoing conversion of our pipeline into contracted assets will define the next 12 months; we look forward to bringing you updates on all of those fronts as the year progresses. Thank you for your continued support. Tyler and I would be pleased to take your questions.
Great to hear about the headway you've made across the sites. My first question is about pricing. Running a quick rough calculation, it looks like the added Stingray lease produces NOI per megawatt that beats the average of the other 2. So I wanted to check, as you hold these added discussions with possible tenants, where is pricing headed? Does it keep moving favorably compared with the deals you already have? Then, secondly, I'd like to ask, broadly speaking, with these leases and negotiations being so strong, what's your view on compute? Is that a type of business you'd consider at all? Has your perspective shifted on it? Or is colo leasing simply so robust that you're staying with it for the time being?
Thank you, Paul. We'll begin with pricing. I believe a single universal response is tough when it comes to leasing, since it's a fluid matter tied closely to how quickly you can reach the market and how quickly space becomes available. When a site is already energized, or will be energized very soon, it clearly commands a premium in terms of the lease rate it can achieve. And when you keep showing you can construct at a faster pace, which we're able to do, those schedules become more believable and your leverage on price grows. So yes, it's accurate to say premium pricing persists in our negotiations. That said, it's also driven by having sites ready in the near term. We still have 2 sites being marketed right now, Reveille and Ulysses, which I'd call near-term availability, and both draw a good level of interest; I anticipate our pricing leverage will hold there. From the discussions taking place, I don't foresee lease rates declining for premium sites that offer solid timelines. Then, regarding the compute matter, it's intriguing. Historically we've stated that if you can secure those premium colocation lease rates, that's simply a superior business compared with owning GPUs, TPUs, or whichever chips you operate. I'd say Reveille gives us an intriguing test case. Even though Reveille remains a large data center at 70 megawatts, perhaps under what the huge colocation tenants are targeting, we're exploring multiple business models there, among them ones where we might take an ownership stake in the computers. Some fascinating shifts are happening lately around credit support for Neoclouds. A range of bigger, more creditworthy backers of Neoclouds are offering credit support to help secure their success. So at Reveille, because of its size, we're examining it as an appealing location for Neoclouds. These Neoclouds are now able to secure investment-grade backing or similar arrangements, be it through a guarantee or via prepayment, and so on. As a result, we're looking at possible structures in which we'd take part in both owning and running the compute located at that facility. From a risk-adjusted standpoint, I believe the returns in that scenario might prove quite attractive, since we'd be able to obtain some support by engaging on that side of the operation. And once you reach that scale, the returns can turn out to be rather compelling; so I'd call that our test kitchen site. We've also received inquiries regarding Ulysses along those same lines. My sense is that our appetite would lean more toward the scale of Reveille, were we to step into the compute portion of the business. Broadly speaking, though, we prefer — whenever you're able to command very elevated lease rates, we prefer the colocation business.
That's quite fascinating. In your view, is that choice mostly pushed by the counterparty on the other side, or does it come from your own business model?
Honestly, it's somewhat of a combination. From our conversations with Neocloud, the general idea is that they're trying to leverage whatever credit backing they can secure, because building out these data centers requires substantial spending as they grow; on top of that, they don't rank among the very largest Neoclouds—the ones with access to broader capital markets, the ones waiting in the wings, ready to scale up or eventually go public or whatever; they may operate within the orbit of a company like NVIDIA or AMD or someone similar. And based on what we hear in those discussions, they've got plenty of compute offtake lined up and ready. Take their situation: suppose they already hold five-year compute offtake agreements, provided a suitable location can be secured inside a favorable timeframe. From the standpoint of what we talk about with them, it ends up being a combination of many elements when we evaluate it, since the questions we raise include which forms of credit backing would make a smaller Neocloud appealing as a tenant. Clearly, having an investment-grade backer stand behind the whole obligation is the gold standard, so to speak. On prepaying fees, if you spread that over time, anecdotally we have observed and been told about some bigger prepayments arriving, and those cut the project's risk substantially. After all that, it genuinely boils down to the numbers. So if there's sufficient prepayment or the credit support is robust enough, that could sway whether we're willing to take part on the compute side. At a high level, that business hasn't been particularly appealing, because even though a compute offtake agreement might return the cost of the computers within 5 years, there's usually a debt overhang sitting on the data center once those 5 years pass. That leaves you bearing what amounts to extended life risk — how much useful life do today's machines actually have? Lately, obviously, those figures have looked quite favorable. But our past experience with ASICs teaches us that this doesn't necessarily stay consistent indefinitely. The next thing I'd ask is, what level of debt are you willing to carry if, five years from now, your compute tenant's lease wraps up and the data center still isn't fully paid off? That's exactly why I keep saying the risk-adjusted returns in colo are incredibly strong — with better lease rates and lengthier agreements, you end up owning a fully paid-off asset at very appealing returns. And honestly, this connects to our bigger thesis around locations such as West Texas, where those sites hold enormous terminal value that the market hasn't really priced in yet. So it's a complicated blend. To put it in a roundabout fashion, our aim is delivering the strongest returns for shareholders once risk is accounted for. As the credit support situation in that area has changed, taking part in ownership of the machines at a location such as Reveille has grown more appealing to us.
Congratulations on everything you've accomplished — it's genuinely impressive to watch. Tyler, do you have anything new to share regarding the Odessa PPA? The market is clearly very robust right now, and I'm certain this topic is front and center for everyone. I also have a brief follow-up question.
Well, here's the thing, Joe — having the lowest electricity cost in Bitcoin mining is a wonderful position to be in, since we earn solid margins daily from mining Bitcoin, and that rate stays fixed at an extremely low level for roughly the coming 14 months. There's considerable interest in Odessa from our side. A hyperscaler has shown interest in that location. As I believe I've said previously, we'd be open to potentially transforming that site the way we transformed Black Pearl, taking it from a Bitcoin mining facility to an HPC campus. Multiple counterparties will be involved in that process since we obviously need to rework that PPA, and we'll be collaborating with the counterparty involved there. So that PPA remains among our greatest strengths, one we secured quite a while back. And it definitely provides us with substantial leverage when we consider what lies ahead for that site. That said, the future of that location might simply be continuing to mine Bitcoin there for another 14 months and generating significant profits from Bitcoin mining at that spot. There's no need for us to hurry, given how advantageously we're positioned thanks to that low price.
Absolutely, that's clear. And keeping with the behind-the-meter subject, Tyler, on earlier calls you brought up looking into behind-the-meter possibilities. Obviously you hold a sizable power portfolio, but could you update us on your approach and your thinking regarding behind-the-meter chances?
Right. This is an area where several of our strongest team members have been devoting the bulk of their hours lately. Given where our sites sit, I believe the opportunity for on-site generation behind the meter is remarkable for both us and those locations. At our West Texas sites, we can tap into an enormous supply of inexpensive natural gas. Assembling every element required for a data center powered by successful behind-the-meter generation is a difficult task. Because of how the load profile works for an HPC data center, you have to work through certain engineering problems. Your own generation won't deliver the same traits and steadiness that grid-connected power does, so those engineering hurdles must be addressed. You also have to handle the gas infrastructure side. Usually independent power producers will be part of this, meaning you must select a generation source, possibly secure air permits, arrange financing for the entire project, and back power purchase agreements worth billions of dollars; it's therefore quite complex. The encouraging part, I'd say, is that we're in the kitchen alongside the finest crew, and every ingredient needed for a Michelin 3-star dish is at our disposal. The trick is that assembling all of those components has essentially never been accomplished. Elon has managed it, though I'd place him in a separate category of his own, one with something like authoritarian control across the whole ecosystem, while the remainder of us ordinary people must work with actual humans drawn from these various, disconnected corners of the ecosystem; so it amounts to a test of engineering, a test of financing, a test of emotional intelligence, and more. I believe our position for seizing that chance is exceptionally strong. Honestly, it might represent the greatest upside convexity potential in our stock, since, to repeat, in theory there's gas capable of fueling gigawatts' worth of generation that we don't even mention in our presentations, simply because it's still early. Even so, the possibility is thrilling enough, and tenants are curious enough, that we're investing substantial time in it. So keep watching.
Congrats on landing that new agreement. Is there anything you can share about possible ways to grow beyond that first 100 megawatts?
Right now it's simply 100 at the outset. That said, like I said earlier, every one of these sites sits on top of what amounts to a whole sea of natural gas; we hold a great deal of acreage, and our positions are advantageous; so keep watching to find out whether that behind-the-meter narrative keeps developing. However, under the terms of the deal, no, the 100 megawatts is what's covered.
All right, that's great. And when you look ahead across the coming several years, is there a moment you envision where you'd be able to leave Bitcoin mining behind completely? And how do you think the company shapes up, say, by 2030? You took some time to discuss Cipher's long-range objectives.
Yeah. Well — I'll tell you, that's a fantastic question, and it's the kind of thing that keeps me awake at night, simply because the upside gets me so fired up. So to begin with, I do expect Bitcoin to wind down. Like I said earlier, there's no intention on our part to put additional CapEx toward Bitcoin mining; what we have is a site that's operational and could keep running through the end of July 2027, and the economics on it are genuinely attractive. So turning it off isn't something we're hurrying to do. Each month it generates positive cash flow — a few million dollars. So having that is just fine with us. With that said, that's more or less a hard stop for the site, and either we might find another use for it, or we could let it wind down over the coming 14 months and say our goodbyes. In all likelihood, our Bitcoin position will also have been liquidated, I'd guess, sometime this year. There's no hurry on our end; we keep managing that inventory downward bit by bit. At low prices we haven't been forceful sellers, because honestly, our view on Bitcoin at this point is fairly bullish. We picked up a decent amount of premium by writing some calls above Bitcoin's market price at various points during the first quarter, since we'd gladly sell at higher levels, and if that doesn't happen, we keep the premium; so we're handling that decline carefully. My sense is Bitcoin won't figure into our story anymore by 2030, as you put it — I'd say by 2020 — by the close of '27 at the latest, possibly earlier. And it's already kind of tapering off when you consider the NOI that the leases will be contributing; before it's fully wound down, it will have become immaterial to our financials. Now, come 2030 — well, this is the part where the upside really gets you excited; we're eagerly waiting on what comes out of ERCOT's batch process. I believe we've tried hard to make it plain that we see ourselves as very strongly positioned, given our sites, to land in batch 0 there. And once more, my take on the HPC business is that it functions somewhat like a flywheel: you sign leases, you show excellence in executing on the quality of what we're constructing, the quality of how we manage relationships, the timeline we deliver against, the capacity to finance at the best rates available in the sector. This morning I looked at our debt for the 2 projects, and the yields came in around 6.2% and 5.7%, with both trading above par; so it's evident that debt investors around the world have faith in our capacity to build and deliver on schedule. That favorable momentum, along with the ability to secure capital, opens the door to building additional large data centers. Our pipeline is in place, and those connections with tenants keep growing stronger. In my view, none of the rivals we face can match the tenant relationships we've built. So when 2030 arrives, to address what you asked, my expectation is that we will hold premium long-term leases with the finest tenants anywhere across every location in our portfolio. Beyond that, I'd like to see us running more than 4 gigawatts of HPC at colocation rates that are genuinely appealing. There's a great deal to accomplish before that point; it's an ambitious goal, yet I have strong confidence this team will reach it.
Nice work on the momentum. Could you talk about which pieces of equipment are still pending for Black Pearl's Phase I and Phase II?
Absolutely. Regarding the majority of the equipment still pending, it hasn't been bought yet. Everything is basically following the procurement schedule we presented at the beginning, and these are generally commodity-type things; so as I believe we mentioned during the presentation, 93% of the Phase I equipment is secured and 80% of the Phase II equipment is secured. The items still pending aren't the long lead time items, which are usually things like, again, cable, office furniture. I think there's some miscellaneous mechanical equipment, and so on.
Excellent, that's really useful, Tyler. Now, one more thing. Given that your signed backlog keeps growing, have you shifted your thinking about how early ahead of a site's anticipated energization you'd be willing to execute a new lease?
Yeah, that's a solid question. In my view, securing the interconnect is a key component of this, so we're monitoring the ERCOT batch process very closely. Should things land where we anticipate they will land, with interconnects secured at a couple of sizable sites that we're targeting for energization in 2028, my expectation is that we'd move into lease discussions fairly promptly. Honestly, a gigawatt-scale site is simply that appealing — plus you require sufficient time to construct it; that, I believe, would fit comfortably inside the timeframe needed to kick off accelerated lease construction. That said, I continue to believe that holding the interconnect has generally served as a validation point that tenants must witness, since it seems like there are roughly 1,000 individuals who have suddenly become data center developers and possess a site in some location; and once you actually scrutinize how advanced and dependable the grid connection truly is, it frequently collapses. We observe this when reviewing corporate development possibilities, when evaluating site purchases — it's reasonable to say that to endure the demands of negotiating a lease with a hyperscaler, you'll need a truly airtight schedule for your interconnect. So my sense is that element is likely the primary driver. Beyond that, consider construction schedules. From our standpoint, one massive edge we possess is our capacity to deliver fast. Our configuration here differs somewhat from what many others have. Specifically, we manage procurement internally. We've assembled a group of specialists, former hyperscaler employees or the construction companies that contract with them. And I'd argue that ranks among our greatest edges — overseeing the supply chain, operating swiftly, and dictating our own future. That's likely why we sit in such a favorable spot across our sites. Honestly, at this juncture, I'd say the odds favor us completing a site early rather than late. Such is the caliber of our schedule oversight; so with any luck, we'll see a few positive surprises in the not-too-distant future. Regardless, I believe overseeing all of that stands as one of our strongest capabilities. And I can see it feeding into the flywheel, just as I outlined previously.
So earlier you brought up the interest you've received about possibly owning your own compute at Ohio at Ulysses. I suppose, just pulling back to the big picture, I'm curious about the wider demand you're seeing for that site. And is there anything worth highlighting since it sits in a different power market compared with the rest of the portfolio?
Yes, sure. I mean, I think interest has come in, but we've also had interest plus active conversations with several hyperscalers regarding that site. So I think it's fair to say that if we're going to put a toe into owning compute, it would more probably occur at Reveille, simply because of the scale and the risk exposure involved there; that said, never say never. Should credit support keep developing among amazing counterparties for Neoclouds and that extends to owning the computers, and we're able to deliver a better return for shareholders, we'll pursue it; but since Ulysses also draws interest from truly great names purely for traditional colocation, that's — I mean, right now, that's probably the more probable path. Stay tuned. On the matter of it being a different power market, PJM is a market we've targeted for a long time. So having a site there is exciting for us. I went to the site 2 weeks ago; it looks great; we're very excited about the potential; so I think there's a high degree of interest in the site, a bit closer to a traditional location since it's in the Greater Columbus area, sort of extended, which is a popular traditional data center market; so I think demand in PJM is also pretty big, because you asked about the nature of the markets. Broadly speaking, deposit requirements there are higher than ERCOT. So I'd say it's a little more demanding. So it's sort of harder, I'd say, to secure an attractive site in PJM; so that's another factor that places a bit of a premium on that site. There aren't as many opportunities.
That's great, really useful. Now turning to how this gets funded, I'm just wondering whether there's anything worth flagging regarding possible project-level financing tied to the new 100-megawatt contract. And I suppose, is there anything about project-level financing for colocation contracts that you'd want to flag?
Greg, would you like to take that one?
Absolutely, glad to chime in. Thanks for asking. So, at this point we've completed 3 project-level bond issuances successfully. I believe we've clearly shown we can tap the capital markets. There's a varied group of institutional investors who now track our name. If we decided to issue another bond comparable to the one we did today, I have strong confidence it would go well and we'd see pricing on similar — maybe even improved — terms; that sits fine with us, and we're pleased with our financing approach so far. Given where we are in our growth trajectory, I want to stress how essential it is for us right now to keep our capital structure flexible. How this entire sector develops going forward is still uncertain. So while noncallable long-duration financings may look impressive on a headline basis, they can, I think, lock up cash and constrain your capacity to optimize assets down the road. What we're really centered on, then, is keeping the capital structure flexible in both directions. We're centered on nonrecourse financing so that construction risk stays isolated throughout construction. And we're centered on refining the capital structure as we expand, adding more assets into the portfolio and more stabilized assets that serve as collateral. So whenever we examine financing options, we'll lay everything out, assess it, and choose whichever option is best. But we're confident we'll be going after something with terms similar to or better than what we've encountered so far.
Based on how frequently bankers have been reaching out to Greg, all of them keen to arrange additional debt financing on our behalf, I have no doubt that demand for our paper in the market is exceptionally strong.
Look, my assumption is that Reveille and Ulysses are essentially locked in at this point. What I really want to dig into is McLennan, Mikeska, and Colchis. Tyler, what stands in the way — what are the biggest obstacles — to getting power flowing at those 3? And what's the demand picture right now for that 2028 power? How are those talks progressing?
Absolutely. Appreciate it, Mike. The way I'd frame the challenge is this: ERCOT is working through a batch process, and that means nonstop meetings, with everyone weighing final adjustments to the implementation approach. Just yesterday there was a session lasting several hours, and we were part of it. So here's where we stand — at those 3 sites, and this is exactly why we call attention to them, we've completed every single item that's been asked of us so those sites can qualify under what will come out of the batch 0 approval as it makes its way through ERCOT, and that ought to wrap up next month. With that in mind, our view is — and keep in mind, this whole thing keeps shifting, correct? ERCOT has revised its requirements, and the batch process has slipped as things have moved forward; so we're making every effort to be fully transparent; our belief is that we've checked every box, but we won't have the interconnect until we get that final green light. Like I said before, securing that interconnect is likely the key that unlocks moving tenant negotiations ahead quickly. So once more, we've got a strong team tracking this and taking part, and I think we're hearing things on the ground quite effectively regarding how the batch process keeps evolving. Our confidence is high that these 3 sites have completed what's required, and that's been true for some time now. So if you think about it in terms of sequence, they're sitting in a solid position. We're anticipating a favorable result once that batch process is finalized in June. And from there, I think we move to the following question, which is: alright, so with roughly 1.5 years to go, heading toward 2028, how much interest is out there. In the past — and when I say the past, I mean roughly the last year, once you enter that window of about 1.5 years out, interest from tenants has peaked at its highest. That's when it lands right in their sweet spot, where they're able to align demand forecasts with their needs. I'm fully convinced there's going to be substantial demand. And when we're talking about sites with a gigawatt on-site or 500 megawatts, those are genuinely massive, appealing sites. They're located in Texas, which welcomes data centers. We deal with serious NIMBY pushback around the country. Here in Texas, at the sites we already operate, we're handling that extremely well. Texas is positioned to deliver a positive result on those types of concerns. Accordingly, my optimism regarding demand for those locations is extremely strong, once the last piece of uncertainty in the ERCOT process is resolved; and after that, I hope there will be plenty of positive developments to share over the next several months.
Understood. Yeah, we're only about a month out from that. Then, if I could ask one more thing: when you look at your pipeline slide around the end of 2026, how would you have us weigh the chances that newly acquired sites show up there? Low, medium, or high?
Honestly, that's tough to call. What I can tell you is that a fair number of inbound opportunities have come our way, I believe because our name recognition has grown, people are doing channel checks on our tenants, they see how our construction projects have gone, and they see our financings. Even so, the bulk of what's coming to us is very far off in the future, and those haven't really fit what we need. Like an opportunity in 2030, that kind of thing. What I think could get interesting is this: as the ERCOT process moves forward and gets finalized, a major piece of that involves putting down sizable deposits. In the past, many of our sites came from speculators with less capital, who spot a good location but aren't equipped to actually build it out or to put down double-digit millions of dollars in deposits as proof they're serious, which is one of the tests under ERCOT. So your question doesn't have an easy answer, but I'd say I'm cautiously optimistic that once this process is finalized, it might create opportunities where people we know and track in that ecosystem of smaller developers reach out to us, possibly seeking a partner. An exact forecast isn't something I can give. The only thing I can tell you is that we're constantly evaluating site opportunities. There may be a wave of them headed our way in a region where we've had plenty of success historically. So that's another one to stay tuned on. That said, building out the 4-plus gigawatts we already have will certainly keep us occupied in the meantime.
Once more, I want to express my appreciation to all of you for joining the call and for the support you keep giving us. There's a lot happening that has us genuinely enthusiastic, and we can't wait to speak with you once more before long. Cheers.
Thanks. That wraps up our presentation for today, everybody. You’re welcome to leave the call now, and I hope the rest of your day goes great.
Read the full transcript via the API
You're viewing the first half of this call. Get the complete Cipher Digital Inc. transcript - plus 132,000+ transcripts from 7,000+ companies, speaker segments and full-text search - through the EarningsAPI REST API or hosted MCP server.
Get an API key View API docs →For developers and AI pipelines
Programmatic access to Cipher Digital Inc. earnings transcripts and 132,000+ others is available through the
EarningsAPI REST API and the hosted MCP server.
Quarterly plans from $162 - full transcripts, speaker segments, full-text search,
and the /api/v1/transcripts/recent polling endpoint for ETL pipelines.