Good morning, everyone, and welcome to the Galaxy Digital third quarter 2026 earnings call. Today's call is being recorded. After today's presentation, there will be an opportunity to ask questions. To ask a question, you may press star and then one on your telephone keypads. To withdraw your questions, you may press star and two. At this time, I'd like to turn the floor over to Taylor Reinhardt, Head of Investor Relations. Please go ahead.
Good morning, and welcome to Galaxy's third quarter 2026 earnings call. Before we begin, please note that our remarks, including answers to your questions, may contain forward-looking statements. Actual results may differ materially from those described in these statements as a result of various factors, including those identified in the disclaimers in our earnings release or other filings, which have been filed with the U.S. Securities and Exchange Commission and on SEDAR+.
Forward-looking statements speak only as of today and will not be updated. Additionally, we may discuss references to non-GAAP metrics, the reconciliations of which can also be found in our earnings release. Finally, the information on this call does not constitute a recommendation, solicitation, or offer by Galaxy or its affiliates to buy or sell any securities. With that, I'll now turn the call over to Mike Novogratz, Founder and CEO of Galaxy, to discuss our results.
Good morning, everyone, and welcome, Taylor, to the Galaxy team. You guys will get to know her well and will like her as much as we do. Listen, we had an awesome quarter. I'm really proud of what the team delivered. But what excites me more is what performance says about where Galaxy is headed. We've spent years investing ahead of two enormous shifts, finance moving on chain, and the build-out of AI infrastructure, and we're increasingly seeing those investments translate into results with the strategy playing out in three important ways this quarter.
First, we saw a strong rebound in crypto markets. Bitcoin was up more than 40% with a meaningful improvement in sentiment and client activity. Our global markets business capitalized on that momentum, growing trading volumes and expanding client engagement.
We also saw a significant appreciation across our investment portfolio, reflecting the value of capital we've deployed across crypto and the broader technology ecosystem. Second, the opportunity in digital assets is getting bigger, and I think we're just scratching the surface of what's ahead. Crypto remains an important and growing asset class, but the opportunity extends well beyond that. With digital assets increasingly becoming the foundation for the future of finance. Increasingly, those two worlds are coming together.
You can see it in the growing adoption of stablecoins, the tokenization of traditional financial assets, and the participation of major financial institutions. I have a tremendous amount of conviction that the financial architecture developed by the crypto industry will underpin much of how the financial system operates in the future.
Galaxy has spent years investing in these capabilities, and while we are still in early stages, I believe that investment will translate into meaningful results over the next several years. Regulatory clarity will be an important part of that. While the CLARITY Act has not become law, unfortunately, the SEC and the CFTC are working towards clearer frameworks. There is still a lot to sort out, but the direction is encouraging. Investing ahead of these shifts has always been part of our strategy.
We did it early in crypto, investing in companies and technologies that help build the ecosystem. Now we are taking that same approach to AI, increasingly investing in the next generation of companies through our investment strategies in our balance sheet. Third is data centers, where we are building a very sizable business. Helios has come a long way in the past year.
We are now generating approximately $80 million of leasing revenue per quarter from phase I . As phase II comes online next year, we expect to exit 2027 at an annualized leasing revenue run rate of between $800 million and $850 million, based on our current delivery schedules. That is a substantial business from the capacity we have already leased, and the momentum is building.
We have significant additional approved power available to lease and a deep development pipeline beyond that. I want to address the timing of leasing announcement at Helios II. Earlier this year, I spoke about the possibility of reaching agreement during the summer. We were in advanced discussions around a potential transaction, but as the market dynamics shifted material in late summer, we ultimately decided not to move forward on the terms being considered.
The market for large scale power via data centers has become increasingly constrained, while demand for capacity beginning in 2028 and beyond has strengthened. With more than 800 MW of secured power and supporting infrastructure, we believe Helios II is an even more valuable asset today than it was just a few months ago. With delivery plan for 2028, we have the flexibility to evaluate our opportunities and get the right deal.
We have built Galaxy by adapting to changing markets and being disciplined about where we commit capital. These are large, long-term agreements, and we are not going to compromise on terms just to meet an announcement date. We continue to have productive and active discussions with prospective customers with a focus on maximizing long-term value for shareholders. As the Helios campus scales, we are fundamentally changing the earnings profile of Galaxy.
We have established a digital asset franchise with significant earnings power, growing opportunities for recurring fee-based revenue, and a data center business is adding long-term contracted cash flows. Together, these businesses give us multiple drivers of growth and more diversified earnings base that we believe will make Galaxy a stronger, more durable franchise over time. We also have the balance sheet to back that ambition.
It is a competitive advantage that allows us to commit capital as a principal, invest in growth across our businesses, and take advantage of opportunities as they arise. We are focused on putting that capital to work where we see the most attractive long-term returns. That includes investing in our own shares when we believe they are trading at a meaningful discount. We have repurchased $200 million year to date, and this morning announced an additional $200 million share repurchase authorization.
As significant shareholders ourselves, we see buybacks as another potential way to allocate capital thoughtfully to create long-term value. Chris will now take you through how we are translating these opportunities into results across our businesses and what we are focused on next. Chris, over to you.
Thanks, Mike. Starting with digital assets, one of the strengths of our global markets business is the breadth of solutions we can offer clients. We are seeing growing demand for more sophisticated trading and financing solutions, particularly as traditional financial institutions look to participate in decentralized markets through products and structures they are familiar with. We have spent years building capabilities for crypto-native clients, and that experience gives us an advantage as traditional finance increasingly moves on-chain.
We understand these markets, can move quickly, and have the flexibility to develop solutions for both groups. Lending is a great example. We have expanded the Galaxy Onchain Financing Rate, or GOFR, which gives clients access to on-chain financing, alongside our growth financing facility, which allows us to support more lending activity through external capital. We are expanding our loan syndication, collar lending, and structured financing capabilities, broadening the range of clients and financing needs we can serve.
We are seeing the same innovation across trading. Our derivatives business executes sophisticated options and hedging strategies for clients. This quarter, we completed our first Compute Link derivatives transaction, extending our capabilities into an emerging market that we believe has the potential to grow into a sizable new standalone asset class. In asset management, we are expanding our investment strategies and product offerings to meet evolving institutional demand.
During the quarter, we launched the Galaxy Sharplink Onchain Yield Fund, an actively managed strategy providing access to yield opportunities in decentralized finance through an institutionally managed SMA for a public company. Sticking with on-chain asset management theme, SWEEP, the first-ever decomposable 24/7 on-chain money market fund in partnership with State Street Global Advisors, is now approaching $300 million of AUM, which we believe is an early signal of strong product market fit among first-mover institutions interacting with tokenized real-world assets.
With Galaxy Digital Ventures, we are progressing towards a first close of Fund 2, expected towards the end of Q4, with a fund target size of $250 million - $300 million. As an investor, you only win the right to raise a second larger fund once you have proven out a strong process and great returns, and the GEV team has done that with Fund 1, currently marked at 2.25 gross MOIC.
In digital asset infrastructure, we are seeing growing momentum with some of the world's largest financial institutions and are expanding our capabilities and offerings across tokenization, custody, and staking to meet their evolving needs. Staking is an area where we are seeing growing engagement with major institutions. Assets under stake grew 64% in the quarter to $4.6 billion.
While this was primarily driven by price appreciation of the underlying assets in the quarter, we've also seen a number of large ETF providers begin delegating their ETH and SOL stake to Galaxy's infrastructure, driving over $400 million of gross new inflows in Q3 and subsequent to quarter end, and we expect additional sizable partner launches before year end. More broadly, our relationship with BNY announced this summer has been an important validation of our capabilities as we continue to deepen our engagement with leading global financial institutions.
These are complex engagements with long sales cycles, but the potential is significant. Our objective is to become a long-term strategic partner as institutions build out their digital asset businesses, creating opportunities to expand relationships beyond an initial mandate and generate meaningful recurring revenue over time. Turning to our data center business.
We're continuing to deliver against our development plans while advancing a substantial pipeline of future opportunities. This was our first full quarter of leasing revenue from Helios, marking an important milestone as our development investments and long-term contracted revenues begin converting into realized contracted cash flows. Phase II, which will deliver an additional 260 MW of critical IT capacity, is fully funded and under construction, putting us in a strong position as financing for new data center developments become increasingly constrained.
Both buildings and the supporting substation are progressing on schedule and within budget, building on our successful delivery of phase I. Structural steel installation is well underway on the first building, construction is advancing on the second, and key electrical infrastructure is being installed. We expect construction activity to ramp significantly over the coming months, with data hall deliveries scheduled to begin in the second quarter of 2027.
On permitting, we've secured the key environmental approvals required for ongoing construction at Helios. While certain additional approvals remain outstanding following Governor Abbott's September directive, we do not currently expect an impact to our construction progress or our anticipated energization timelines. Turning to Helios II. To build on Mike's earlier comments, we saw several important shifts in market dynamics beginning in late summer that we believe have meaningfully strengthened our competitive position.
Forward power availability has become increasingly constrained, with new power approvals more difficult to secure and delays in competing developments limiting the amount of large-scale compute capacity that can realistically be delivered between 2028 and 2030. At the same time, we're seeing a meaningful increase in demand from hyperscalers, frontier labs, and neoclouds now looking to secure capacity for 2028 and beyond to support existing backlogs and future growth.
Underlying demand for compute remains as strong as we've ever seen it. Financing markets have also become more selective, and we believe access to capital will increasingly favor developers with proven execution capabilities, high-quality counterparties, and financing structures that appropriately allocate risk. Having successfully delivered Helios phase I, with phase II fully funded and progressing on schedule, we are building a track record that we believe will differentiate Galaxy in this environment.
Meanwhile, the evolving regulatory and political landscape in Texas has created additional uncertainty around future data center development. Based on our current understanding of the applicable requirements, we do not believe Helios II is subject to the same uncertainties facing certain other projects, further strengthening its relative attractiveness. Taken together, these developments have changed how we evaluate the opportunity at Helios II.
We're taking the time to engage with the right counterparties and evaluate structures that we believe can maximize long-term shareholder value for Galaxy. The focus of our active conversations is squarely on investment-grade hyperscalers or counterparties supported by investment-grade credit enhancements, considering not only headline economics and counterparty quality, but also critical lease terms, probability of execution, and future growth potential outside of the Helios campus. Beyond Helios II, we've made meaningful progress through ERCOT's Batch Zero process.
As announced in September, five of our Texas data center projects received conditional Batch Zero classifications, representing approximately 4.2 GW of capacity. That includes more than 1.6 GW of previously approved base load capacity at Helios I and Helios II, which remains largely unaffected by the ongoing ERCOT process. Our three expansion projects, Caspian, Selene, and Helios III, received studied load classifications, representing an additional 2.6 GW of potential capacity.
While we submitted Caspian for base load classification, ERCOT ultimately reclassified it as studied load, and our subsequent dispute for reconsideration was denied. Although that wasn't the outcome we had sought, Caspian remains part of the Batch Zero study, and we expect greater clarity on power allocation and anticipated energization timelines for these projects in April 2027, based on the timelines communicated to us by ERCOT.
These classifications cover a substantial portion of our broader development pipeline of more than 5.7 GW, providing a foundation for potential growth well beyond the initial phases of Helios. Beyond our data center business, we're also investing in the growth of AI across the firm, with a focus on opportunities ranging from frontier model developers to emerging technologies at the intersection of AI and financial infrastructure.
Through Galaxy Ventures and our broader investment activities, we're building on our experience gained from investing early in the digital asset ecosystem. To date, we've deployed more than $80 million across 11 portfolio companies via 17 investments, including SpaceX, Etched, Karta, Ornn, TRM Labs, and Architect. While these companies are all early in their growth journey, our current unrealized mark to market on that pool of investments is already 3.4x.
Importantly, our ambition extends beyond just generating returns for our balance sheet. We're using these investments to build a track record, establish relationships, and create commercial opportunities across the Galaxy platform. Across our businesses, we're focused on translating the investments we've made into tangible results and long-term earnings growth. I'll now turn the call over to Tony to discuss our financial results.
Thanks, Chris and Mike. For the third quarter, Galaxy reported net income of $323 million, adjusted gross profit of $569 million, and adjusted EBITDA of $455 million, marking a sharp improvement from second quarter performance. Our results reflected the continued earnings power of our digital assets franchise, the first full quarter of leasing revenue from Helios, and significant appreciation across our investment portfolio.
Firm-wide operating expenses, excluding grossed-up transaction costs, digital asset impairments, and pass-through data center operator costs, were $161 million, compared with $147 million in the prior quarter. This increase was primarily driven by higher depreciation and amortization expense related to Helios phase I entering full operation. Our operating businesses generated combined adjusted gross profit of approximately $150 million, up more than 75% from $86 million in the prior quarter, and adjusted EBITDA of $60 million, compared with approximately $1 million in Q2.
On a segment level, starting with digital assets, the business generated $71 million of adjusted gross profit, up from $66 million in the prior quarter, reflecting increased client activity and the breadth of our platform. Within global markets, adjusted gross profit increased to $53 million from $49 million in Q2. Our digital asset trading volumes grew 12%, outperforming the broader spot market, where volumes declined, demonstrating the strength of our trading franchise.
In lending, our average loan book grew over 25% to approximately $1.8 billion, reflecting continued demand for our financing solutions. Asset management and infrastructure solutions generated $19 million of adjusted gross profit, compared to $17 million in the second quarter. Assets under management increased by over 30% to $5.9 billion, driven by growth of more than 40% in ETFs and 30% in alternatives.
As Chris mentioned, assets under stake also increased more than 60% to approximately $4.6 billion, reflecting both market appreciation and new capital inflows. Fee growth was more modest during the quarter, as reported AUM reflects quarter-end balances, including capital raised late in the period, while fee revenue reflects activity throughout the course of the quarter.
Turning to data centers, we generated approximately $79 million of adjusted gross profit and $69 million of adjusted EBITDA, representing the first full quarter of contribution from Helios phase I. Capital expenditures were approximately $757 million during the quarter, compared with $448 million in Q2. We expect spending to increase as phase II construction progresses, with a significant portion of the remaining construction expenditures anticipated in the first half of 2027.
As a reminder, there are a few moving pieces related to interest income and expense in our data centers business that will affect our financials over the coming quarters. As financing proceeds are deployed into construction, our short-term investment balances and related interest income will decline. At the same time, a greater portion of our borrowing costs will be capitalized on the balance sheet rather than expensed through the income statement.
At quarter-end, data center segment had approximately $6.8 billion in total assets and approximately $5.3 billion in total liabilities, reflecting continued investment in Helios and the financing secured for phase II. We do expect this segment's total asset base to grow as construction progresses and capital is deployed.
The treasury and corporate segment generated $419 million of adjusted gross profit during the quarter, compared with a $42 million loss in Q2, driven primarily by significant appreciation across our digital asset holdings and private investments during the quarter, including notable contributions from Etched and Forward Industries, among others. As Mike mentioned, these investments reflect our ability to identify compelling opportunities across the digital asset ecosystem and broader technology ecosystem and represent an important source of long-term value for Galaxy.
We also held approximately $1.7 billion in net digital assets and investments at quarter-end, up nearly 50% from Q2, including exposure to Bitcoin and other liquid digital assets. Firm wide, we ended the quarter with approximately $4.9 billion of cash, stablecoins, and treasury bills, compared with $2.5 billion at the end of June.
This increase was driven largely by proceeds from the $3.5 billion financing completed during the quarter to fund construction of Helios phase II. Looking ahead, we anticipate pursuing phase III financing in the first half of 2027, with the structure of that financing being informed by market conditions, comparable transactions, and our assessment of the most attractive terms available. We also have $445 million of exchangeable notes maturing in December 2026.
We have the necessary resources on balance sheet to meet that maturity today, but are always evaluating different financing options in the context of our overall liquidity position and broader capital priorities. We exited the quarter with approximately $3 billion of book equity, with more than 80% allocated to our operating businesses, providing a strong capital foundation as we continue to scale.
On share repurchases, we deployed approximately $79 million during the third quarter, and as Mike mentioned, we purchased the remaining portion of our previously authorized $200 million during the month of October. These repurchases have more than offset dilution from equity-based compensation, contributing to a net reduction in shares outstanding year to date.
The board's approval of an additional $200 million authorization announced today allows us to continue opportunistically buying back stock when we believe it's trading at a meaningful discount to intrinsic value, and it fits within our broader capital plan. With that, operator, let's open the line for questions.
At this time, we'll begin the question and answer session. To ask a question, you may press star and then one on your touchtone phones. If you are using a speakerphone, we do ask that you please pick up your handset before pressing the keys to ensure the best sound quality. If at any time your question has been addressed and you would like to withdraw your question, you may press star and two. Again, that is star and then one to join the question queue. Our first question today comes from Patrick Moley from Piper Sandler. Please go ahead with your question.
Hey, good morning, everyone. This is Will Copps for Patrick. I want to go back to the comments on the Helios lease discussions from over the summer. Can you help us understand what led you to step away? In what terms you'd need to see the deal you're willing to sign? Was it driven mainly by economics and revenue or potentially more by counterparty and credit support or something else? Thank you.
Hey, Will, good morning. Thank you for joining, for the question. As we said in the remarks, we had been pursuing opportunities throughout the year at early stages, given that Helios II's energization timeline really isn't until the middle of 2028 at the earliest expected right now. Those conversations got pretty serious going into the summer, and then a lot of stuff happened.
We saw actual projects around us getting delayed. We saw regulatory headwinds, not just in Texas, but across the country, tightening up, and therefore an expectation that available power on the forward was going to be tighter than I think we or the market really understood. At the same time, we saw a number of financing structures start to come into the market. Think like investment grade wraps around non-investment grade potential counterparties that we thought were interesting, and that we were evaluating.
The issue there is the devils are in the details, and so an investment grade wrap that doesn't actually achieve investment grade protection for both our debt and our equity isn't really an investment grade wrap. You don't really know those things until you get into the details and you get to the finer points.
Finally, to sort of back that up, the financing markets, obviously, around that time also started backing up and the focus on structure counterparty really came into light for us and for the whole market. With all that as the backdrop, we decided to take a pause, take a breath, reignite conversations more broadly. That sort of happened at exactly the same time that a number of the big hyperscalers started turning their attention to 2028 available power.
That's what led us to decide to broaden out the process. That's why now we're in active conversations with a number of high-grade potential tenants who are now very focused on figuring out what their power strategy is for 2028.
Got it. Thanks a lot.
Our next question comes from James Yaro from Goldman Sachs. Please go ahead with your question.
Good morning, and thanks for taking the question. I was hoping you might be able to update us on your current capital deployment priorities, in light of the new buyback program, the maturing notes at the end of the year, as well as the continued need to deploy capital into the data center segment. Could you speak to your appetite, I guess, or maybe just expand on your appetite to potentially replace the notes maturing in December, with additional debt?
Yeah, sure, James. I'll take that one. First, when we think about capital planning, we start with what is the business need across our two businesses, which obviously have different capital need profiles. That is the first focus for capital requirements. We obviously have a funding and financing plan for the data centers. The digital asset businesses broadly tend to be self-financing in many ways as financial services oriented. We think of our business needs as the first priority.
Secondly, I would say we think about a capital buffer and cushion on the balance sheet. Maintaining a healthy, strong, robust balance sheet I think is a very important principle from our standpoint. As we talked about, we have significant cash, stablecoins, and liquidity on balance sheet in addition to liquid investments on the balance sheet that we have the ability to shift around as necessary.
As you'll see, we've got about 80%, actually more than 80% of our capital now deployed into our operating businesses. We've talked about this over the last couple of quarters. You should expect that number to increase, as we see the investment in those businesses grow out. Then third, we extinguished the buyback, the $200 million initial.
We thought it was appropriate to have an additional authorization. But we will be very, I would say, opportunistic in how we think about deploying additional capital from a buyback perspective. The second part of your question, remind me. What was the second part of your question?
If you could just maybe just expand a little bit on the comments around the exchangeable notes and I guess appetite to replace that. I think you pointed out you have enough resources to not have to do so, but just maybe the moving parts there.
Yeah. Got it. Thanks. Yeah, the exchangeable. We have $445 million of exchangeable notes maturing in December. Look, the way we think about that is, first and foremost, we have the capital resources to pay that back in full, without refinancing it. As you know, we did a convertible issuance Q4 of last year. Those resources were deployed across our operating businesses and to sort of pre-fund some of this. We are not going to pre-commit to anything in terms of what we may or may not do for broader debt or capital planning. But suffice to say, we have the resources to pay that back, in the near term.
That is super clear. Thank you so much.
Our next question comes from Peter Christiansen from Citi. Please go ahead with your question.
Good morning. Thanks for the question. I wanted to ask about the digital assets part of the business. It generated a decent gross profit this quarter, yet EBITDA still negative. I am just curious, what are some of the key investments that is suppressing profitability there? How do you think for digital assets to become more sustainably EBITDA, what do we need to see occur?
Yeah. Thanks for the question. This is Tony. I will start, and Chris can add as he sees fit. I would say, broadly, the second quarter, as we talked about in our comments, the broader landscape was volumes and activity in the crypto space were down. Obviously, the price of crypto is still something underlying that drives a lot of the activity in our trading and derivatives businesses. We continue to outperform in that space.
Our volumes were up on a quarter-over-quarter basis. We believe we are continuing to take share in those businesses. Our lending book grew about 20%-25% quarter-over-quarter, and we are adding recurring revenue streams, which we think is the most important part of building a more durable business. We do continue to invest. We are making investments in a few different areas that we have talked about. Digital infrastructure areas where we see future growth.
As activity continues to pick up, I think you should see us move closer towards consistent profitability in that segment. The other thing I would mention, it is a bit nuanced, but we did have a small accounting adjustment in terms of how we accounted for certain things in the business, that if not for that change on a quarter-over-quarter basis, we would have been profitable in that segment just this quarter. It is sort of a one-time shift, but other than that, we are. Chris, I don't know if you have any other comments.
Yeah, I'll just supplement that a little bit. I think there's a few things going on in digital assets in the broader financial infrastructure markets that drive our thinking and our planning around profitability and what we're investing in. The key is one, volumes in the narrow niche crypto market obviously are suppressed and are a bit challenged at the moment.
Our business generally is a pipe business, where the more volume that flows through it, the more opportunity there is for us to generate revenue for the business, historically at least. We do have a long-term view on crypto assets, like Bitcoin, for example, and the decentralization theme, particularly in a more complicated, digitally driven agentic AI world. That aside, the other thing that's happening is a transition to real-world assets of the legacy financial system moving on-chain.
What we're doing today is prosecuting the business that we can in the crypto markets while investing in technology and people, and licensure, for example, so that we are going to be in a position to be on the forefront to be able to service the transition of volumes from traditional legacy real-world assets on-chain into a digital form.
While we never like to lose money, ever, being able to sustain the amount of revenue we have in this niche market while having the luxury of having the capital to be able to invest in growth so that we're positioned to capture that mega opportunity is really where we're at in the business today.
Thank you. Very helpful.
Our next question comes from James Faucette from Morgan Stanley. Please go ahead with your question.
Thanks very much. Wanted to ask quickly on Helios I and working with CoreWeave there. Just any insight you can share with us around what you know about their agreements with their partners for not only phase I, but also if that same partner is likely to also take on phases II and III.
Sure. Good morning, James. Thanks for joining for the question. I will set the stage for our relationship with CoreWeave, how it is set up, and how we understand their contractual structure and how their business model works. I think it is important, and it is illustrative of how we think about the risk and also the insight we have or do not have into their ultimate operations.
Our relationship with CoreWeave, with the three leases that exist today for Helios I, also that CoreWeave's parent company. We face CoreWeave, Inc. The way we think about that is that means that ultimately we face CoreWeave's entire portfolio of customers. We face effectively CoreWeave's residual equity and interest in those contracts after their site-level GPU secured financings. That is pretty much the business model. That is our exposure.
We like that exposure because what it means is that we are not subject to, at our facility, a single counterparty look-through exposure in the instance where there is an idiosyncratic problem with that one counterparty. CoreWeave's portfolio of pretty much all the entire hyperscaler universe and most of the frontier labs is the credit that we look to across their whole portfolio.
CoreWeave's contracts, while we do not have insight into all of them or into the specificity necessarily, CoreWeave generally contracts with its customers under four, I believe now the average is closer to five years as they reported, on a GPU hour basis generically, not on a facility level agreement. What that means is CoreWeave has the ability to balance their loads to their customers across their data center footprint.
That is important because it means that the answers to the question of who is the underlying tenant at Helios, by definition can and we expect will change over time, both because they commit generic GPU loads to their clients and because we know that the tenors of their contracts are shorter than the leases they have to sign, by definition.
We have been communicated from CoreWeave that the loads running today that we see every day on the screen in Helios I are coming from a customer contract that is currently allocated to a large multi-trillion dollar investment-grade hyperscaler. We do not have insight yet into ultimately who is going to be in phase II, let alone phase III.
That will happen as we get closer to energization and data hall turnover, and they decide at that point in time, in 2027, where their load is needed and where they want to allocate load to a particular facility. That is really the setup. I know that investors want to hear a direct look-through to investment-grade counterparty on the other side. It is just not the business model and not the way the contracts are set up at the CoreWeave side.
It is the platform and the risk that we underwrote when we originally did the transaction. That is not to say, I will add one thing. That is not to say that there won't be opportunities if it is credit enhancing to look through to a direct customer and drop it down in an SPV, for example, which we have seen happen once or twice in the market so far.
That sort of is downstream from CoreWeave financing their GPUs, and getting better execution on their GPU financing through an SPV direct deal and would come with our ability to approve or disapprove of any change in the tenant from the CoreWeave parent company down to a particular SPV. We will look at that on a case-by-case basis and evaluate it and assume it is credit enhancing. It is obviously something that we would engage with.
That is really helpful. Back on the Helios II and kind of the engagement there. Obviously, I think we are all sensitive to the changes in some of the market dynamics, et cetera, that Mike alluded to and mentioned at the outset. Are there key incremental market changes or clarifications that we should expect to come through before we would be able to announce a new tenant for Helios II? I am just trying to get a sense of what the timeline and checklist is we should be paying attention to for that. Thanks.
Yeah. The dynamics there, I tried to lay it out as clearly as possible in the prepared remarks. There are a few things that I would look towards. There is nothing today holding us back from executing on what would be a deal with an end tenant. The speculation around late summer and the timing and the conversation we are in, we could have signed a deal had we wanted to. A lot of that was in our control to make a decision to go forward.
We proactively chose not to because of the change in market dynamics. That still sort of exists today. There is obviously Texas. There is a lot going on in Texas, from a regulatory perspective. There are some key dates coming up. In December, we expect to have clarity on audit results that we have submitted across the assets, including Helios II.
All of those things. In April of 2027, we expect to get allocations of Batch Zero, Studied Load, power, et cetera. All of those things are on potential tenants' minds. What I would say quite clearly is, there are absolutely investment-grade tenants who, despite questions around the ERCOT process, want to contract and lock up power.
When they do the work on our portfolio, in particular on Helios I and Helios II, we think they generally come to the same conclusion we have, which is Helios II is a pretty unique asset in terms of when its studies were approved prior to the batch process and the likelihood that Helios II is going to be unimpacted generally by any of the uncertainty.
Nonetheless, those dates are key, and for some of the tenant base looking at it, they're looking for us to pass through those unscathed as we expect to, in order to sort of finally contract. Again, that's not the entire base. There are absolutely deals we expect we could do. The entire universe, clarity on the regulatory front, in Texas in particular, would help the process along.
Great. Thanks, guys.
Our next question comes from Devin Ryan from Citizens Bank. Please go ahead with your question.
Thanks. Good morning, everyone. I want to come back to Chris's comments on kind of broadening the pipes to be able to trade more, kind of couple parts here. What else are you watching on regulatory guidance that you guys think is critical from here? Assuming that progresses, how are you thinking about where Galaxy is best positioned to participate in tokenization? Meaning, how do you see Galaxy being able to compete with obviously well-funded incumbents that will likely be coming here as well?
Are there specific real-world assets that are more attractive to Galaxy? Even beyond just trading and trade finance, how do you see Galaxy's potential infrastructure partner here as we kind of progress of this world of tokenizing more real-world assets and trading them? Appreciate there's a lot in there, but just would love some thoughts. Thanks.
Yeah. Hey, it's Mike. Listen, so our big bet is that the infrastructure that crypto created is going to end up being the infrastructure that a lot of the traditional world operates on. When that happens, we'll see. It is starting to happen. I would tell you that almost every TradFi shop is building towards that, partly because they have to, because if it happens, they need to be there.
One of our businesses that we're putting a lot of effort into and is growing is actually being the partner to help build that infrastructure for TradFi companies. So that's in some ways a fee or a SaaS business, but the real upside of that is once you get partnership, the ancillary services, the trading, the potentially offering other products, that's where the growth of those things come.
We announced the one deal with Bank of New York. We are very engaged with a few other large participants. I think that's a good strategy. It's not going to dominate our revenue in the next 12 months. But when I look out 24 and 36 months, it's where I think we'll be a key part of that play. The second real angle is on-chain credit.
Chris talked about a few of the deals we've already done. We have credit as part of our DNA. Chris came from a credit shop. It's been a big driver of our business since we started. We are one of the leaders of on-chain credit today. We think that business is going to continue to grow. So building on-chain prime, building on-chain credit, on-chain asset management lives really on top of that.
I'd love to say in two years, our digital asset business is really an on-chain digital asset business, and that's what we're driving towards. When someone asks Tony about the investment dollars, they're all tech dollars, and they're all tech dollars and regulatory dollars going into that build-out.
Great. Thanks, Mike.
Our next question comes from Edward Engel from Compass Point. Please go ahead with your question.
Hi, thanks for taking my question. Back in February, I think you kind of acknowledged maybe aspirations that you'd have a deal for Helios II within the next 12 months. Should we assume this timeline maybe gets extended by a little bit, just given that you walked away from the deal this summer? Then just quickly on Helios I financing, I think you last mentioned that the phases I and II should have enough free cash so they kind of self-fund phase III. Has that also changed, just given maybe the exchangeable notes that are due and maybe some of these recent buybacks? Thanks.
Yeah. Hey, Ed. Good morning. I'll take the first part of the question. On Helios II leasing timeline, I think we've covered a lot of it so far. I think the expectations around timing of signing a deal, let's start from the fact that Helios II's energization timeline is expected to be in mid-2028. If you work back from mid-2028, that gives us 1.75 roughly years of room for both finding the right partner, negotiating the right deal, design planning, and then ultimate construction.
While that's not infinite time, that is a lot of cushion that we have, which we appropriately think we should take given how dynamic the market is, to find the right partner, get the right economic transaction. Because these are deals we have to live with for effectively the rest of our professional lives, given their tenor. So we take it very seriously. It's very important.
I wouldn't peg a specific time. I would tell you, though, that it's very difficult to enter into direct conversations with some of the largest companies in the world without an intent to ultimately transact, and we are in direct conversations with some of those companies right now. I am pretty positive and pretty bullish on the opportunities sitting in front of us for that right now. Know that our end date solving backwards is to deliver those potential clients' data halls in 2028.
On the second part of your question, Ed, as you know, phase I and phase II are fully funded through the completion and end of construction. We will have the opportunity to refinance phase I if we think the economics make sense from a refi standpoint, which may give us the opportunity to take some additional capital out.
But importantly, we do not need to do that. As we've said before, phase III is fully equity funded through the project. We intend to pursue debt financing sometime in the first half of 2027, and the equity financing being fully available already is not changed.
Great. Thank you for that clarity.
Our next question comes from Brett Knoblauch from Cantor Fitzgerald. Please go ahead with your question.
Hi, guys. Thanks for taking my question. On just the permitting environment, I think you guys touched on it briefly in your prepared remarks regarding Helios II. How critical is it to have all those permits in place before maybe we look to sign a lease there? How are your customers thinking about maybe a permitting freeze and, if the election goes the other way in Texas, that potentially could be extended?
Yeah. On the permitting front, we could take this in a couple different tranches.
Yes
For Helios I, II, and III under development, we have all the permits in hand needed to begin and continue construction. There are still a handful of permits that the TCEQ primarily would issue and or approve that are not permits that would prevent us either from continuing construction or ultimately delivering and energizing data halls for the client. They are unrelated to cooling.
They primarily relate around the potential to operate diesel gen backups and some other potable water and things on site. They are not limiters. All the permits we need for those things are in hand already. For future developments that will require sort of those level of permits, they are months to years out from when we would actually need them to begin construction.
We are just very confident that while today there is a permitting freeze, while ERCOT and all the stakeholders are attempting to comply with Governor Abbott's announcement, we are not of the belief that those timelines are going to intersect anywhere close to one another, that we have concerns that permitting is going to be an issue for us at Helios.
Perfect. Thank you, guys. Really appreciate it.
Our next question comes from Jon Petersen from Jefferies. Please go ahead with your question.
Oh, great. Thanks. Sorry to keep going on the data center stuff, but you outlined that some of the hyperscalers might be willing to secure power now even before final approval. Or, sorry. Yeah, secure a lease here and power even before final approvals from ERCOT.
Curious what you are hearing in the project financing market and if the debt markets are, I guess, willing to be that forgiving or if we have to get past that April deadline for any potential leases you sign today to be financed. Then maybe also just a follow-up just, I guess, as you talk to potential lenders, how much stricter are they being on different debt covenants and LTVs and LTCs and all those sorts of metrics?
Yeah. So, it is a very good question. From a counterparty perspective, as you pointed out, we are not seeing broadly really any pushback to, "Hey, let us just not engage until we see clarity in Texas." People know power exists in Texas. People know power is affordable and getting cheaper in Texas. People know that that is going to be a Tier 1 market, we believe and they believe for a lot of time to come for these kinds of assets.
So, the conversations we are having really, they want to understand the dynamics at play with our assets, but they are certainly willing to move past that pretty quickly. On the lender side, first and foremost, the existing facilities we have in place, are not impacted at all by any potential perceived uncertainty around the Texas regulations. So there is no covenant breaches. There is no risk to capital availability, in those facilities.
Those are fine. On a go-forward basis, I do believe the risk on investing capital in advance of knowing for sure whether you have the right to energize is a risk that we, in theory, would be taking should we choose to do that, and therefore, what lenders would be choosing to take along with us. Lenders generally are a lot more conservative than we would be, as a developer in the equity. The good news for us is, as Tony pointed out, our potential financing needs outside of our own capital start to get highlighted as we turn the corner into Q1 and to Q2 of 2027.
From a timing perspective, we, for better or for worse, have the luxury of being able to assess the regulatory environment, see how December goes with respect to the RFI audit results, choose to deploy our own capital at risk based on that landscape. We really don't have to go to the markets to have conversations or will want to go to the markets to have conversations until there's a lot more clarity in 2027. Not all of our competitors are in that situation. We happen to have the luxury of being able to do that, and we're happy with that.
Okay, great. Thank you.
Our next question comes from Joseph Vafi from Canaccord. Please go ahead with your question.
Hey, guys. Good morning. Nice to see progress on all fronts here. I was wondering if we could kind of just drill down on the digital asset side of things. Is there anything particular to point to here on trading share gains in the quarter that you'd like to call out? Then just the appetite to continue to expand the lending book from here from a risk management point of view. Comments there would be appreciated too. Thank you.
Yeah. Hey, Joe. Good morning. On the trading side, I think factually the numbers are that our volumes grew 12% quarter-over-quarter and industry volumes were roughly flat. I think from a pure share gain perspective, we continue to make incremental gains of market share, albeit in a fairly slow and tepid environment for traditional crypto assets. But the team is performing extremely well given that condition, and Galaxy's getting stronger.
On the lending side of the business, we did expand our lending footprint over the quarter pretty dramatically. I think we were up 27% quarter-over-quarter average loan book size. We think there's huge opportunity in that space. We think that we could probably grow faster, if we wanted to. That's not always the right thing to do, as a lender is grow whenever you can. And so we don't.
We pick our spots, and we choose them wisely. As I said in my remarks, the other thing we're really focused on on the lending side is we're pretty cognizant that you can't really just grow a lending business ad infinitum, with a finite balance sheet, particularly a finite balance sheet that has a lot of uses across the company.
The good thing is we do have the capabilities internally, and we think there's a very large opportunity to expand our structured credit offerings to originate loans, package them, sell them off, think more traditional financial debt capital markets businesses, in a whole new design space on chain that really has not been tapped yet in any size.
I think you should expect us to both grow our balance sheet exposure to our loan book, but also be able to grow volumes outside of just our own balance sheet exposure. The team is hard at work on creating structures, creating risk boxes so that we can do that responsibly for ourselves and potential clients and offtakers. More to come on that, but I think that's the future you'll see of our lending business.
Great. Thanks very much, Chris.
Our next question comes from Chris Brendler from Rosenblatt Securities. Please go ahead with your question.
Hey, thanks. Good morning, and thanks for taking my question. I wanted to follow up on the idea of the sort of modernization of financial infrastructure and this whole space moving on chain and Galaxy's position within that opportunity. You had the BNY this last quarter. Any other sort of color you could provide on your progress with partnering with traditional financial institutions and helping with this process?
Then zooming out a little bit, what does the revenue mix look like in digital assets as, if you are successful in capitalizing on this opportunity, how much will the business move away from sort of transactional and sort of lending and trading activity and more towards either asset management, custody, or staking activities as you progress on that opportunity? Thanks.
Well, certainly the goal is to move it away in a large fashion, right? I think part of that is how fast this stuff moves on chain. It is one of these things that my intuition is it will go slow and then it will go fast, kind of like the way people go bankrupt, right? Like it happens, it happens, and all of a sudden you are like, "Well, how did that happen?" So we have not seen it yet.
We are seeing lots of positive signs, like stablecoin usage is a big part of it. But we have not seen a wholesale shift to tokenization. People still trade normal equities, not tokenized. Partly that is because most of the traditional players have not built out their infrastructure.
There is a scramble going on, that we're part of, for these institutions to build out their digital wallets, their custody, their staking abilities, or to hire people to do their staking abilities for them. I think once that infrastructure shows up, you very well might see that part of the business really start to grow. We're walking that tightrope between making sure our core business that we have operates as best it can and makes as much money it can while investing for this future that we're very confident will come.
We just don't have the forward calendar, with where the checkered flag is. I know it's not a perfect answer for you, but that's exactly the mindset that we have. Again, we wouldn't be making this investment if we didn't have strong conviction that that's where the world is going and seeing lots of green shoots.
I'll just add, if you think about the revenue mix, I think that was your second question. To Mike's point, this build-out in infrastructure and adoption will take time. But one way to think about it is our more traditional businesses around trading, derivatives, lending, et cetera, will grow, the opportunity set will grow with the overall growth of the pie. So the mix may shift for us, but we think that the opportunity set, for the more recurring fee-based revenue, as well as the more transactional-based, can grow together, as the industry grows and adoption continues.
Yeah, the last piece I'll add on that is just from an infrastructure business model perspective, which I think is part of the question. We see the financial players and institutions around the world in two categories, or at least they pick and choose these two categories depending on the service needed, right? There's a, "We need to own it, own the underlying technology, own the underlying operating risk," or, "We are okay white labeling and contracting out."
So what we are doing today is we are selectively partnering with some of the largest institutions who need to own certain components, think base layer custody, think maybe like a custody allocation engine, et cetera, to help build that, and we get paid to help build that.
That increases not only their capabilities, but our portfolio of capabilities to then offer as a platform service to those companies and everyone else for the components of those services that they are okay not owning and would rather contract out to create recurring revenue for us. The model is getting paid both upfront to help build, but also over time, stacking a recurring revenue business for a platform service for both those customers and the long tail of middle market and lower market financial institutions who would never need to build and own their own capabilities.
That's fantastic color. Thanks so much.
Our next question comes from Ben Sommers from BTIG. Please go ahead with your question.
Hey. Good morning, and thank you for taking my question. Kind of to build off the last one a little bit, wanted to ask around the individual investor adoption of digital asset product offerings through GalaxyOne. Seems like the TradFi institutional interest is significant, but kind of curious how GalaxyOne could potentially contribute to the revenue mix over time.
Yeah. We haven't talked about GalaxyOne in our prepared remarks or really in questions a lot recently. We're actually under the hood really excited about not only what's going on there, but also the product roadmap. Our assets on platform, while still relatively small there, have grown quite substantially. We're well north of $125 million, I believe, the last number I saw there, and growing fast.
On the forward, this year was really a build baseline capabilities and ensure that our retail-facing offering had digital asset, crypto-related capabilities that were on par with what's available in the market. We still got a little bit of work to do on there, but that was the big lift so far this year. What we look forward to next year in 2027 actually is rolling out expanded brokerage capabilities across a number of different asset classes.
Our real focus there is not only just baseline capabilities to compete with existing big platforms, but actually offer much more higher-end bespoke wealth management style access products and capabilities for high earners, individuals with net worth that they want to preserve and grow. We have a really cool lineup of product roadmap on that front that we're really excited to start rolling out in Q1 2027. So more to come on that, but we think the opportunity set is much broader than just offering crypto access, and is going to be quite differentiated in 2027.
Super helpful. Thank you for taking my question.
Our next question comes from Michael Kim from Bloomberg. Please go ahead with your question.
Hey, good morning. Thanks for taking my question. I noticed that the average annual revenue assumption for phase II of Helios I just ticked down a bit from last quarter's forecast. I am just wondering if you could maybe flesh that out in terms of, or maybe with respect to your earlier comments around supply remaining constrained and demand strengthening as well as your competitive positioning. Thanks.
Yeah, Michael, it's Tony. I will address the first part of the question and then if Chris wants to hit the other. Nothing has changed with regard to phase II and the revenue. It's simply a clarification of the recognition under lease accounting for GAAP purposes that we just wanted to help analysts clarify for their model building. Nothing fundamentally has changed. Phase II is exactly the same, so.
Got it. That's helpful. Thanks.
Our next question comes from Bill Papanastasiou from Chardan Capital Markets. Please go ahead with your question.
Yeah, good morning. Thanks for taking my questions. We're hearing that long lead time item equipment procurement is becoming increasingly harder to come by. Just quickly here, how is the team strategizing equipment procurement for its uncontracted sites and lining that up with energization timelines? Any pressure there on the Helios II build-out timelines? Thanks.
Sure. On the procurement side, our internal mantra is be early and be deliberate, but don't get too far ahead of yourself on demand. It's always a delicate balance for us. Again, for us, we have the lowercase L luxury of having a balance sheet and resources available to us to allow us to lean in where we see potential future supply chain bottlenecks.
We're now becoming a large consistent buyer of a lot of the infrastructure, and so the big suppliers, while not on the same level as the hyperscalers, we are a key customer to critical infrastructure suppliers now. And so we have the access and the ability to place deposits, lock up future capacity for ourselves. For example, for Helios II, we've already committed approximately $230 million in deposits and purchase orders for future electrical infrastructure.
Mostly focused on long lead time equipment that early on we saw the potential for some supply chain disruption. I would not characterize supply chain for infrastructure today as something that's completely prohibiting and limiting from growth and expansion. It's available. There are also new suppliers globally that are emerging that we have traveled around the world to meet with and evaluate as to whether the equipment meets the highest standards so that we can install into our infrastructure. And we've made substantial progress on that front too.
Our final question today comes from Martin Toner from ATB Cormark Capital Markets. Please go ahead with your question.
Thanks for taking my question. Mike is a very unique CEO because of his market sense. I wanted to ask Mike what he thinks of the credit market. Your customers need a lot of money or seems like kind of a standoff to me. I would love to hear what Mike thinks about it and how he thinks it will be resolved, and on what timeline, if he is willing to speculate.
Yeah, listen, it is a great question. You would be living under a rock if you did not realize there was stress in the credit markets recently. The equity markets have sold off some, but not as dramatic as the stress one might have seen in the credit markets. That has happened for a few reasons, right? This war in Iran has kept oil prices over $100.
That pushed up the 10-year yield, which pushed up the credit yield, and just the recognition of just how big this infrastructure build is going to be. I think partly pressure or higher rates, as a base case, is starting to be absorbed psychologically by the market. We are not going to go back anytime in the short term to short-term interest rates at 1% or 2% and long-term rates below 4%, right?
There was a whole generation of traders and investors that started to live with that as their base case. I think we are going back to an era where short-term rates are probably between 4% and 6%, 4% and 5%. The yield curve tending to be more steep. In some ways, this is a one-time adjustment in psychology that is coming off of what was really an era that was abnormal historically. I do think this election is going to come and go. There is going to be a lot of readjustment for the first week or two, and then like always, economics take over from politics when it comes to markets. If the war ends and oil comes down.
If Trump attacks and we end the war, or we pressure them to end the war, there are people negotiating like crazy right now to try to end the war. You get oil, going back from 100 to 60, it is going to take a tremendous amount of pressure off the fixed income markets and hence the credit markets.
Chris was saying we have the luxury of being able to wait to the turn of the first quarter, second quarter before we really need to tap credit markets. I am thrilled with that because I know one thing, the tension never lasts forever. It always resolves itself. There are bullets that Scott Bessent has in his gun. We will see if he uses them. Kevin Warsh will most likely raise rates, not the week before the election, but soon after.
At one point you slow the economy because you have to, if that's what it takes. We have indigestion in the credit market. It is probably close to peak nervousness. There's not an easy path out. But I think time is going to heal this thing. Mostly that's going to be the reaction to the oil prices. The good news is Meta's coming out and doing a monster equity raise.
Anthropic said they have $130 billion on their balance sheet, I think that was the number, is going to raise another $100 billion. Every time I see giant equity raises, I breathe a sigh of relief for the credit market. You could argue is it good for the equity holders of those companies or not that they're being diluted?
That's a separate discussion, but more equity, less credit is one way to think about it. This last week to some degree was good for the credit markets because we saw this massive amount of equity. I don't think the belief in what's coming in AI is going away. Just yesterday, four of the most, to a layman, absurd math problems got solved by these machines.
The breathtaking accomplishments of AI are going to start translating into medical advancements and even consumer. If you look at the consumer apps from Instinct to Muse to Grok Bot, as consumers start using them, the demand for AI picks up, but also the acceptance of AI. It's going to be a politically charged issue. But this is the transformative technology that is literally going to change the world. In some ways we need it to win.
Galaxy is betting on it that it's going to win. But quite frankly, as an American, we need it to win. When I think of $41 trillion of debt, the only way out is we have above average growth with lower inflation, which is one of the promises of AI, or at least one of the potentials of AI. One of the reasons we've made this bet is because we're good patriots.
But we actually believe that this really. Listen, there's going to be a new regulatory framework that comes out sometime in the next 12 - 18 months. That meeting at the White House, which looked like it was performative, wasn't just performative. They've come out with a bunch of things that are going to slowly be, or quickly be instituted to make people feel better about AI safety. But that conversation's not going away either.
We plan on being part of it, being at the table, trying to help guide. I think the best AI policy is one that keeps people optimistic. You might see some things that are perceived as small step backwards to go two steps forward. But this is, again, the country's making a long-term bet on AI, Galaxy is, and I think we're pretty optimistic it's going to turn out well.
That is great. Thank you very much, Mike.
With that, we will conclude our question and answer session. I would like to turn the floor back over to Mike for any closing comments.
Yeah. Listen, I want to give a shout-out to the Galaxy team. First of all, people have been working their tail off, in all aspects of this thing. Also to our investors, nobody likes when their stock goes down, or dislikes it as much as I do. We are working hard to build a company that someone always. Stock markets are weighing machines, and we really hope that in six months, 12 months, 18 months and 36 months, when we are weighed, the conclusion of the equity market is that we are worth a lot more than we are today. So that is our goal. Appreciate the questions, and appreciate the support.
With that, we will conclude today's conference call and presentation. We do thank you for joining. You may now disconnect your line.