DigitalBridge Group Inc - Ordinary Shares - Class A Stock

Hello, and welcome to the DigitalBridge Group, Inc. Annual Meeting of Stockholders. Please note this meeting is being recorded. Questions may be submitted via the questions box to the right of your screen. [Operator Instructions]

Good morning, ladies and gentlemen. I am Nancy A. Curtin, Chairperson of the Board of DigitalBridge Group, Inc., and it's an absolute pleasure to welcome you to call to order our 2026 Annual Meeting of Stockholders. In accordance with our bylaws, I will serve as Chairperson of this meeting. We will first conduct the formal business of the meeting by voting on the proposals described in our proxy materials. As is our custom, we will answer questions about the proposals on the company following presentation of the proposals and the conclusion of the meeting. Only stockholders may submit a question. [Operator Instructions]

Please note that in the interest of all stockholders, we will only address those questions that are pertinent to the business of the meeting. Before proceeding further, I'd like to introduce the company's other directors and certain officers who are present at the meeting: James Keith Brown, Jeannie Diefenderfer; Mark C. Ganzi, CEO of the company; Gregory J. McCray, Shaka Rasheed, Dale Anne Reiss, David M. Tolly; Jay Wintrob; Ben Jenkins, President and Chief Investment Officer; Thomas Mayrhofer, Chief Financial Officer; Liam Stewart, Chief Operating Officer; Geoffrey Goldschein, Chief Legal Officer and Secretary; and Severin White, Head of Public Investor Relations.

We have received an affidavit from our proxy solicitation agent certifying to the mailing of notice of this meeting commencing on April 28, 2026, to stockholders of record at the close of business on the record date, April 24, 2026, wherein the company furnished to such stockholders a proxy statement, annual report for the period ended December 31, 2025, and a proxy card. This affidavit, together with copies of the notice meeting, proxy statement, annual report and proxy card will be filed with the minutes of the meeting. Barry Rosenthal of the American Stock Transfer & Trust Company has been appointed Inspector of Elections and has taken his oath. A copy of his oath will be filed with the minutes of the meeting.

As of the record date, there were 182,378,179 shares of our Class A common stock and no shares of our Class B common stock outstanding. The Inspector of Elections has advised me that a quorum is now present. Therefore, this meeting is duly organized for the transaction of business. We will now commence with the formal business of the meeting.

The first item of business on the agenda is the election of the 9 individuals to serve as members of the Board of Directors of the company for 1-year terms until the 2027 Annual Meeting of Stockholders and until their successors are duly elected and qualified. Under the company's bylaws, the only persons who have been properly nominated are those nominees listed in the company's proxy statement. Myself, Nancy A. Curtin, James Keith Brown, Jeannie Diefenderfer, Mark C. Ganzi, Gregory J. McCray, Shaka Rasheed, Dale Anne Reiss, David M. Tolly, Jay Wintrob.

I there declare the nominations for directors are closed. The proposal to elect the 9 nominees is described in detail in the company's proxy statement distributed in connection with this meeting, and the following resolution is deemed duly presented at the meeting.

Resolved that the stockholders of the company hereby vote to elect each of the 9 nominees identified above to serve until the 2027 Annual Meeting of Stockholders and until their successors are duly elected and qualified.

The second item of business on the agenda is the approval on a nonbinding basis of the compensation of the company's named executive officers, which is described in detail in the company's proxy statement distributed in connection with this meeting. The following resolution is deemed duly presented at this meeting.

Resolved that the compensation paid to the company's named executive officers as disclosed in the proxy statement pursuant to Item 402 of Regulation S-K, including the compensation discussion and analysis, compensation tables and narrative discussion is hereby approved on an advisory basis.

The third item of business on the agenda is the approval of the amendment to the DigitalBridge Group, Inc. 2024 Omnibus Stock Incentive Plan. The following resolution is deemed duly presented at this meeting.

Resolved that the amendment to the DigitalBridge Group, Inc. 2024 Omnibus Stock Incentive Plan in the form presented in the proxy statement is hereby approved.

The fourth item of business on the agenda is the ratification of the appointment of Ernst & Young LLP as the company's independent registered public accounting firm for 2026, which is described in detail in the company's proxy statement distributed in connection with this meeting. In addition, representatives of Ernst & Young LLP are here today to respond to any appropriate questions stockholders may have during the question-and-answer period later in the meeting. The following resolution is deemed duly presented at this meeting.

Resolved, stockholders of the company hereby ratify the appointment of Ernst & Young LLP as the company's independent registered public accounting firm for the fiscal year ending December 31, 2026.

The polls are now open for voting on these proposals via the Vote My Shares tab at the top right of your screen. The floor is now open for questions or comments from the stockholders of the company concerning any of these proposals. Please note that in the interest of all stockholders, we will only address those questions that are pertinent to the business of the meeting.

There are no questions.

If there are no further questions, we will proceed with the meeting. If you've already voted your shares and do not wish to change your vote, no action is required at this time. If you have not yet voted or would like to change your vote, you may do so now by clicking the Vote My Shares tab at the top right of your screen. If you previously voted by proxy, submission of a new vote will revoke your prior proxy.

I will ask the Inspector of Elections to review all submitted votes. There being no further ballots, I declare the polls are closed. All ballots and proxies are now in the custody of the Inspector of Elections. The Inspector of Elections has determined that our preliminary counts indicate that the proposals 1, 2, 3 and 4 have been approved by the necessary votes. Therefore, the resolution for such proposals have been duly adopted. The report of the Inspector of Elections, which contains the final vote totals will be filed with the minutes of this meeting in the company's minute book and in our 8-K that will be filed with the SEC.

This concludes the 2026 Annual Meeting of the company's stockholders, and I now declare this meeting adjourned. I want to thank all of you for attending today's meeting. We are grateful for your interest and support of DigitalBridge. Stay safe and well.

Thank you. Ladies and gentlemen, this does conclude today's presentation. We thank you for your participation. You may now disconnect, and have a wonderful day.

”

”

”

”

” TD Cowen

” KBW

” B. Riley Securities

” Raymond James

” JPMorgan

” Wells Fargo

Greetings, and welcome to the DigitalBridge Group, Inc. Third Quarter Earnings Conference Call 2025. As a reminder, this conference is being recorded. It is now my pleasure to introduce your host, Severin White, Managing Director, Head of Public Investor Relations. Thank you, sir. You may begin.

Good morning, everyone, and welcome to DigitalBridge's Third Quarter 2025 Conference Call. Speaking on the call today from the company is Marc Ganzi, our CEO; and Tom Mayrhofer, our CFO. I'll quickly cover the safe harbor.

Some of the statements that we make regarding our business operations and financial performance may be considered forward-looking, and such statements involve a number of risks and uncertainties that could cause actual results to differ materially. All information discussed on this call is as of today, October 30, 2025, and DigitalBridge does not intend and undertakes no duty to update it for future events or circumstances. For information, please refer to the risk factors discussed in our most recent Form 10-K filed with the SEC for the year ending December 31, 2024, and our Form 10-Q to be filed with the SEC for the quarter ending September 30, 2025.

With that, let's get started. I'll turn the call over to Marc Ganzi, our CEO. Marc?

Thanks, Severin, and welcome, everyone, to our third quarter 2025 business update. We appreciate you joining us on the call and look forward to answering your questions.

Let's get to the quarter. This quarter really exemplifies what we've been building towards at DigitalBridge from our near-term financial goals to our longer-term strategic priorities. Let's get started with the key highlights that align with our strategic road map.

First, financial performance. DigitalBridge delivered another quarter of robust growth with fee revenues reaching $94 million, up 22% year-over-year. Our fee-related earnings grew 43% to $37 million in the third quarter, reflected continued margin improvement as revenue growth continues to outpace expenses.

Second, capital formation. We raised $1.6 billion in new capital during the quarter, bringing our year-to-date to $4.1 billion. Look, we're well positioned thinking through the fourth quarter here as we remain on track to meet our full-year objectives. As most of you know, the fourth quarter is historically our strongest quarter.

Finally, and this is the most important story of the quarter, the relevance and strategic value of our power bank was on full display. We saw record data center leasing activity across our portfolio that will build and accrue significant value for you, our investors, over time. Our portfolio company, Vantage Data Centers announced the Frontier mega campus in Texas, a $25 billion, 1.4 gigawatt development, serving the leading AI infrastructure build-out.

This was followed up by a second campus, dubbed Lighthouse in Wisconsin, a $15 billion-plus development to support the expanding OpenAI and Oracle Stargate project. These landmark transactions demonstrate that our years of securing power across the portfolio are now translating into the largest leasing commitments in data center history. I talked about it last quarter, having a power bank that is ready to go for our customers is a comparative advantage.

Let me put this quarter's performance in a broader context. Continued financial performance and capital formation that advances us towards exceeding our full-year objectives. What makes this quarter truly distinctive is how our strategic positioning around power is creating differentiated outcomes at the portfolio level. For years, we've talked about the importance of power as the critical constraint in the AI era. Today, we're seeing that thesis play out in real time, and DigitalBridge is leading on the front.

As I referenced, year-to-date capital formation of $4.1 billion positions the firm to surpass our financial targets. We achieved our $40 billion FEEUM target 1 quarter ahead of schedule, reaching $40.7 billion as of the third quarter. This milestone that reflects both the strength of demand for digital infrastructure and the execution capabilities of the DigitalBridge global platform. The record FEEUM today translates directly into revenue and earnings growth.

We're seeing particularly robust activity in co-invest, where third quarter fee rates continue to expand relative to historic levels, up to 70 basis points in Q3. I talked about this earlier this year in multiple quarters. We're very focused on expanding margins in our co-investment program, and we're getting it done. That's the key. We're executing.

We're finalizing our flagship strategy capital formation, targeting over $7 billion in the next few weeks as we head into the end of the year, our focus has pivoted to the second credit strategy and our new offerings in power, stabilized data centers and private wealth that will drive our 2026 capital formation.

Having a new product pipeline that sets you up for success is really what it's about in terms of being an alternative asset manager where we have a multi-strategy platform. This is the full effect of DigitalBridge as a full alternative asset manager. This is on display for all of our investors as we push forward into 2026.

Next slide, please. Now I want to talk about a key component of our private wealth strategy, the partnership we announced with Franklin Templeton in the third quarter to launch our first programmatic private wealth distribution channel. At its heart, the partnership is about democratizing access to institutional quality, differentiated digital and energy infrastructure investments that were previously reserved for institutions. Franklin Templeton is a $1.6 trillion global investment leader and their CEO, Jenny Johnson, has prioritized this initiative as growing alternative investment portfolios.

Importantly, Franklin Templeton are building a diversified open-ended infrastructure solution that will have the ability to invest across all infrastructure subsectors. They intend to compete head on with the mainstream supermarket asset managers. On our side, we're bringing our $100 billion-plus in assets under management and our position as the leading digital infrastructure specialist across data centers, cell towers, fiber networks, digital energy and edge infrastructure.

We're partnering with our friends at Copenhagen Infrastructure Partner, the world's largest dedicated greenfield energy fund manager with $37 billion in AUM and Actis backed by our friends at General Atlantic with their deep sustainable infrastructure expertise. For their part, Franklin will focus their accredited investor product on the mass affluent segment in the market, a difficult segment to access without significant investment in sales infrastructure.

They have a sales force of over 600 people, giving them strong distribution capabilities and reach. The strategic rationale here is compelling. Together, we're focused on a massive investment opportunity. There's a $94 trillion global infrastructure need by 2040. We're positioned at a pivotal inflection point as AI, electrification and connectivity megatrends accelerate infrastructure demand.

Now, why does this matter for you, our DigitalBridge shareholders? Look, first, the 3 reasons: One, evergreen capital. This is an incremental source of capital and FEEUM that layers over time in a long-duration structure.

Second, it's an earnings contributor. Fee revenues convert to fee-related earnings as the platform scales. Then third, earlier carrier realization. The potential private wealth carry is paid as accrued earlier than our traditional institutional structure. This partnership launches exactly at the right time, and it supports our strategy of building a multichannel approach to wealth sales. It enables us to reach multiple client segments across the broader wealth universe.

There is a secular migration of wealth management allocations to private infrastructure. This is happening. The institutional quality solutions were designed are meant to provide stable, inflation-linked cash flows with resilience through economic cycles. We're capturing what we believe is a massive opportunity and Franklin Templeton gives us distribution platform and private wealth client access to do it at scale. That's the key component that we're doing this at scale.

Next slide, please. Let me bring this all together with what I believe is the defining characteristic of the DigitalBridge portfolio today, our power bank. To be credible and to be honest with our customers today, if you don't have a power bank, you really can't have a conversation in terms of leasing megawatts and gigawatts. Last quarter, I highlighted this. We have over 20 gigawatts of total secured power across our data center portfolio. That's not a projection. That's actual power that we can access. That's critical to understand that, that this is not a book dividend or something that we're trying to accomplish. This is power that exists inside of existing land, existing facilities, existing campuses with our 11 existing platforms.

In the third quarter, we put that power bank to work and leased a record 2.6 gigawatts across the DigitalBridge portfolio. To put that in perspective, that represents 1/3 of total record U.S. hyperscale leasing for the quarter. 1/3. That's not market share. That's market dominance in the most important segment of the data center industry today. Here's what it means in practical terms.

When the world's largest technology companies need to deploy AI infrastructure at scale, they come to our portfolio companies. They come because the portfolio companies have a long track record of delivering for them and because they've got the power. In today's environment, power is everything. You cannot build a 1 gigawatt AI campus without 1 gigawatt of power. It's just that simple. Ultimately, the 2.6 gigawatts of third quarter leasing translates directly into new capital formation, fee revenues and carried interest and long-term value creation. These are decade-plus contracts with investment-grade counterparties. The revenue visibility is exceptional, and the returns are improving relative to what we underwrote when the power was originally sourced.

As you think about DigitalBridge's positioning today, think about it this way. One, we have the power; two, we have the platforms; three, we have the customer relationships; and four, we are executing. That combination is creating outcomes that very few firms in the world can deliver. I would argue we are actually the only firm that can deliver it on a global basis, and we're only in the early innings of this cycle. I cannot be more excited about this development.

Again, this has been set up. This has been our conversation with you, our investors, for the last 3 quarters. How would we translate this 20-plus gigawatt power bank into comparative advantage? This is as easy as you can see it for investors today. We have the capability, we have the advantage, and we're executing.

Next slide, please. Now let me put the power bank into broader context of what we're building across the entire DigitalBridge portfolio. Look, across our 11 data center platforms, we're deploying significant capital to support the growth of the AI ecosystem on a truly global scale, catalyzing development from hyperscale to private cloud to the edge, spanning North America, Europe, Asia Pacific and Latin America. The key to this is it's a customer-driven investment model following the logos where the hyperscale, enterprise and cloud customers are demanding capacity.

In North America, Switch, Vantage, DataBank and Expedient are each scaling to meet differentiated customer segments from the largest hyperscale AI workloads to enterprise edge computing. In Europe, Vantage EMEA and Yondr, our newest platform, is building out critical capacity across multiple markets. Vantage Asia Pac and AMES are positioning us for rapid growth in Asia Pacific, while Scala continues to lead in Latin America, and AtlasEdge is capturing the emerging opportunities at the intersection of connectivity and compute in Europe where inferencing will come into full focus in the next decade.

We have the products for every type of workload. We have the products for every type of workload in every geography. This is by far the most unique and differentiated data center platform in the world. What makes this powerful is the diversity and complementary of these platforms. We're not a one-product shop. We have the right platform for hyperscale GPU compute, for private cloud workloads, for enterprise colocation, edge infrastructure, and of course, now we move to inferencing.

That breadth means we can serve the full spectrum of AI infrastructure demand. It means our customer relationships deepen as their core requirements evolve. That's what I love. I love evolving with customers. Just like we did 30 years ago when we evolved the towers from analog to digital into multiple different technologies over the last few decades. We're capturing that same business model with our customers today in data centers.

The capital we're deploying across these platforms is measured in tens of billions of dollars over the next several years. It's directly tied to contracted customer demand and secured power positions. This is DigitalBridge's competitive advantage at scale, a global platform with local expertise backed by institutional capital following customer demand and enabled by our market-leading power bank.

With that exciting overview, let me turn over the call to Tom to walk you through the financial details, and I'll come back later to wrap it up. Tom?

Thanks, Marc, and good morning, everyone. As a quick reminder, the full earnings presentation is available within the Shareholders section of our website. As Marc discussed, we had an exceptionally strong third quarter, supported by continued capital formation in our flagship fund series, which generates high-margin catch-up fees. Throughout my remarks, I'll highlight the impact of these catch-up fees in order to provide a baseline for our prospective performance once we complete the fundraise for our current flagship funds in the fourth quarter.

Starting with the financial highlights. In the third quarter, we recorded $93 million of fee revenue, representing an increase of 22% over the third quarter of 2024. Our fee revenue this quarter benefited from the cumulative effect of organic growth in our flagship fund series and co-investments over the last 12 months with a $8 million contribution from catch-up fees in the third quarter.

This growth in fee revenue resulted in $37 million of FRE in the quarter, an increase of 43% over Q3 of last year and putting us on track to hit or potentially exceed the top end of the range for our 2025 full-year FRE guidance. Excluding catch-up fees, FRE for the quarter would have been $29 million, an increase of 36% year-over-year. Growth in FRE resulted in distributable earnings of $22 million for the quarter, representing a double year-over-year.

As of quarter end, our available corporate cash was $173 million, providing material liquidity and flexibility for us as we continue to evaluate both our capital structure and opportunities to invest in and grow our business. We also currently hold $54 million of warehouse investments on our balance sheet to support the launch of new power and private wealth strategies, which we expect to recycle over the next year as we raise third-party capital for these products.

Moving to the next page. Fee-earning equity under management increased to $40.7 billion as of September 30, representing a 19% increase from last year. This growth is primarily driven by capital formation in the DBP series and co-investments as well as fees activated upon deployment of previously raised capital. We closed $1.6 billion in new fee earning commitments during the quarter, led by strong co-investment activity and new commitments to our latest DBP flagship fund.

Turning to the next page, which summarizes our non-GAAP financial results. As mentioned earlier, we reported $93 million of fee revenue in the quarter, representing growth of 22% over the same quarter in the prior year. Our LTM FRE margin was 38% as of the third quarter. We expect FRE margins to remain elevated through the final close of our flagship fund in the fourth quarter of 2025, supported by the continued contribution from catch-up fees.

Moving to the next page, which summarizes our carried interest and principal investment income. We reported a $20 million reversal of carried interest during the quarter. As a reminder, the company accrues carried interest based on quarterly changes in the fair value of our fund investments. As discussed previously, many of our vehicles are in the early to middle stages of their life cycle and have not fully worked their way through the J curve to be entirely clear of the preferred return.

At this point in their life cycle, small changes in the fair value of the fund assets can have an outsized impact on the quarterly accrued carried interest that we report, including causing reversals as we've seen this quarter in periods when the appreciation in the portfolio does not exceed the preferred return hurdle for the quarter.

As we've discussed in prior quarters, carried interest compensation expense tracks these changes, and therefore, there was a commensurate reversal of a portion of the unrealized carried interest compensation this quarter. Principal investment income, which represents the mark-to-market on the company's GP investments in our various funds was $25 million.

Turning to the next page. This chart continues to highlight the stability and consistency in growth, both in revenues and margin that we've experienced over the last 2 years. We've included this quarter FRE metrics, both gross and net of catch-up fees, given the more meaningful contribution from catch-up fees this year as we close out the fundraising period for our most recent DBP fund. LTM margin, excluding catch-up fees, has grown to 33% as of September 30.

This quarter, we saw $1.1 billion of FEEUM inflows, a significant portion of which was related to the activation of fees on previously raised co-investment capital. These inflows were partially offset by approximately $100 million of outflows.

Finally, the company continues to maintain a strong balance sheet with approximately $1.7 billion of corporate assets, largely reflecting our material investments alongside our limited partners and available corporate cash. We're pleased with our results through the first 3 quarters of the year, and we're very excited about the opportunity set that we see ahead of us, both in our core business and some of the new initiatives that we're working on.

With that, I'll turn the call back over to Marc.

Thanks, Tom. Now I want to shift gears and talk about our investment activity and how we're creating value at our portfolio companies. As a reminder of the framework we outlined last quarter, our competitive advantage is built on a 3-decade operational framework that delivers repeatable value creation. The DigitalBridge development model has 3 phases.

Phase 1, we establish platforms. We back great CEOs. We build great companies. We identify and acquire the right platform and the team to capitalize on unique digital infrastructure opportunities. This is about pairing capital and operating expertise with the right strategic business plan around both greenfield development and strategic M&A. You've heard me call it before, this is the build and buy.

We move to the second phase, which is about transforming and scaling. Once we have that platform, we have the right team, we execute operational transformation to improve margins, grow the business and scale it efficiently. Then Phase 3, follow the logos. This is our customer-driven investment framework. We allocate capital and resources to support network growth where our customers are demanding that capacity. We don't build data centers or cell towers or fiber networks on spec, never have. Haven't done it in 32 years, we wouldn't start now. We follow the logos. We go to where Microsoft or Oracle, any of the hyperscalers are telling us they need capacity, and we show up for them with strong intention and execution.

This framework has delivered repeatable value creation for 3 decades, and it's what's driving the results you're seeing at Vantage, DataBank Switch and our other platforms today. We've been applying this playbook consistently, and it works. It's worked for a long time now. Let me take you through several new initiatives and transactions that demonstrate this model in action.

Next slide, please. In September, we announced that GIC and ADIA, both existing Vantage partners are investing $1.6 billion to scale Vantage Asia Pacific platform to 1 gigawatt of capacity. This investment supports the Johor Campus acquisition and broader regional expansion across 5 markets. Let me unpack that for you and why it really matters.

Singapore used to be a traditional data center hub in Southeast Asia, but land, power and regulatory constraints have led to a moratorium at one point. That leads to limited growth. Now, along comes Johor, just across the border in Malaysia has emerged as a natural overflow market, much like we executed that strategy in Reno, Nevada when it became clear to us that Santa Clara had the same type of issues. It offers lower cost, proximity to Singapore, less than 1 millisecond and dark fiber connectivity to the Singapore hub.

From a customer's perspective, it functions almost like an extension of Singapore, but with better economics and available power. This sounds really familiar, doesn't it? This is exactly what we did with Switch and Reno, where we have today over 1.8 gigawatts of compute available for our customers in Reno. We're running that same playbook here in Johor.

The APAC data center market is growing at double-digit growth rates and is expected to reach $77 billion by 2030. 72% of organizations tie their data strategy directly to AI initiatives, which means the demand for data center capacity in the region is only going to accelerate.

Again, using the same playbook, we brought in new leadership last year to position the region for growth. Jeremy Deutsch joined as the President of APAC in October 2024. He previously served as the President of APAC at Equinix, an organization that we have a lot of respect for with over 20 years of operating experience. He expanded Equinix into 5 countries during his tenure and was the inaugural Chair of the Asia Pacific Data Center Association. Jeremy is exactly the kind of world-class operator I love partnering with, and I'm looking forward to building this platform with him. He's doing a great job for us.

The strategic growth drivers here are clear. Singapore spillover demand creates a natural customer base. AI fuel demand is accelerating across the region. We're positioned with the right platform, the right leadership, the right capital partners in GIC and ADIA. The investment is expected to close in the fourth quarter of 2025, and it represents another example of how we're following the logos in the key markets, not only in Johor, but of course, Kuala Lumpur, Melbourne, Sydney, Osaka. These are the growth markets for AI in Asia. Our hyperscale customers are telling us they need capacity in the region, and we're getting ready to set to deliver to that at scale.

Next slide, please. Let me talk about 2 landmark developments that demonstrate the power of our strategic positioning, Vantage's Frontier and Lighthouse mega campuses representing a combined $40 billion investment and over 2.4 gigawatts of GPU compute capacity. What you're seeing here in DigitalBridge backing the build-out of an entirely new generation of compute infrastructure. The AI revolution requires fundamentally different infrastructure than what the cloud demanded. Higher power density, different cooling technologies and most importantly, access to power at giga scale. These campuses represent our response to that transformation.

Both are long-term contracted, pre-leased facilities, not speculative development. We have long-term commitments from Oracle, OpenAI and leading cloud providers. The revenue visibility is exceptional. The returns are attractive, and the strategic importance to our customers is undeniable. This is DigitalBridge enabling the infrastructure backbone for the most advanced AI workloads in the world.

Frontier is in Texas, $25 billion across 1,200 acres in Shackelford County, delivering 1.4 gigawatts of ultra-high-density racks supporting 250 kilowatts and above. Lighthouse up North in Wisconsin represents $15 billion delivering 1 gigawatt with potential for more with the Stargate program distinguished by the development of new renewable capacity, the largest behind-the-meter renewable commitment in the United States today.

Together, these projects create over 9,000 jobs, represents billions in regional economic impact, but more importantly, to our investors, they demonstrate that DigitalBridge has the capital, the expertise, the customer relationships and the power positions to support infrastructure development at scale that very few platforms in the world can match. Construction is underway with the first deliverables beginning in the second half of 2026.

These are mission-critical facilities for some of the most important AI initiatives in the world. They validate our thesis that controlling power and backing world-class operators creates differentiation and creates value in the AI area. We're very excited about these 2 developments, and it really catalyzes a lot of hard work. Congratulations to Sureel and the team at Vantage.

Next slide, please. What do Frontier and Lighthouse mean for DigitalBridge shareholders? Well, let me spell it out because this is where the value creation happens. First, higher fee co-invest. The attractive development economics of these projects enabled us to raise additional co-invest capital at advantaged fee rates. We're deploying and activating that co-invest capital as FEEUM over the next 2-plus years, which means growing fee streams for the projects and for you, our shareholders.

Second, carried interest generation. We're expected to create significant value through carried interest as these developments stabilize over the next 3 to 5 years. Think about the math. We're investing in these projects at development yields. They're going to stabilize at much higher valuation and appreciation flows as carried interest to DBRG and our shareholders over the next few years.

Third, developing our LP base. These projects position our platform to attract institutional capital, specifically targeting AI infrastructure exposure. This broadens our LP base beyond traditional infrastructure allocators to include technology-focused investors, sovereign wealth funds focused on AI and other pools of capital that are specifically interested in the sector that have not yet entered.

Now let me talk about the key strategic considerations because that's what creates our competitive advantage. First, scale advantage. Operating at gigawatt scale generates structural advantages, superior unit economics, access to constrained power and exclusive positioning for multi-gigawatt hyperscale requirements. Our customers cannot get this kind of scale from anybody else.

Second, premium workloads. These campuses are built for the most advanced AI applications. That means higher average pricing in investment-grade hyperscale counterparties. This is the best risk-adjusted business you can do in real estate today.

Third, power differentiation. Distributed power delivery at scale is our critical advantage. Frontier represents the largest behind-the-meter power development in the United States today. That didn't happen by accident. That happened because we've been working on power for years. We talked about in our last earnings call, the power of our partnership with ArcLight, our digital power strategy and what we're doing to ultimately put power back into the grid and baseload.

Together, these $40 billion developments represent watershed investments for Vantage and for DigitalBridge. They deliver unprecedented scale for the build-out of cornerstone AI hubs, serving hyperscale demand, and they demonstrate better than anything I could say in a prepared remark about our power bank strategy is working. I don't need to talk about it. This is the evidence. This is on display.

Next slide, please. Let me close by putting it all together. In terms of our context for our 2025 priorities and where we stand heading into the final quarter of the year. What we delivered year-to-date, we're achieving organic growth with management revenue and fee-related earnings both up 20% year-over-year, even excluding the impact of catch-up fees. We achieved our FEEUM target 1 quarter early, exceeding our 2025 target in the third quarter, and we're tracking to meet or exceed our 2025 metrics on FRE and margins.

We launched new investment strategies and channels, specifically our programmatic private wealth strategy, Franklin Templeton. We've continued to maintain a strong balance sheet and liquidity with over $170 million in corporate cash and a growing asset base. We've delivered breakthrough record leasing across our global data center portfolio with 2.6 gigawatts in the third quarter and $40 billion in new development contracted.

Looking ahead to our year-end priorities for the fourth quarter. Again, we're focused on delivering and exceeding our 2025 financial metrics with FRE achieving or exceeding our guidance. We're formally launching our new digital Energy and stabilized data center strategies, and we're working to secure initial anchor commitments for one or both. We're building on our early private wealth momentum with targeted asset-specific investment opportunities, and we continue to evaluate strategic accretive M&A opportunities centered on adjacent asset managers.

Let me wrap it up with the final thoughts before we go into Q&A. This has been an exceptional quarter for DigitalBridge. What a change a year makes from where we were a year ago in terms of our inability to meet our guidance to where we are today, which is, to be honest, a little bit on the front foot versus our back foot. Yes, we delivered the strong financial results, but this is beyond the numbers. You have to look through what we're doing here.

The quarter really proves out my strategic positioning around power and AI infrastructure and what it's doing is it's creating real substantial differentiated value for our portfolio, their customers and our investor base. Advantage announcements, Frontier and Lighthouse represent over $40 billion of committed developments. These are not speculative projects. These are contracted pre-lease facilities with the world's leading technology companies. They're generating fee streams today. They'll generate carried interest tomorrow, and they will demonstrate that we are having and we've created capabilities that cannot be replicated.

The Franklin Templeton partnership opens up a new distribution channel to Evergreen Capital. The APAC investment positions us to be one of the fastest-growing global data center markets where we can grow in Asia Pacific with our key customers. Look, I've been in this business for 3 decades, and I've seen -- and I've never seen a more compelling structural advantage than controlling power in the AI era. The demand is massive. The supply is constrained, and we're positioned better than anyone in the world to capitalize on this dynamic.

The value that will accrue to our shareholders from this positioning over the next 5 to 10 years will be substantial. I want to preface this by saying, we're literally in the early innings. Let's wrap it up. Super simple quarter, strong financial performance, record leasing driven by our power bank advantage, continued capital formation momentum, strategic expansion of our distribution channels and a clear path to delivering on our long-term value objectives. We're executing our plan. I'm excited about what's ahead, and I look forward to updating you on our continued progress.

With that, let's open up the line for questions. Thank you very much.

[Operator Instructions] Our first question comes from Michael Elias with TD Cowen.

Congrats to you and the team, including Surrel on the massive leasing in the quarter. Great job. In the past, Marc, you've talked about $1.55 a share in carried interest for every gigawatt of data center leasing. Can you just help us understand when in the life cycle of the data center, that unrealized carried interest is recognized? Is it when it's leased? Or is it when it's delivered?

Then also, I'm seeing more of these gigawatt scale deals on the market. I'm curious, as you think of your power bank, how would you describe your ability to take on more of these massive projects?

Yes. Michael, thank you. Let's start with the last part of your question, then we'll come around to the front part. These gigawatt projects are really tough, Michael. I don't think you're going to see tons and tons of them going forward. I think you're going to see more workloads kind of in that 350 to 800 megawatt more bespoke, a bit more tailored. I think building these 1-plus gigawatt campuses are really, really tough. I think it's tough from a capital formation perspective. I think it's tough from a resource perspective. I think a lot of the big gigawatt campuses for LLMs are being delivered now and will be delivered over the next 3 to 5 years. Because remember, these things take about 24 to 48 months to fully build them.

What we are seeing an uptake in is in across all of our portfolio is this what I would call kind of 250-megawatt to sort of 500-megawatt workloads. Our sales funnel has gotten a lot bigger in this quarter. We have over a 7 gigawatt sales funnel right now. When you deliver -- when you lease 2.6 gigawatts in a quarter and your sales funnel grows by 7, you obviously -- you can start putting that math together that there's a lot of big chunky deals sitting in the pipeline.

Our pipeline has gotten bigger. We delivered over 1/3 of the leasing for the industry in terms of actual power delivered against our power bank, what was actually delivered by the industry, we probably delivered about 50% of actual delivered capacity in the quarter. These metrics are important because we keep talking about the power bank and that 21 gigawatts of power. Leasing another 2.6 gig, delivering another 2, I mean, this is hard to do.

Again, when you start -- we started over 10 years ago doing this. We've been able to really keep our pipeline moving and keep our delivery schedules moving, and that's what's giving us massive comparative advantage. I think you'll see other really good quarters from us from a leasing perspective. I just think you're going to see more distributed compute. As we move to inferencing, I think you're going to see -- then you're going to see a whole generation of deals that are going to get done in that kind of 20 to 200-megawatt range, which will be in the sort of secondary markets, which follows what happened in public cloud.

Inferencing will really kick in, in kind of '27 through 2032 as we're still building kind of big LLMs. Nothing kind of slows down. Actually, this quarter, we saw acceleration. Kind of the third time I've told you this year, just when everyone thinks maybe there's a bit of a breather, we're not getting a breather, largely because we have this enormous power bank at these 11 portfolio companies that allows us to keep going.

For example, later last year, earlier this year, we were talking about some of the big deals that Switch was doing. Now we're talking about some of the big deals that Vantage is doing. Next quarter, maybe we're talking about DataBank or Yondr or Scholops. We don't rely on one platform. I think that's what makes us pretty unique and why investors need to own our stock is really simple because you don't have to hang your hat on one story. There's a great -- look, we have enormous respect for DLR and Equinix. They are great businesses, but those are -- you're relying on one management team and you're relying on one pipeline.

When you buy our shares, you got 11 teams Chris crossing the globe, focused on different types of workloads, different types of customer requirements and most importantly, different designs in different locations. I like our bets when you play it with 11 guys on the field versus 1, it kind of gives you a really big leg up. That's why when you're looking at this 21 gigawatts and the 2.6 that we delivered, there's nobody even close in AI data centers to DigitalBridge right now. We're playing the game at just a very different speed and at a very different scale. This quarter really manifested that, and it will continue. I think you'll continue to see that out of us.

Now to your question around carried interest around the $1.55 per gigawatt, how do you realize that carry? To fully realize that carry takes anywhere from 3 years to 5 years. Some of that carry is accrued when you get the entitlements and you get the power. Some of that carry gets accrued when you sign the lease. Some of that carry gets accrued when you deliver the first data hall. Some of that carry gets accrued when you deliver the final data hall. Then some of that carry -- well, all that carry gets realized if it ends up -- if the data center ends up getting purchased, put into a continuation fund, gets acquired as part of a portfolio deal.

Generally speaking, our monetization and our DPI track record in data centers goes back over 4 years ago. We started returning capital back to investors 4, 5 years ago, whether it was our DataBank continuation fund with Swiss Life, whether it was the North American Stabilized Data Center Co, or Valkyrie, the European Stabilized Data Center Co. We've been a consistent returner of capital to our investors, which then triggers carried interest.

Now the problem with DataBank and Vantage was those vehicles were outside of our funds and predated the Colony merger. Now as we get these other companies growing up, whether it's Gala, whether it's Switch, whether it's Vantage Asia Pac, these are now vehicles that sit in our funds. As you know, Michael, roughly about 28% of the carry across our fund products sits with public investors. That episodic nature of carried interest as we deliver these new facilities don't accrue just to the management teams, it now starts accruing to public shareholders.

All this hard work that we're doing now, we're accruing that carry, we're building it, but ultimately, we anticipate in the next 24 to 36 months to start delivering that carried interest out to our shareholders. In the meantime, the math is holding up. I think that's one of the key things that you could say with a straight face today is the arithmetic that we put in front of investors a year ago in terms of what a megawatt means, what a gigawatt means, that flow-through to investors now is really quite clear.

Our next question comes from Jade Rahmani with KBW.

What's your overarching view on how the new data center projects achieve a stabilized capitalization given their size? Do you envision they will be owned long term by a combination perhaps of large REITs, infrastructure funds and hyperscalers themselves? It seems like the digital bridge structuring expertise could provide solutions to that eventuality.

Yes. Thanks. Jade, great to hear from you. Look, we announced last quarter the formation of a strategy called the Data Center Income Fund. Now that strategy is in flight, and we're having a lot of great dialogue with a new set of investors, Jade, that are different from our investors that we had before. Remember, in our flagship funds, we're talking to infrastructure allocators. When we're sitting down on the DCIF product, Jade, we're talking to real estate allocators.

For example, last week was the PREA conference in Boston, our new Head of Capital formation for the DCIF product, winning Price is up there with Ramel Marseille. We had over 60 meetings with real estate investors. Real estate investors are really eager to get allocation, Jade, to these amazing stabilized data centers that we have a deep pipeline in. We're excited about that because for us, it's an entirely new swim lane of capital. Every year, there's about $3 trillion of capital that's allocated to real estate Jade. That's bigger than the $1.7 trillion that's allocated to infrastructure.

If you think about swimming pools, that's actually a bigger swimming pool than we swim in today in terms of general infrastructure. We get kind of excited about that. Real estate allocators today are looking at industrial properties. They're looking at shopping centers. They're looking at downtown office buildings, which, as you know, are kind of -- is a tough asset class. Along comes these 15-year investment-grade data centers, low incremental CapEx going forward. The tenants rarely call you. It's a pretty hands-off real estate product. Most importantly, 95% of the cash flows that we're seeing are investment grade.

This is a really hot new development for us. As we've launched that strategy, and we're really delighted to bring Wendy on the team. We've got [ John Diev ] and Jon Mauck help running the strategy. We've got a big team working on it. It's an entirely new opportunity for us, and it's a new set of investors. It's another way that we're growing our FEEUM, that we're growing our FRE and we're growing our AUM at the same time in a discipline that we know incredibly well.

We have the advantage. We know all the developers out there. We know all the other GPs that need liquidity, and we're pretty excited about it. I think the key to that is we do have the solution. We do have the team. We have the right pipeline of ideas and product, and we've already identified the right set of investors. Everything is lining up to be really successful, similar to kind of what we've done in digital power, Jade, a very similar type of approach, very focused, very surgical and not competing with what we're doing.

As we look to the future, 2026 will be driven by this real estate product, our digital power strategy and private wealth. We've lined up 3 new products for next year. At the same time, as you can see, we'll always be in the market with some sort of co-invest vehicle. Everything is setting up quite well for next year, and we love the product set.

Can you, as a follow-up, give some insight into what fund outflows look like for DBP I, II and InfraBridge perhaps in 2026?

I'm sorry, repeat the question again.

What do fund outflows look like for DBP I, II and InfraBridge? In other words, how much legacy fund runoff should we expect in 2026?

Yes. Thanks, Jade. We don't get into that exactly. We don't give guidance specifically on how we're realizing or monetizing assets. I think as we monetize those assets, we'll report them in our normal cadence. I think, obviously, Fund 1 is starting to near its natural turning point where you begin to think about monetization. We're thinking through that pretty carefully. At the right time and the right speed, we'll do that. Credit is constantly turning over. That turnover is -- loans typically have a 24- to 36-month lifespan. As we're turning over loans and returning capital, we're booking new loans in the second fund and our SMA strategy.

InfraBridge continues to do exactly what we think it should do. In due course, we will continue to monetize assets in InfraBridge I. then as we look forward in the coming years, we'll look at InfraBridge II, but again, as we monetize stuff, we'll share that with you in the quarter. All of those fund products are moving at the speed at which you'd expect them to move. We feel good about the speed and the cadence at which we're delivering DPI.

Our next question comes from Timothy D'Agostino with B. Riley Securities.

On the Franklin Templeton strategic partnership, in the deck, you outlined the $15 trillion opportunity through 2024 -- 2040, sorry, with wealth management allocations to private infrastructure. I was wondering if this strategic partnership is sort of a one-time partnership or if we could see more of these down the road?

Sorry, we lost the beginning of what you said. Could you repeat the beginning of your question?

Yes, sure. Sorry. With wealth management allocations to private infrastructure estimated at $15 trillion through 2040, I was wondering if the Franklin Templeton strategic partnership is sort of like a one-time thing or if we could see more partnerships like this down the road?

Yes. Look, I mean, for this particular product, alongside of Actis and CIP and it's a strategy that they've launched on their platform. We have other partners in private wealth. We did a fantastic product offering last year with Goldman. It was really successful. It was wildly oversubscribed. We do intend to be on other platforms and have other partnerships. That is something that we'll reveal in due course.

We are not exclusive nor limited just to the Franklin Templeton platform. We are working with other allocators, and we agree with your arithmetic. It's a $15 trillion opportunity. It's really big, and there's a lot of great partnerships to be had, and we're excited about it. For right now, we love what we're doing with Jenny and her team. They're fantastic partners. So far, it's been a really successful launch. Great question, and we'll reveal more next year, but right now, my focus is supporting Franklin Templeton and making sure that, that product is wildly successful and oversubscribed.

Then as a quick follow-up, could you provide a little bit more color on the Evergreen capital on the long term, on the long duration and then kind of why the structure is so attractive to you all?

Well, look, I think we have both types of structures across our product set. We don't believe that you have to be all permanent capital nor do we believe you have to have 10-, 11-year closed-end funds. Commingled funds work well for certain allocators, particularly pension funds and other types of sovereign wealth funds, they like that because there's a finite end to what they're doing.

I think other allocators like the open-ended structure, for example, like real estate investors. That's something they're quite familiar with and private wealth clients. They're very familiar with that structure as well. Our continuation funds, the things that we're doing around the real estate asset class, the stuff we're doing around private wealth tend to be open-ended and be permanent in nature. Then our flagship vehicles tend to be closed-end in nature, along with our digital power strategy. That tends to be a strategy that you would lend itself to being closed end.

I think that the great thing about DigitalBridge today is that we're a multi-strat platform. We have multiple strategies really focused on allocators very specifically and then pairing that allocation with the right products, the products that investors want, whether it's a data center platform, a tower platform, a stabilized asset, an energy project that's tethered to the AI economy. These are the things that investors really want today. We have the right products and the right structures, so having that multi-strategy capability is really what differentiates DigitalBridge today.

Our next question comes from Rick Prentiss with Raymond James.

First, to follow-up on Michael's question a little bit. I mean, Marc, you guys as an all asset manager, but focused on digital infrastructure AI. Good to see you communicate numbers, hit numbers, achieve numbers, maybe exceed numbers. The piece that seems to be missing from the stock price is really recognition of carried interest.

You touched on it a little bit to Michael's question, but help us understand the pacing as your portfolio companies now ballpark 50 companies or so, I guess, how should we think about a stable, consistent kind of monetization path into the future to try and get some of this value realized because clearly, AI is hot. power banks are hot, data centers are hot. You're not getting the credit for that. Just trying to think through how do we see that monetization path play out to help the realization of that. I'll come back with a question for Tom on that the bookend impact.

Yes. I'll let Tom handle the bookend piece. I'll just stick the 50,000 feet and then flip it over to him. He gets the hard part. I get to answer the easy part. He's smiling at me right now. Look, I would say, Rick, what I said earlier with Michael, which is it's really important to note that we have certain fund products that have now turned the corner and they're entering into that phase where we begin to monetize.

If you go back to a 2019 vintage fund, where we invested that fund in 2020 and 2021, it's logical to assume that realizations begin happening in '26, '27 and '28. That's really where our first flagship fund sits. Then logically, you can start thinking about how we exit some stuff in Fund II. I'm not going to, on a call, speculate which companies are going to have those realizations, but what I can tell you is we're now in a steady cadence where we have certain portfolio companies going through strategic reviews.

I think that we believe just by the time line and nature of our original legacy flagship funds, you can begin to see a steady unwinding of those funds and return of capital. Along with that return of capital comes the realization of carried interest. Now, Fund II has a little more carried interest for investors. The third flagship fund has a little more carried interest for our investors. I think as time goes on, you're going to see not only more frequency in carried interest, but you'll see more carried interest. We're going to work on that pretty hard next year. I think that will be one of the differentiators about next year versus this year is that you are going to see more realizations next year than this year.

The other thing that I would say is on our data center business across the 11 platforms, we have been Rick, pretty creative around creating DPI and creating carried interest. When you've got great vehicles that are in permanent capital vehicles like DataBank or now Manage North America and some of these -- some of the growth metrics around some of these other businesses, we have to recycle capital, return capital, and when we do that, that also triggers carried interest as well.

Data centers is an area where we think we can do quite well. I think stay tuned, but we do believe that Rick, it's reasonable to assume that given the age and vintage of some of these fund products that you can begin to count on carried interest instead of being episodic.

I don't know, Tom, if you want to add anything to that.

No. I mean, you really covered it. The only -- I guess the only thing I'd add is sort of on the margins, if you had sort of backed up 4 years ago and projected a bell curve of where we would be distribution-wise, I would say probably over the last few years, there's been a more significant and extraordinary opportunity to continue investing in some of our portfolio companies than maybe you would have thought 4 or 5 years ago. That may have pushed out exits that you -- several years ago would have thought were happening in '25. Maybe we've continued to invest in some of those companies because the opportunity there is extraordinary. I wouldn't add anything different than what Marc said in terms of the bell curve and the vintages.

The way the book works, I think you touched on that in your prepared remarks a little bit about the J curve and where you're at on it because it does confuse people. I think sometimes when they look at the book carrying value, you're not marketing it to market on transactions that are occurring or things in the marketplace. You're literally looking at your asset-by-asset. There some people, I think, look at it and go, well, how can the book doesn't show more? Maybe just elaborate a little bit more on that.

Yes. Look, I think there's sort of 2 main components to the performance and how we mark things. Marc talked a bit about kind of our investment playbook, making sure we make the right investments that we are focused on the customers, deliver for the customers, drive operating performance of the companies. That is one significant contributor to the performance. You've seen some of that to date.

Then I would say the second material contributor to performance and accrued carry or realized carry is the exit process and selling the companies well. That's where you capture kind of the second stage of value. I would say, we're sort of in between those 2 kind of step functions right now where we've achieved real value and real performance at the company level. We will capture the second step of value as we create distributions and liquidity and realize the full value of the companies, if that makes sense.

Does. You're not marking to what transactions might be, you're not marking to comps until you actually achieve a sale basically?

Look, if there's -- you think the valuations and performance are 100% quantitative and there's always a bit of qualitative to them. We do our best to mark to where we think -- what we think the asset is worth, but there's always a bit of uncertainty and how much of that will you actually realize in a sale. You're always including some sort of contingency to address that.

Marc, one of the interesting things we saw this quarter was the story that Elon Musk is now following you. Can you give us a little background color? Did that happen? How did it happen? What's your kind of relationship with Elon?

Yes. Thanks, Rick. I don't comment about social media stuff. It's just kind of, unfortunately, it becomes like water cooler banter. What I will say is, I've got a lot of respect for him. Yes, I do know him, not particularly well, but we know each other, and he's an important customer. He's a very important customer. He uses a lot of our different portfolio companies, and we serve Starlink, we serve SpaceX. We serve Tesla, we serve xAI, and we use his batteries at some of our data centers for storage. It's a really super important customer that we think can get bigger over time and that can be a lot more strategic.

Anything we can do to work with him and support what he's doing, I'm very supportive of him, and I'm always happy to put capital into his businesses through infrastructure, we're side-by-side with him. A lot of deep respect there, and I think he's one of the great minds of our generation. If he chose to follow me, that's great, but I'm super focused on servicing him and trying to figure out how we can help enable his businesses to go faster.

Our next question comes from Richard Choe with JPMorgan.

I know you spent a decent amount of time on it, but I think it's important to kind of go through it a little bit more. With Vantage in these -- the Frontier campus and Lighthouse campuses, these are 2 really big deals. Over the past few months, we've seen a lot of companies come out and say that they have a lot of power and that they have a lot of capacity deliverable, but you've actually won these leases. Can you take us through the process a little bit? What gave DigitalBridge and Vantage the, I guess, advantage over the other companies and winning this deal?

Yes. Look, I think that we continue to believe that this industry will be unfortunately marked by a lot of amateur and a lot of tourists in the next 24 to 36 months. People that do certainly know how to buy land and how to get entitlements and certainly have to reflect that they have some source of power.

I think it's a different level up when you actually build a data center and you're responsible for delivering something at a Tier 3, Tier 4, Tier 5 standard and that you've done it for a decade-plus. Customers know the difference between people that are new to the sector and someone that's a trusted set of hands that has over 400 data centers and 11 different companies. Differentiation isn't about press releases. Differentiation is about execution and the ability to show up for a customer and deliver on time.

Again, we don't get a lot of credit in our share price for being around for 30-plus years as executors. What we have to do is we have to go out and we have to deliver a quarter like this. We have to deliver a quarter where we clearly demonstrate leasing volumes that are differentiated, a power bank that's differentiated, but most importantly, to the metrics, Richard, that you now know that we're judged on, which is FRE, FEEUM, distributable earnings, AUM, all of the things that we're being judged on. We know who our peer set is now. We're an alternative asset manager.

The most important thing that we now need to do is just like we've been executing for customers, we now need to execute for our public shareholders. The framework and the prism that Tom and I are judged on is by other alternative asset managers. When I said in the earnings call, what a difference a year makes, we had a very tough last year third quarter. Tom and I thought long and hard about it. What we tried to do this year was deliver something that is -- that we can -- investors know they can count on us and that we're credible around the numbers that really matter.

I think execution is critical. Executing for our public shareholders is critical, executing for our customers like Oracle and OpenAI is important, and we're executing for hundreds of other customers this quarter. We just don't have enough time or pages in a deck, Tom, to sort of share all of those wins for all those customers, but we're delivering dark fiber routes. We're delivering towers. We're delivering small cell infrastructure. We're delivering WiFi offload. There are so many things that we're delivering right now that we don't have time to talk about. We just got to keep delivering for our logos. That's really what is differentiating about us right now.

I think at the end of the day, customers not only vote with their wallet, but they vote with their -- the integrity of their network. I think one thing that we've proven to be for a long time is a very trusted set of hands. Hopefully, that will work out well. I think that a few people are actually turning on capacity right now, and we are. We're turning on capacity, and that's because we started planning this 8, 10 years ago. We didn't start -- we didn't say a year ago an investment committee because we thought it was a hot idea to get into data centers. That's not how we're built. We're built differently. We'll be here. We'll be here tomorrow. We'll be here in 10 years. We were here 20 years ago.

I think it's that consistency is what customers really like. When it gets down to choosing, having great teams like Sureel, who knows how to deliver for a customer, that's what it is. We're very fortunate to have Surrel and Jeff Tench and Dana Adams, that entire team is just a bunch of pros. Adults that have been there, have been doing it for a long time, and that is, thankfully, Surrel is our partner. We're fortunate to have a great management team that was able to deliver for the customer.

You mentioned it a little bit, but will these projects be meaningfully contributing to FEEUM early in '26? Or is it more later in '26 and '27?

No, 100%. It will be a big contributor in 2 years. We took in a lot of new capital, co-invest capital specifically to these 2 projects. We get paid on that capital on that committed capital. There will be an immediate impact when you have really good co-invest. Look, Tom said it earlier in the year, we've got to improve our margins. Both Tom and I committed to that. Tom has been working on the cost. I've been working on making sure we get our fees up and co-invest. I think we can look at each other and say we've delivered. Tom has done a great job on the cost side. We've done a great job in improving our margins on co-invest.

When you put those 2 things together, you get a result like this in the third quarter, which is improved margins, incredible year-over-year growth, as you can see. I mean, just looking at fee-related earnings, were up 43% year-over-year. I don't think there's another publicly traded alternative asset manager that's up 43% year-over-year. This has been a lot of hard work, and we're not done. I think there's a sharp focus on what we got to keep doing. The fees build as we build. As we keep leasing megawatts and we keep deploying capital, so does our FEEUM raise go up and so does our FRE go up. We're heading into a historically strong fourth quarter. I think Tom and I are excited to continue to work hard between now and the end of the year.

I don't know, Tom, if you have any voice over on that.

No, I think you covered it all.

Our next question comes from Eric Luebchow with Wells Fargo.

Marc, I'm curious about your data center power bank that you've been highlighting in the last few quarters. Maybe you could talk about the split you're seeing between kind of behind-the-meter power solutions versus more direct grid-connected power and perhaps how that behind-the-meter opportunity, which seems to be much bigger today, kind of ties into the size and scope of the energy fund you're fundraising for.

Look, the energy strategy is super important, and it's a big part of what we're doing in the back half of this year, but more importantly, Eric, what we're doing next year. We've already closed a couple of deals in that strategy. Takanock is one of them where we provide digital power solutions for our customers and to other data center developers. We've got another project where we're tethering some -- a specific power solution to an existing DigitalBridge data center where we're adding 500 megawatts of power.

What I love about what we're doing in digital power, Eric, is these aren't ideas. These are very, very focused solutions in very specific locations tethered to very specific outcomes for customers. That's what makes it so unique is that a lot like what we're doing in the data center space, we're doing in the power space. We're not taking risk. We're entering into long-term contracts with counterparties, investment-grade counterparties. We know that the offtake is in place.

I think what's interesting, Eric, is as we've been doing this for about 2 years, we've also learned that the grid doesn't go away. You have to learn how to use the grid, you got to learn how to work with the grid. You've got to have battery storage capabilities inside your microgrids. You've got to be able to build up that power during the day, sell some of that into the grid, so you're basically an offtaker into the grid and putting power back into the baseload on.

Then during the nighttime when baseload is more available, you can buy back from the grid. It's a very fluid relationship in a microgrid. The key to that is interconnection and being interconnected into the -- that state's Public Utility Commission grid and having an active relationship with the utility. What we're doing is not in contrast or in competition with our utility partners in each state. In fact, we are a trading partner with those utilities. A lot of like what we do in interconnection and fiber, we're doing the same thing actually in power. It's a really interesting business model, and we've had a lot of good early success so far.

Someone once told me, I think it was -- I was listening to another alternative asset manager, I think it was Brookfield that said it, but there's $7 trillion in AI, there's a $7 trillion CapEx AI spend. If you think about the power that's required, the incremental power to deliver 300 gigawatts, Eric, there's another $1.3 trillion in power to be built. We look at that, again, that swimming pool is $1.3 trillion. We look at the opportunity there, investors are really excited about it. I've been on the road talking about digital power, talking about microgrids and how we're delivering these power solutions to customers. As you can probably imagine, LPs get really excited.

If you look at the fundraising this year in infrastructure, Eric, infrastructure is having a record year in fundraising. I think there'll be over $200 billion of capital raised in infrastructure. -- if you take a look at that micro [Technical Difficulty].

Sorry about that. Technical glitch. Anyway, when you look at the total $200-plus billion of fundraising and infrastructure this year, over 50% of that is to energy transition, not data centers, not digital. It's really interesting to me that allocators are really focused on this issue of power. We really see that as a big opportunity because if we think about building large-scale campuses and if you're going to spend $11 million to $12 million per megawatt building a data center, you could end up spending $3 million to $5 million in grid independent power to that data center.

We really look at this as a $0.30 to $0.50 incremental spend on power because we're taking that risk and building the data center. We know how to build our microgrid infrastructure. We know how to source LNG, best solar, wind, hydro and most importantly, use the grid. I think everyone gets obsessed a little bit, Eric, with, oh, you're either on the grid or you're off the grid. It's not that way actually. It's that if you're going to really build a scale and you look, for example, at the last 2 big projects that we've done, we do use the grid. At certain points of the day, we're putting power back into the grid. Then at certain points of the day, we're taking power from the grid. It's a fairly dynamic and fluid relationship between a microgrid and the actual grid itself.

I think other people are finally figuring out what we're doing. It makes a lot of sense. We've got great industrial partners. We've got a big pipeline of projects that we're building. We're excited to talk more about it next year. I mean it will be a big part of what we're doing in 2026. As I said, 3 new strategies for next year, and that's -- digital power is one of those strategies. When you and I get together next time in person, Eric, we'll do a little bit of a digital power teaching whenever you're ready.

That would be great. I appreciate it. I guess just one follow-up, Marc. Some of these mega deals we've seen, including the ones that Vantage did, obviously, there's a lot of kind of newer hyperscale tech like LLMs that are directly or indirectly backing some of these new builds. How do you think about the pricing, the lease terms, the credit risk of some of these newer players that are obviously not profitable today, but growing incredibly fast. There's certainly been a lot of industry chatter about some of the LLMs, their ability to pay for some of the future commitments they've made. Just curious how you guys think about it at the infrastructure level.

Yes. Look, I think that the -- there's different types of LLMs, Eric, right? There's different types of quality of credit tenants. We look at some of the NeoCloud business models, and we've chosen not to put our equity capital to work there. We've been very selective about those types of credits. I think, again, when you've got a really substantial power bank and you've got on-demand capacity ready to go, it does allow you to be a little selective about what we can do.

I think it's pretty unique, our ability to choose customers we want to work with. I do get a little worried about some of the credit profile risk around NeoCloud versus AI providers versus the hyperscalers. We've been very cautious about not overweighting one particular customer or one particular story. I think what's unique about our platform is given the scale of our platform, we're invested across all of these customers and all across these workloads so that we're not beholden to one customer or one technology or one LLM.

I think that's where scale matters, Eric. We talk a lot about scale in the alternative asset management space. Well, in our specific swim lane, we have a lot of scale, and we have a lot of customers. That diversity in customers is what's really important. We've been able to demonstrate that we can lease to a lot of different logos at the same time. I think at the end of the day, having a diverse set of cash flows and a diverse set of customers and a diverse set of data centers is really where you want to put your capital today.

Again, same thing we talked about, buying our stock is a proxy for that, right? You're not making one bet on one specific data center platform, you're making a bet on a global portfolio of 11 platforms, 21 gigawatts and record leasing. That, that flow-through will come through in a function of FRE, FEEUM and carried interest.

There are no further questions at this time. I would now like to turn the floor back over to Marc Ganzi for closing comments.

Well, thank you. I appreciate all the thoughtful questions from the analyst community. We look forward to engaging with all of you over the next couple of days to bring more clarity to the quarter. Let me finish in thanking our team. We have an incredible team.

Again, I want to bring focus back to third quarter last year against third quarter this year. Specifically, I want to thank my CFO, Tom Mayrhofer, for delivering a very clean quarter and delivering on the promises that we made to you, our public investors. Again, fee revenue up 22% FRE up 43%, distributable earnings up 102% and fee equity under management, FEEUM up 19%. This was a clean quarter. This was a quarter that we -- to be candid, we felt capable of delivering, and we owe it to our investors to deliver results like this in a quarter like this.

We have strong liquidity. We're allocating capital. We're raising money. We've got a great suite of new products that will come to market with here in '26, and we're well-positioned to be the leader in digital infrastructure and the power that's required to fuel it. Really looking forward to the end of this year and looking forward to catching up with all of you on the road as we hit different investor conferences. Please follow up with Severin and our team to get access to the team. We're always happy to have a conversation with you. Thank you for your continued interest and your ownership in DigitalBridge shares. We appreciate it. Have a great day.

This concludes today's teleconference. You may disconnect your lines at this time. Thank you for your participation.

Good morning, everyone, and welcome to Day 4 the Goldman Sachs Communacopia and Technology Conference. I have the privilege of introducing Marc Ganzi, CEO of DigitalBridge.

My name is Josh Frantz, and I cover telecom and communications infrastructure here at Goldman. Thanks for being here this morning.

Thank you, Josh. Good to be here.

Marc, you've been in this industry as long as anyone and your portfolio companies touch basically the whole set of infrastructure globally and so, I would imagine you have a pretty good view of what's happening on the ground. And we'll go into each of the different subsegments in a bit. But from your vantage point and looking out for 10 years. How do you see the whole stack of infrastructure evolving? What do Fiber networks look like? What do wireless and tower networks look like? And how does AI change how the data center market looks?

Well, I think, look, the ecosystem has never been more challenged. I think we have 57 companies around the world that go out and compete for business every day. And this will be far and away the biggest year in terms of CapEx deployment across our ecosystem. And as the year has gone on, those numbers have crept up, Josh, as we've gotten more bookings more orders for towers, more orders for redundant dark fiber routes. And then we'll get to data centers. We'll save the best for last.

But we're sort of drinking through a firehose. And to do that requires enormous coordination. It requires a lot of supervision. But I would say there are 2 things that you can't live without right now. One is capital and 1 is power. And so that unique combination of being able to raise capital and deploy it in an efficient way and then at the same time, be able to hopefully predict, where you're going to need power and then ultimately, either procuring that power or as we've talked about in our last quarterly conference call, where we're building that power ourselves. And we call that sort of digital power grid independent solutions. Some people call it behind the meter, in front of the meter, I don't know where it is on the meter. But ultimately, you do connect to the meter at some point.

So I spent most of my time today. Someone asked me yesterday, like how are -- how am I spending my time? I'd say I'm probably spending 50% of my time on power. I'm spending 40% of my time on raising capital. And then the other 10% of the time I got to run the business, which means I need sort of a clone of me to do that part. But it is important, though. I'm not sort of making light of it.

There's a lot to do. And I think the ability to triage between what's real and what's not real, is really important right now because we get a lot of customer inquiries across all the verticals, right, mobile infrastructure, fiber infrastructure and data center infrastructure.

And part of our job is also, Josh, discerning which customers are real because you have a lot of window shopping, right? You have customers saying, "Oh, I really like that location because you've got that power." And then that sort of that moment of liking a location to ultimately signing a lease, that takes about 9 months, right? And so, we invest a lot of time and energy with our customers. At the same time, those conversations for sites aren't just a singular site conversation. There's multiple customers looking at the same sites, we have to make decisions.

Now thankfully, we've got a powerbank of 22 gigawatts, which allows us to be somewhat discerning and pick who we want to use -- but at that same time, you've got a bunch of other implications, which is how do you keep buying land? How do you keep identifying power? And then how do you pick the right sites to build the power at.

So there's just a lot going on there. I'll come back to fiber in a second. But on the tower side, AI is absolutely impacting towers, and it's impacting the demand for towers. I think, if you listen very carefully to the subtext of SBA, Crown and American and then when Vertical Bridge talks publicly, the leasing demand that we're seeing in towers today is probably the best leasing we've seen, since 2013.

So sort of unpack that a little bit, what gives us this optimism around tower infrastructure, mobile infrastructure. There's 3 things happening in mobility today that I think are -- that require careful thought. One is mobile data traffic is up materially. And if you were listening in the second quarter to what Jensen said and what Masa said in the first quarter, mobile data traffic will move up somewhere between 3 and 5x.

And I think that's interesting because the last time we saw a really big step function in mobile data traffic was when you -- when the public cloud came to mobile devices, and we were starting to put active compute, which is applications on devices you saw mobile data traffic from 2010 to 2016 go up 10x.

And so what you also saw at that same time period was a massive investment in network infrastructure. It was the last time you really saw the tower companies growing at double-digit organic growth. We're seeing that same type of cycle occur again, except instead of sort of cloud-based applications, we're seeing inferencing. And so, as AI moves from LLMs to generative Ai to inferencing, where do we think inferencing happens? It happens here.

80% of AI is going to happen in a mobile environment. Now I'm not saying that 80% is all here on a mobile device. But what I would -- I would offer to people listening today is that we're at 30 billion connected wireless devices today. By 2033, we go to 60 billion wireless devices. Now what's a wireless device. It could be a wireless meter reading, it could be your fitness band, it could be a mobile device, it could be a phone, it could be a car.

Think about all the connections that are happening and then compound that in an environment, where 90% of those connections are machine to machine. So the fastest-growing area of data consumption in AI is not enterprise. It's not data sovereignty, it's not consumer, it's machine to machine learning. That's where you're going to see most of the activity. And all of that machine-to-machine activity is going to happen, predominantly 90% will be wireless.

Look at a factory. So you go to a General Motors factory today, how many wireless devices, how many wireless sensors are along that production line at GM. And so, if you go to their Lansing production plan, which we actually had a very interesting deployment by ExteNet there, which is one of our mobile infrastructure companies, where we're doing private 5G networks for GM.

Along that production line, they have over 300 different wireless devices, communicating to each other without a human being involved. And so that requires a node or an antenna or some form of connection. And all of that is running in a private 5G environment. This is just 1 example of machine-to-machine connectivity and how wireless is being used to proliferate that activity.

So I'm pretty bullish on mobile infrastructure right now. And so 1 is machine-to-machine connectivity that's driving that growth. 2 is inferencing to the wireless device. And then the 3 things you probably picked your head up this week is you saw satellite spectrum being bought by wireless companies. Why? We don't have spectrum. The FCC doesn't have a plan to issue new spectrum right now. And so what are the carriers doing? They're looking at what's available, which is the [ Sat ] spectrum.

So that introduction of spectrum and inferencing and the amount of devices that are coming online, 30 billion new devices, what are we going to do? Well, what do we do is we build infill. The carriers hate it when you say densification. So we'll call it infill for the time being. But the reality is what you did see in this quarter from all of the tower companies is that it wasn't amendment traffic that was leading leasing. It was Colos. Colos are back.

And so for the first time, since 2013, at least in our private business, we saw de novo Colo applications and de novo leases, which are Colos, outpace amendments. And that's important because, as you know, a Colo today is anywhere from 2,000 to 2,500 a month. An amendment is on average between $250 and $500 a month. You got to do 5 amendments to equal 1 colo. I'd rather do 1 colo than 5 amendments, right?

And at the same time, new construction is way up. So our domestic U.S. REIT vertical bridge has a pipeline of about 5,000 build-to-suit towers. And so last year, we delivered about 800 towers. This year, we'll deliver 1,000. Last year, American Crown and SBA combined delivered 180 towers. So our private little tower company is delivering 4x more BTS than the public guys.

So we're pretty bullish about what's happening in towers. And I think towers are going to come back in a material way, I think, if interest rates get cut, I guess what happens to tower stocks. It's pretty easy to figure it out. So -- and it's not just at Vertical Bridge, we own 10 other tower companies around the world in Asia and in Europe and in Latin America. So and we're seeing this across all of our tower businesses. We're seeing a surge in demand.

So that's kind of the first category. And look, across ExteNet, Boingo and FreshWave, we own 3 small cell providers. FreshWave serves the U.K. Boingo serves WiFi and private networking and ExteNet serves the carriers. And all 3 of those companies are recording a massive uptick in their pipelines and bookings. So I think mobile infrastructure, which generally has been falling out of favor because of what's happening in data center land.

I think, if investors are paying attention, they'll see the towers and small cell infrastructure is a good place to go right now. Fiber, it's been a great year for Fiber so far. We look at fiber potentially from a different prism than the rest of other people do. I think the Fiber business today is 4 different businesses, and I think you have to decide where you're going to put your capital.

We think there's 2 businesses that focus on resi and 2 businesses that focus on commercial. And on the residential side, consumer-facing businesses, where we are connecting homes, those businesses or businesses we're investing in. We announced an investment in a while, WideOpenWest. We're going to put a bunch of capital in that business. We're going to refresh the network. We're going to groom the network and give it the necessary capital to go grow it. We're excited to close that, work with Crestview on that. And there we just -- we were opportunistic. We bought it at the right price. We like it when we can buy business as a percentage of replacement cost. That's how we think about it.

So I'm constantly thinking about what am I buying that business per home pass per home connection and then per route mile. Those are sort of 3 metrics we look at in resi Fiber. At the beginning of the year, we bought an MDU business. That's the second vertical inside the Fiber business. So an MDU business is where you're a wholesale provider. There, we have a business called FiberNow that operates in the Southeast U.S.

All we do is supply Internet and cable service to multi-dwelling units. So high-rise condos, HOAs, planned communities. We enter into a 10-year contract with that HOA and then we provide the Internet service, the broadband connectivity to those communities. So what I like about that business, you probably saw Brookfield bought Hotwire, which was like a 40x multiple, which was fantastic. We get excited about those multiples and Brookfield pays that.

But what I think Hotwire, what Brookfield saw in Hotwire is they saw thousands of communities, where they had a 10-year exclusive. And so when you light up a community or you light up a high-rise building, you have a 100% penetration day 1. because the broadband fee is included in the HOA fee. So if you're in a Hotwire building, you don't have a choice, you have hot wire. You literally plug in your router and you're active. The same thing in Fibernow, which is the sort of challenger brand, the Hotwire in the Southeast.

So we bought a really good business called OpticalTel, put a bunch of capital in, refresh the management, put some growth CapEx in and now we're growing that business and competing in the MDU space. And we also do that down in Latin America. We've got a business called Mundo Pacifico that also provides wholesale broadband fiber for the carriers and for MDUs. So it's a good space. It's a space I think that telcos value because it's really hard to run around and sign up those HOAs.

But what I think you're seeing in the multiple differentiation you see a business like WOW trading for 5x, then you see a business like Hotwire trading for 40 times, you say, that's 35 turns of value. What am I missing? And I think it's not what you're missing. I think people believe that Hotwire and FiberNow have long-term contracts and you don't have to run around signing up the homes. You get that 100% penetration day 1.

On the wholesale, on the commercial side of fiber, we have Zayo performing really well for us. It's been a great turnaround story we sat here 5 years ago at this conference and said there's 5 things we got to do to fix Zayo, when we took it private. And we've now completed all 5 of those things. Now the next phase is how do we keep that double-digit organic growth going at Zayo.

It took a long time to get the sales engine going and get the growth going. And really, today, Zayo is the Tele2 cities, right? It's 2 businesses. 1 is an enterprise business. We're connected to close to 60,000 buildings -- we're excited to close Crown and integrate all those buildings. And then the other side of the business is our wholesale transport business. So that's a great business.

Our average contract there is over 20 years in duration. That's really our long-haul network. Our transport network, sub Oceana cables, our metro rings and the laterals that connect data centers and the laterals that connect small cells and towers. So that business, really good business, greater than 80% investment grade. Long-term contracts, and that's infrastructure. And I think, if you're thinking about fiber in the commercial sense, you really need to think about how you divide it.

So on the enterprise side, we compete with Cogent. And on the long-haul transport side, we compete against Level 3, which is embedded inside [ illumin ]. And we really like that long-haul transport business. If I could sort of -- if you had to rank fiber from sort of 1 to 4 I would go that business, which is transport. I would take residential MDU, then I would go enterprise fiber in a commercial sense and then I'll go residential fiber. That's how I rank those.

And look, we're investing across all 4 we think broadband connectivity and providing that -- those capillaries that connect data centers and towers and small cells is still a great business, and we'll see how the rest of the year turns out for Zayo. But that's mobile infrastructure, that's broadband infrastructure, and that leaves the last 20 minutes for what you really want to talk about, which is data centers.

Sure. Before we get there, the 1 thing that I think investors in DigitalBridge stock kind of think about is you have this strategic vision, you have the underlying trends that you talked about and all these different subsegments -- so when does AI work at the portfolio level and when does that manifest into kind of DigitalBridge's earnings?

Well, I think the direct flow-through for the DigitalBridge investors is a really simple algorithm, which is we said it in our last earnings deck, which is for every megawatt that we're lighting and building we're on average, we're spending about $10 million per megawatt. And I would say those costs have crept up into $11 million to $12 million per megawatt.

And the way we think about that is in a normalized world, we're generally spotting 60% leverage and 40% equity. And so again, if I take that $10 million, I've got to go raise $4 million of equity to go build that 1 megawatt of power. Now historically, over the last 12 years, since the inception of DigitalBridge, if you look at our realizations in data centers, whether it was DataBank or Vantage or any of the $9 billion of DPI we've returned to investors, we're generally returning that capital north of a 2x MOIC.

Now why is 2x MOIC important? 2x MOIC is important because every dollar I raise, I generally have a performance hurdle around that 2 point MOIC range. And so if I'm raising $4 billion of equity capital to build 1 megawatt, what's really interesting about that is I can turn that $4 million of equity into at least $8 million. And if I turn it into $8 million on that $4 million what's really interesting is we're earning a 20% performance fee on that $4 million of profit that we create.

And so that sets up really nicely because if you're really thinking about that $4 million of profit, that's $800,000 of carry that's being generated back to DigitalBridge. And so that's the flow-through, right? And ultimately, if you looked at our last earnings deck, 28% of our carry goes to our public shareholders. We send that back to our shareholders.

And so we think it's really interesting because our portfolio has been growing. We highlighted that we'll be close to 6 gigawatts of capacity shortly. And we've got a power bank of 22 gigawatts. We'll lease through that 22 gigawatts in the next 3 years. And so again, that's 22 gigawatts, which is 2,200 megawatts.

And if you start running the carry math on that, what's happening at DigitalBridge is we're accruing a massive amount of carry. And on that $40 billion of equity we've raised that's going to work for our investors, what's the missing link in the DigitalBridge story for investors of going from a $12 share price back to '23 is fully valuing the embedded nature of what's in the portfolio.

And so we have to go out and make that argument. We have to sort of spoon feed it. We got to show investors what we're doing. And then at the same time, right now, we're deploying $50 billion of CapEx. It's the most amount of CapEx we've ever deployed. We announced this morning, we raised another $1.6 billion of equity at Vantage Asia.

We announced a couple of weeks ago, a huge project we're doing in Lancaster. We raised effectively another $1.7 billion of equity, we announced a couple of months ago, Yonder, where we raised $1.6 billion of co-invest on top of our fund commitment. So we keep raising equity, we keep putting it to work. But most importantly, when you put capital to work, we're tethering it to a lease which is a megawatt.

And the faster that investors can correlate that 1 megawatt to, whether it's Microsoft, Amazon, Oracle, Google, Meta, whoever it is, [ Korev ], NVIDIA, we do business with the top 12 hyperscalers in the world. And just on a sheer volume, if you took all 10 of our portfolio companies that 22 gigawatts, just look at the Power Bank of Digital Realty and Equinix, right? Equinix is about 6 gigawatts, DLRs at 8, we're at 22.

So 14 gigawatts with the 2 biggest digital REITs in the world that own data centers against our 22. So we think we're playing at scale. We've been we've had this plan in place for 8 years. It's all starting to manifest itself now. And so we find ourselves in a really lucky position because, as you know, people are struggling for power and people are struggling for real estate.

And so we're sitting back saying, well, look, we don't come to the fight with 1 company. I don't just own QTS, right? I own 10 QTSs. And so -- and I'm fighting the fights locally. And further to that, we're also saying, 1 data center company doesn't address the opportunity. You have to have an edge business, you have to have a hyperscale business, and you cannot go to sleep on private cloud as we've shown a switch. What we've done with Switch in the last 2.5 years has been incredible.

And so we're segmenting the market by geography, by workload, and we're showing up with a lot of capital, a lot of capabilities and a lot of power. We're doing it in a way that nobody else is doing it yet. The market is starting to appreciate it. It will appreciate it when we start returning capital, and we turn those megawatts into carried interest, and then that turns into EPS.

And so that's our architecture. And maybe it requires a little patience because we've been building this for 8 years. And I think the real sort of -- this year, it's starting to turn for us. You're seeing it in our earnings. We've had 3 very consistent quarters, where we beat our FRE metrics. We plan to do that for the rest of the year. We plan to do it next year.

And ultimately, the metric that we're going to be measured by is really 2 things: 1, our fee-related earnings. And so as we raise capital like we did today for Vantage Asia, like we did for Vantage North America and for Yonder, investors are looking at, okay, I like the carry. I got to wait 4 or 5 years to get to the carry, but what's going to feed me in the time being, FRE fee-related earnings. So when we do raise that capital and we put that capital to work, it has a fee associated to it. So we get fee and we get carried interest.

And I've always said this when we did the merger with Colony, when we became a REIT and then we de-rated I said, look, the only way you're going to play this game and the way that we're going to be at American Tower, Digital Realty and Equinix and Crown is you've got to go asset-light. If you're going to be in the digital infrastructure business, capital is everything. And we couldn't stay a REIT. We tried to stay REIT, but ultimately, to go build 1 gigawatt of capacity is hard, right?

You got to show up with a lot of money and you got to show up with a lot of capabilities. And the reason DLR and Equinix have missed all these big bookings is because they don't have the capital, they don't have 10 different platforms, 10 different CEOs, a big land bank and a big power bank.

And so the way that we have surpassed these guys is that we made the decision 6 years ago, whether it was popular or unpopular, we dereadit, we win asset light. We had a series of silos and funds, where we could raise capital very quickly and ultimately, we felt the free cash flow conversion in fee and carry was a better bet than just only being able to deploy, I mean, Andy is going to have a great year at Digital Realty. I love Andy, and I love what he's doing, but he's going to deploy a couple of billion dollars of capital.

I'm deploying $50 million, so the quantum at which we're playing and the level which we're playing is just a step function higher than the other digital REITs. So we've really created some differentiation. The only problem is I went from this sandbox, which was competing against my friends that run DLR and American Tower and now we're in the alt space. And so now people look at us and say, okay, you're essentially the Blackstone of digital infrastructure, right?

And so I'm like, okay, fine, I'll accept that. The only problem is they manage $1 trillion in assets, and we manage $106 billion of assets. So -- now my competition isn't so much whether I can win the bookings or I can win the power. We're winning the bookings. We're winning the power battle. Now the real challenge is how do we match up against Blackstone, Brookfield, Ares, Apollo, KKR -- and there, we've got to go out and we've got to compete for capital, and we're doing a great job. I think for our size, we punch way above our weight class. I was just with an investor, and they said, well, how do you do it? How can you show up and beat someone like Blackstone on a site like Lancaster.

And I said it's really simple. I think investors, private capital investors really appreciate specialization. And I think when we sit with an LP, whether it's the government of Singapore, whether it's Abu Dhabi's investment authority, the ability at which we speak about this asset class is at a very, very intrinsic level. We understand land use, we understand zoning. We understand will serve letters. We understand how to build a microgrid. I can negotiate a master lease agreement against Oracle, Microsoft and Amazon.

These are the things that candidly, Blackstone can't do. They can outsource it to a portfolio company and then go do it, but we do this at a very, very fundamental ground level. Now at the same time, I don't have private wealth distribution. I don't have a BREIT. I don't have 200 salespeople running around selling the 30-year track record to Blackstone, which is great. So we got to earn that.

And so we're -- the stock is in transition because we move from being a digital REIT into this world of the financial alt world and I went from being the biggest and digital infrastructure to the smallest in alts. So it's a real challenge, and I think I'll wake up every day, and we're tackling that, and we've done a good job, I think, conveying our value proposition to people that own financials.

And as much as I love coming to conferences like this, which is really what I've been doing for 30 years, I now have to spend half my time going to financial conference and explaining FRE and carried interest in EPS and all the things that we're doing. It's a challenge, but it's fun.

Got it. To the AI point, training has been what's been driving a lot of the leasing in the past few years. It feels like inferencing is starting to take over a bit and I think there's still a lot of companies that are trying to figure out what these enterprise applications are and how this actually manifests at the end of the day. From your seat, what are you seeing? Where do you think we are in the transition from training to inference and -- and as you have discussions with enterprises, like how do you think this all kind of shakes out?

I think we're really nascent in inferencing. I think, if we were playing kind of a 9-inning baseball game, we'd be in the top of the first. And I think from a training perspective and from an LLM perspective, we're kind of in the third inning. And if you track the CapEx and you track the amount of power that's been lit, -- we're -- today, the industry is at about 68 gigawatts. That's up from 54 gigawatts last year. There's a couple of different curves and growth in terms of how you look at it.

Some people think we'll get to about 137 gigawatts. Our forecast is 196. The Jensen forecast and Masa forecast is like anyone is like 300, of course. Somewhere in between is probably the truth, which I think hopefully is our forecast, which is 200 gigawatts I was in L.A. yesterday speaking at the McKinsey Global initiative, which is, candidly, the whole conference was took over by the talk of power.

So for 2 days, top CEOs in the U.S. from the power industry, from the infrastructure industry, all came together, and I think it just turned into a power conference for 2 days, where everyone just talked about power. And 4 or 5 really good CEOs in the room that have been in the power industry a really long time. And I think the consensus is that the United States has to basically build 200 gigawatts of new power.

And that's daunting because -- the United States has been on a cadence, since 2005, we've been lighting 4 gigawatts of new capacity pretty consistently. If you look at that chart, it's like -- it's literally like this, like somewhere between 3.8% and 4.2%, but it never deviates. And so, the utility industry in the U.S. has been on this very cozy and comfortable kind of 4 gigawatt ride. And then you see this curve, you see what's happening in terms of consumption, then you see where the grid is.

And so base load is kind of clipping along like this and then up here is actually what's needed to keep the country running. And it's daunting. I mean, the gap just keeps widening and widening and widening. So you have a supply and trade imbalance -- and the supply side essentially is in the next decade, we're going to be at somewhere around 120 to 130 gigawatts of base load available for data centers.

And again, we need probably 200 to 300 gigawatts, so there's this big gap. Now I look at -- whenever I see a chart that shows a big gap in a supply and trade amounts, I get excited because it's opportunity. And so we made the decision 3 years ago to start thinking about how do we build power how do we think about building grid independent power to our portfolios, to our 10 captive companies that own over 430 data centers, and we have over 30,000 customers.

That really struck me as opportunity because I own the real estate, I own the building, I own the customer. And so why not take the approach that we go customer first. And we think about places where we have excess land, where we can build microgrids and we can source power and bring it into our microgrid and start selling power to our customers, create backup battery sets, interconnect to the grid and then actively trade to the grid where we put power back into baseload.

That's how we've approached the problem. So we've been so far successful, it's early. I'm not going to claim any sort of victory because we've had some challenges along the way, but we've built now 3 distinct microgrids. We've got another call it, another 20 projects on the drawing board tethered to our data centers and tethered to our customers.

That's the key. And the approach we took to power was kind of the same approach I've taken to serving digital infrastructure for 30 years, which is let's start with the concept that you have to have a contract with a customer. Let's just start there. And any project we get involved in power, we have an anchor customer, just like I have an anchor customer on a tower, a fiber network or a data center.

And I don't think the power industry has really looked at it. I think the power industry has always thought, okay, I'm going to go build the power -- and then I'm going to eventually just sign a PPA with somebody and I'm going to offtake it, whether it's to Microsoft, whether it's to NextEra, Constellation, whoever it is, there's always an offtake.

And by the way, that's correct. If you go build a really good solar farm or you go build a great wind farm, there's always somebody waiting to take because the grid again, that 4 gigawatts and you got this chart that's going like this. So there's always offtake. Now offtake is interesting. What we learned in the last 2 years is that offtake is an 8% to 9% IRR business.

A little sleepy for us. We generally like returns kind of in that 14% to 22% range. And we said, well, wait a second, if you just do an offtake agreement and you print that 8% to 9% IRR, that's essentially a yieldco. And by the way, it's not too dissimilar from when towers get mature data centers get mature, but someone is making money on the development piece. And so that's what we've kind of cracked the code on a little bit is how can we buy power or build power adjacent either through leasing a transmission line or building our own private wire, we bring it into the microgrid.

And that microgrid serves as essentially as a massive substation that serves our data center. But the key differentiator in microgrids is you don't have to have a single source of power. You actually, in some microgrids, we have 2 to 3 or 4 different sources of power. So take, for example, what we've done in Reno -- we've got solar, we've got wind, we've got hydro and we've got LNG, and we have grid connectivity.

I have 5 sources of power at the SUPERNAP for Switch. That's unique. I'm not going to tell you every micro grid has 5 sources of power. It's really hard. But most of the microgrids have 2 to 3 different sources of power. So we've involved in a project, for example, in the U.K. for 1 of our big data centers there, and we're using TCG and wind and grid connectivity.

So 120 megawatts of wind, 200 megawatts of coal and about 58 to 60 megawatts of grid connectivity. Combining that all together, optimizing it, reducing intermodulation, by the way at the end of the day, we have an excess 100 megawatts that we can sell back to National Grid so that we can trade power with National Grid, peak and off-peak.

I'm going fast because I'm looking at the clock, we've got 3 minutes left. But this is where we're spending our time. It's capital formation, and it's how do we crack the power problem. And again, we don't have all the solutions. But what we are doing in terms of building grid independent power is working, and it's allowing us actually to go faster.

And if you think about some of the big wins that we've recently booked at Yonder and Vantage and Switch, how are we differentiating ourselves, power. And not like we have power in 2 years, we have power today. And that's the big differentiator.

In the last 2 minutes that we have, 2 questions for you.

We don't get to go another 20?

When does power get fixed? When do we close that gap. And secondly, what are the 1 to 2 things that people are underestimating or overlooking in your company?

So just on Power quickly, I think by 2033 to 2035 we should bring online about 60 to 80 gigawatts of NUCs. That will really help not the data center industry, but it will help baseload. And that's what we need to do. And so the U.S. has really 2 problems right now. We have a baseload problem. And then we have really antiquated transmission infrastructure.

That was 1 of the thematics coming out of the McKinsey conference yesterday with most of the utility CEOs is like, hey, yes, we can go build a bunch of NUCs and build a lot of generation capabilities, but if we don't fix our aging transmission infrastructure that was built in the '50s and '60s, how can we stay ahead or stay competitive with China in terms of our ability to produce power.

So it's a 6- to 10-year journey. It's going to transcend this White House. It will transcend the next White House. So we really have to take the long view. I've been pushing all of our relationships in Washington to think longer to think more holistically. A lot of conversations with FERC and the DOE and those conversations are going to continue.

The 2 things that I think people are missing on Digital Bridge is pay very close attention to this conversion of our Power Bank into active megawatts into active carried interest. Everyone can do that math. I give our Head of Public IR Severin White, a lot of street credit for this. He created a very simple algorithm that allows you to convert megawatt and to carry.

Once you start doing that, you begin to see this accrued amount of carry that starts to build, and it becomes a very, very big part of our NAV. It comes -- becomes bigger than 1/3 of our NAV actually over time. That's a lot of value. And again, we've got to come back we're coming back, right? We were at north of $20 in terms of share price bottomed out in the high-5s and low-6s. And so now we're building our way back through consistent earnings, building the carried interest, growing our portfolio, raising capital and most importantly, just doing what we told the street we would do.

The second thing I would say is we put a big emphasis this year on focusing on co-investment because I think when you have great ideas and great portfolio companies, private capital will come to you. And so we've done a great job of increasing our margins in co-invest, which in turn has a direct flow through to the margin or our FRE.

We've generally been kind of in the low-30s. Our goal is to get to a 40% margin on a run rate basis by the end of this year. We've instituted our first 2 levels of cost cut a third level of cost cuts is coming. And then on top of that, we're forming new capital, which we think has a 100% margin straight through.

So keep your eye on FRE. Our FRE is growing, and we're trying to put little crumbs on the trail to give people the sort of the ammunition to build that model. But I think between the growth in FRE, the growth in margins and most importantly, the growth in carried interest. We've -- we've got the ship now turned in a direction for a lot of growth in the next 24 months.

Great. Great place to that. Thanks so much for being here.

Thank you. Appreciate it.

All right. Well, we'll keep moving here. So welcome, everyone, to our fireside chat on Power and AI and the emerging data center energy crunch. I'm Will Thompson. I'm a member of the thematic investment team here at Barclays. I've spent a lot of time running extensively on the Power and AI theme.

So I'm really thrilled to be joined by Marc Ganzi, the CEO of DigitalBridge. Marc has decades of experience investing in digital infrastructure. Digital Edge, maybe I'd describe it as a relatively small asset manager, but a major investor in digital infrastructure that is pursuing somewhat of a differentiated grid-independent power solutions.

So Marc, I think you've been pretty adamant about this topic. So maybe just a quick introduction of DigitalBridge for the audience.

Yes, sure. So DigitalBridge is a $106 billion AUM global asset manager. We focus on the development, ownership and management of digital infrastructure assets. Today, that portfolio is roughly about 40% data centers, roughly about 30% towers and then the rest would be fiber and other associated forms of infrastructure.

As I like to say, we're the sort of accidental tourist in the power space. We sort of backed into being in the power space through our footprint, which is over 400 data centers. We have a power bank of about 22 gigawatts globally across 9 different data center businesses. And there's every form of power that this conference would know is tethered to one of our data centers. So we are every -- we're sort of in the vortex of all of this as it's evolving.

So Marc, I think you've quantified the AI opportunity of $7 trillion, while powering AI would require maybe $1.3 trillion to $1.5 trillion of new power infrastructure. The biggest question I'm getting today is when or if we will hit the power wall, does this force hyperscalers to actually cut back on CapEx? I maybe love you to dig into sort of DigitalBridge's secured power bank, which I think you just talked to, and its venture with ArcLight. Maybe give us a sense on how are you -- what's your vision for powering AI and general cloud needs?

Sure. And look, the energy crunch situation that's coming through data centers and through power, it's really just data, right? I'm just a big believer in data. Everything I've been doing for 32 years as the CEO is data-driven. We don't make decisions without really good data. What's happening is there is a trade and supply imbalance between power on our grid and how much leasing activity is happening in data center land. So there's a really interesting slide that a certain research firm put out, not named Barclays. But it was very telling in the sense that the sector has been kind of chugging along at -- you go back to 2022, the sector signed 3 gigawatts of leases. You go to 2023, that jumped to about 4 gigawatts. You jump to 2024, it was another 5.7 gigawatts. This is the first year the sector will sign 6 gigawatts of leases.

The problem is the grid is only turning up about, on average, about 5 gigawatts of power per year of incremental power on the U.S.'s backbone infrastructure. And what's happening is leasing is going from 6 gigawatts to -- which will be this year's number, next year, somewhere around 7. The year after that, 8.6, then it jumps to like almost 10. And then by 2032, it goes to 20, okay, 20 gigawatts of new leasing. Meanwhile, the grid is chugging along at 5 to 6 gigawatts per year of new power. So each year, last year was the first year we had a trade and supply imbalance, which means the U.S. power sector didn't deliver about 900 megawatts against what was leased.

So we started this year with a 900-megawatt deficit. This year, that deficit grows by 1.1 gigawatts. So if you're accumulating that deficit and you're just doing math, it's a 2-gigawatt deficit already, and we're in 2025. And as we look forward and you get out to that 2032 year, where the grid is turning up about 8 gigawatts and the sector is leasing 20, that's a 13 gigawatt deficit just in 1 year.

Now the question is what AI number do you believe? And there's sort of -- there's 3 forecasts around the total amount of AI infrastructure needed to make this all work. There's the conservative guidance, which is 137 gigawatts. There's the midrange of that guidance, which is the number we're anchored on, which is about 196 gigawatts. And then there's the -- we call it the -- we always pick on [indiscernible] because he's such an optimist, but he believes that AI will consume about 300 gigawatts. Now if there's anything near kind of the AI Sam Altman number, which is the 300 number, we're in a world of trouble. But I don't think that number actually happens.

I think to your point, CapEx does curtail. And why does it curtail? Because whether it was digital PCS, the Internet, cloud computing, I've been a CEO across all these thematics and all these things have a 7-year cycle, right? You get that steep slope in the first 3 to 4 years, which is we're sort of in the year 3 of AI infrastructure. It's going to go hard for another 2 more years, 4 to 5. It's going to taper and then it's going to fall off a little bit. And by the way, that's been every technology CapEx cycle for the last 30 years has followed that bell curve. So we think that bell curve lands us right at that 196 gigawatt number.

Today, data center capacity in the U.S. is roughly about 60 gigawatts of compute. And again, with the sector adding 6 gigawatts this year of leases, you get to 66. And then if you follow that trajectory through 2032, you land right at about shy of about 190 gigawatts of power, which is getting through large language models, generative AI and inferencing. Inferencing is kind of the next big sort of leg up into compute.

So again, I'm a realist. I'm not an alarmist. I've been around 4 or 5 different technology cycles. I really have a good feel for what our customers are doing in terms of CapEx. And we have a very good feeling for how fast these guys are getting the return on their investment in AI, which is faster than cloud. It took us just to frame cloud for a second.

Let's go back in time. Let's go back to 2011 when the public cloud was formed. Public cloud to today, the cloud is essentially 14 years old. We're 80% built on public cloud. We're not finished yet. And so -- and remember, of that 60 gigawatts I just told you, AI is only 35% of that. The other 65% is public cloud, and we're just now starting to build private cloud, which is a whole new vertical. So everyone talks about AI, but one of the adjuncts of AI is data sovereignty and building private cloud in the sovereign cloud. Those are also building private large language models, which is its own language model for AI, which is outside of the hyperscalers.

So I'd like to frame all this which is math and not to confuse people, but really to embrace it and to understand the gap. And so we saw this gap about 3 years ago, and we started going down the path of finding other forms of energy to our data centers. And it really started when we took Switch private about 2 years ago. Switch was really a very interesting company that really focuses on private cloud, but we weren't really so much focused on Rob Roy's passion for private cloud. I was more focused for his passion for these giga campuses and his ability to procure power outside of the grid. I thought that was really interesting.

And so it was a public stock that we took private for $11 billion. The market really didn't understand it as a story. We've gone on to quadruple the size of the company. We've done a ton of bookings, and we've built new data center capacity. But as we built that capacity, we've been building power. And we've been building grid independent power across a series of microgrids. And those microgrids are sourced with -- we work with utilities. We have interconnection agreements in all of our microgrids. In some instances, we're building private lines from renewable power directly into the microgrid where we're leasing that infrastructure through the regulated utility company.

But at the end of the day, we found a way to aggregate 4 or 5 or 6 different sources of power into a microgrid. And we've been able to create our own set of backup batteries where we store that power. And then we've optimized a 24-hour clock on how we use that power where we're constantly trading in and out of power with the regulated utility in that geography.

A great example of that is we have a really positive relationship with Nevada Power & Light. And in fact, our 2 mega campuses, one in Vegas and in Reno, those 2 campuses together consume almost 3.4 gigawatts of power. So we've got a 1 gigawatt microgrid in Vegas. We now have a 1.8 microgrid in Reno, and we're expanding both of those microgrids now, supplementing them with LNG. Both those microgrids have 5 different sources of power, including hydro, wind, solar, LNG and grid connectivity.

And I think when people ask us, why are you doing this? What's the purpose of this? And we say, look, we just can't be beholden to one source of power, it's just not feasible for what we're doing in these campuses, particularly when we're powering NVIDIA and CoreWeave and some of these really high-power density compute modules that we built out. So part of this has been necessity and survival and also our ability to embrace renewable power.

The other microgrid we built, we have 2 small microgrids in Sao Paulo of all places. So Sao Paulo, Brazil is really interesting for us. There, we have a 500-megawatt microgrid. We have a 300-megawatt microgrid all in the city of Tambore. And there, we lease the transmission infrastructure, but we have 2 sources of hydro. So we're 100% hydro and across 14 different data centers. Today, we have excess power of about 300 megawatts. We sell all that power back into the São Paulo grid. And we also sell power to Digital Realty and we sell power to Equinix, our 2 competitors.

Why? Well, we make money on it. And what we found is that building our own grid independent infrastructure has actually been a great return because we own the data center, we have the relationship with the customer. We figured out how to negotiate PPAs directly with them and the excess power we're trading in all day long.

So we have found -- we've kind of turned a negative. I'm not going to tell you it's a positive yet because there's a lot of hard work to do. But we do have a really unique relationship with ArcLight. We have a fantastic relationship with them. We've combined up strategically. We have a pipeline of about 9 gigawatts of new power projects we're building with them. And it doesn't have to be a microgrid. It can be -- we may just build a solar farm and have an offtake agreement to Google. We may build an integrated solar farm and data center like in a place, for example, like Zaragoza, Spain.

So we're coming up with unique ideas on how we build power generation adjacent to data centers or we're building our own microgrids, so we're sourcing the power and bringing it into our data centers. But in all instances, we're tethering that to a forward 10- to 15-year commitment with our customers who are looking for power and need power. And so we're taking kind of a negative and we're turning it into, we think, a positive, particularly for our portfolio companies.

And so there seems to be parts and sports that the U.S. seem to compete with China on AI, yet there's no consensus on how we're going to power it. And so to your point, it seems like a push for all of the above energy sources. Can you maybe talk about just the priorities now? Because it seems like speed to power is the priority. We've seen new policy measures in terms of something like Texas Senate Bill 6. requiring some level of on-site power. But it seems like secondary measures have been emissions, capital costs, electricity prices, relative to speed to market. How is the data center industry prioritizing what's the requirements to...

Again, I'll try to keep it simple. I don't believe we are in an AI arms race with China. I maybe have a very different perspective on it, which is kind of a 12-year view of watching China build their state-controlled LLM. A lot was said about DeepSeek, not to take a tangent, but everyone asks me, they ask me my opinion what about DeepSeek? I say, look, it's really simple. DeepSeek doesn't exist without Meta. It's really simple. If Mark Zuckerberg's open source LLM doesn't exist, DeepSeek does not exist because if it did, it would just be another adjunct LLM that runs off of China's state-controlled LLM. Here in the U.S., we're building 7 large language models, 7 privately funded through a series of different hyperscalers that are highly sophisticated, incredible data gatherers and generally speaking, pretty secure in terms of the structure of that data.

You sort of look at that in contrast to China, there's one LLM being built, which is the state's LLM, highly controlled, highly manipulated and doesn't have the funding and the capability that the 7 hyperscalers have. So -- but what China does have is they have a state regime that is very focused on power and making sure that China has an edge on power, no regulations in terms of the size of their LLM and where it goes and how it supports Tencent, Alibaba, ByteDance. Those are all companies that are supported by the Chinese sort of infrastructure.

But DeepSeek only came to prominence by using a U.S. LLM, right? Not a Chinese LLM. They didn't use the state-owned apparatus. And so DeepSeek's first version came out, anyone know how accurate it was? About 63% accurate, their first version. Their second version, which is now being run on the state's LLM, has dropped to 53% accuracy. So that should tell you everything about the difference between China's LLM capabilities and the United States' capabilities. I'm betting on the U.S.

Now our constraining factor is we don't have fields and fields of solar farms in the middle of nowhere, which China has done a very good job of doing. They've decided to weaponize their apparatus to go build as much renewable power to power AI. But ultimately, if you don't allow an LLM that sounds weird, it needs to grow. Large language models need to grow and learn and keep moving. If along that road, you're manipulating the data, you basically destroy the real sort of truth behind AI, which is that it has to get to that phase of inferencing, where it begins to think for itself. But if you have a model that's constantly being told, I got to help it think what it needs to think, you've sort of corrupted the whole concept of inferencing.

Now will China get there? I don't know. It's not -- it's -- as I say, it's not my monkey, it's not my circus. But we keep our eye on China because it's -- some of those customers like Tencent and Alibaba and ByteDance are customers in my data centers in Asia and they're customers in my data centers in Europe. Our portfolio is a global portfolio.

But coming back to the U.S., I think at the end of the day, -- as most everyone at this conference knows, what I've always been sort of looked as the sort of limiting factor to where we go is just our PUC structure. It's just very antiquated. You've got to go state by state. A lot of people in the data center space don't understand that. We've been building infrastructure for 30 years. So I know that building towers and fiber networks and data centers is a highly localized business. You then take that localization and you put the PUC on top of it and then you got FERC sitting on top of it. You have these layers of regulatory that I don't even think the White House fully understands or appreciates.

At the end of the day, power is really the gatekeepers of the public utility commissions in each state. That is going to be the limiting factor. We can remove all of the red tape in Washington. But until you remove the bureaucracy at the state levels, each state is looking at their baseload and saying, okay, I'm concerned because as you said correctly, people are waking up to the fact this is going to hurt consumers. So what do we need to do? And we say, look, at the end of the day, if I can be a net contributor to the grid or if the microgrids we build or we aggregate power and I can sell power back into baseload, I turn from being an enemy into a friend at the PUC level.

What I do worry about is that this weaponization of the cost to consumers is going to get politicized and it's going to slow us down. And I think that will be a real problem for the hyperscalers. Look, the last administration, our former Secretary of Energy, she -- I met with her twice. I think she's really smart. I really like her. But her answer was, well, we'll just wait around and let the hyperscalers pay for it. That's never the right answer. There has to be a solution that brings DC together, PUCs together, private investors like us and the hyperscalers to build some of these grid independent solutions, which is what we decided to just go do on our own, kind of we were left our own devices.

We often throw the different workloads in the same bucket in terms of data centers. Can you talk about what are the restrictions or requirements when we think about cloud, which can be hundreds of different cloud products and you talk about public and private cloud. And now we have AI inferencing AI workload. And there's different latency land requirements. There's potentially different power fluctuation requirements and then obviously, the five-nines often gets brought up in terms of reliability. Maybe just talk to the different considerations when we think about those different workloads.

Well, what's interesting is there are very distinct workloads now. And they tend to sit in different types of data centers, and they're using different types of GPUs. And ultimately, for AI inferencing and large language models, you want to use the highest-powered GPU and get your hands on. So when we talk about NVIDIA's next-generation chips and you talk about the Blackwell chip, they're really expensive. But the hyperscalers want to get their hands on them because ultimately, that processing capability and the ability for that large language model to learn is much faster. Now the adjunct to that is there's -- as most of you know, power people that is more power density. So you're trying to squeeze more power into a smaller GPU, which is a smaller rack.

If any of you ever get the chance to tour Switch in Las Vegas, it's actually where CoreWeave recorded their entire IPO roadshow, it's where they did the IPO. And you can go in a couple of those data halls and you hear these Blackwell chips and you've never heard a sound like this. The hissing, the sort of the pitch to that is defining. But what it is, it's power density. And those chips are no longer cooled with forced air. In fact, a Blackwell chip melts if you go through forced Air. The only way you can power NVIDIA's next-generation chip is through liquid cooling, which is what we're doing at Switch.

And I think the other reason that Jensen and Mike at CoreWeave have chosen Switch is, to your point, they're the only Tier 5 operator out there. So it's not five-nines, it's 100% uptime. A Switch data center has never failed and people will pay for that. The customer will pay more rent to be there. And our liquid cooling system, our EVO system, which is our patented cooling system, is really revolutionary. We don't lose one drop of water, which I'm actually pretty excited about. I'm from Colorado. So we get a little excited when we talk about water because we're losing water every day.

Once we're done talking about the degradation to the consumer in AI, people are going to move on to water next. That will be the next topic that we'll be talking about a year from now. But for right now, I've got to just solve the problem that's in front of me, which is how to convince public utility commissions that AI is not the devil, and it's not what's going to be driving up consumer prices, which right now it is. If you look at the data, the amount of power that's being consumed in the data centers is impacting baseload, which is impacting consumers.

So I think some of that's going to have to come back to the state level, which unfortunately, those PUCs are going to have to be rethinking about rates. There's going to be so many rate cases in the next 12 months. You're going to see rate cases in every state, and I guarantee you're going to see very few rate cases where the rates are going down. My suspicion is rates will go up, and it will be segregated between AI rates and consumer rates. We will begin to see a parsing of how government charges the consumer versus how they charge data centers. Now for us, as an owner of data centers, that's a pass-through. We don't pay for the power. Now if we have our own microgrid and we're producing our own power and we're bringing it into the data center, then we do have an offtake agreement with our customer. So I do -- looking around corners, I worry a little bit about that.

And we get the sense that some of the utilities are driving these sort of data center-specific tariffs, right? We see in Ohio. I mean is that what you're talking about in terms of both the public commissions and at least recognizing that there is an inflationary effect to other retail and industrial.

Look, if I were running NextEra -- my friend, John runs NextEra, I really like him. He's a great CEO. If I was running a big public utility company, I would probably be doing that because I'd want to get out in front of that before I'm stuck in a rate case. Rate case is the 2 dirty words for them. So I think that the industry can stomach it. to a certain degree. I think at some point, there will be pushback and then the customers will seek their own solutions. But the reality is the solution set, if you're a hyperscaler and you're trying to build a 1 gigawatt data center, you don't have a lot of solutions.

Where are you going to get the turbine for your LNG solution or CCG solution? Turbines right now are backed up 2 years. We have relationship with all the producers of turbines, and we have -- our forward log of turbines is we're set until the end of 2026. And then we even have a problem in terms of sourcing for our microgrids. But look at this thing is complicated. And like I said, the one thing we have learned in the last 3 years is, one, -- our friend is the public utility companies. We work with them. We're interconnected to them. And most of our power is bought from them across our global portfolio. And I think what we're trying to do is create solution sets that augment that and ultimately are a net contributor. I think if we stay in that swim lane and we keep going, I think we'll be pretty -- we should be successful.

And do you see a situation like we've seen in Texas with Senate Bill 6 is sort of forcing the hand to require microgrids that are interconnected or on-site power generation is back up just become when the grid does become constrained?

I think, look, the -- I think Abbott is smart. I think he knows his base, his base is the gas industry. There's an abundance of gas in Texas. We've proven that microgrids can work in Texas. The first Stargate site at Crusoe is half grid, half ERCOT, half microgrid. We're working on a solution in Lancaster, Texas. It's very similar. It will probably be about 75% microgrid and about 25% ERCOT. And so Texas is pretty unique because it's ERCOT. It's one of those sort of for us, it's kind of a fish out of water. And so we've had to spend a lot of time with ERCOT trying to understand what the problem is. And by the way, ERCOT has its own problems, which is they're on a 10-year project to completely rebuild Texas.

So anything, I think the governor is being smart because he knows you got to have supplemental power because the baseload on ERCOT right now is already fragile as we've seen in the last 3 years in the wintertime. So Texas is pretty unique. And most of our grid independent infrastructure is being built in Texas. 3 out of our first 5 microgrids are in Texas. And by the way, he's made it easy. I mean, he hasn't made it hard. He's been very clear about what Texas wants. And ERCOT has been also very clear, too. If you want interconnection to us, here's what we expect from you. I kind of like building data centers in Texas because the rules of the road are quite clear. You may not like the cost, you may not like the final outcome, but at least they make it very clear where you can go and where you can't go.

In the last 5 minutes here. Maybe you made some comments earlier about the digital infrastructure cycle. We often get a lot of questions on the ROI of AI and the sustainability hyperscaler CapEx. You suggested we're sort of inning 3 of a 7-inning game maybe in your viewpoint. But I get the question like how is this different than the telecom boom bust of like the early -- earlier this decade?

Well, what's interesting is if you go back to the advent of the PC, the mobile phone, Internet, sort of mobile data and cloud computing, those are kind of sort of 5 tectonic shifts in technology. And there's a slope around adaptation and how long it took to adapt. So to get to widespread adaptation in the PC took 12 years. It took essentially less than a year to get almost 92% of Americans to touch AI. So adaptation and AI, the slope on that was the fastest we've seen in any sort of technology introduction. What's also interesting at that same time is the cost to produce AI or to reduce a token, which is a measure of AI, is radically falling, and it's fallen 40x since the inception of AI.

So what's really interesting is cost per token is down, adaptation is like this, where you look at the PC, which was the curves were like this, adaptation came down like this and cost was like this. So sorry, adaptation like this and costs like this. And in between there were these other introductions of different technology. So what's interesting to me is the positive revenue impacts or the positive ROI in AI took 3 years. Public cloud took 5 to 6 years. So if you go back to the earnings of Microsoft, it really wasn't until 2016 that Azure was producing positive EBITDA and then it just ramped.

And so what's happening right now, like with, for example, with Chat and you look at the earnings from Amazon, you look at earnings from Microsoft, these guys are all now producing positive net income from AI. We're 3 years in. It's early. And the use cases for AI in public cloud, there were 2 use cases. So in public cloud, you had public cloud, which is what we use for Internet and document storage. And then you have consumer, which are all the applications that sit on our phone. All of those run on the public cloud.

When we go to AI, we actually have 5 different use cases for AI. So there's 3 new use cases in AI that didn't exist in public cloud. And so you've got -- obviously, you've got enterprise, you've got industrial applications. You've got consumer, which is the same as public cloud. You have enterprise, which is the same as public cloud. But there's 2 other new verticals that come out of that, which is data sovereignty, which is huge. And then the one that nobody talks about is machine-to-machine learning.

So today, there's about 30 billion connected devices to the Internet of Things, to IoT. In the next 7 years, that goes to 60 billion devices. Remember, machine-to-machine connectivity means there's no human in between that. So you got one language, you got one model, talking to another model. One AI agent over here, one AI agent over here, going back and forth. And so it could be a public safety network talking to an autonomous vehicle. It could be a wireless electricity meter talking to your credit card company. So imagine a world where you've got 60 billion wireless devices, accelerating the conversation between 2 machines. That will be 80% of AI consumption will not involve a human being. That is staggering.

And is that -- because that's where part of my debate seems to be is that -- I mean my view is that we're now entering the agentic AI era. And that's time test scaling on steroids.

Correct.

And then you have this potential for Tesla humanoids, which is physical AI, which again would be time test scaling on steroids, and that's the surge we're seeing in maybe AI inference demand. Is that a fair way to think about it?

It is a fair way to think about it, but I think the Tesla robot for me is like an anomaly. It's like an outlier to a certain degree because there's robotics inside of that, that are less independent sitting on a factory floor that are a lot more productive than an Elon robot.

Robotics is a part of that machine-to-machine learning, right? But the amount of data that will be consumed to deal with machine-to-machine learning there, we haven't even started the baseball game yet. We're like top of the first. The pitchers are out of the mount warming up. So -- and there's so many revenue models that pop out of that. And there's so many implications to fiber and to cell towers and to mobile infrastructure. The ecosystem is just getting warmed up. And again, to your point, we're in the third inning of a baseball game. So there's another 6 innings left of a lot of infrastructure spend coming.

All right. Well, obviously, I could talk to you for another half hour just to cover the topic, but we're out of time. I appreciate Marc coming, and thank you, everybody, for joining us.

Thank you for having me. Appreciate it.

Citi's 2025 Global TMT Conference. For those of you I haven't met, I'm Mike Rollins and I cover communication services and infrastructure for Citi. Disclosures are available at the back of the room and if you don't have access or would like another copy, please e-mail me at [email protected]. We're pleased to welcome Marc Ganzi, CEO of DigitalBridge. Marc, thank you so much for joining us today.

Yes. Thanks, Michael. It's good to be back here.

It's great to see you. And maybe just to kick us off, maybe you could share with us how DigitalBridge is differentiating itself with its platform. Maybe in a couple of ways. One, pursuing the investments in AI and digital infrastructure but also how the combination of having all these companies within this portfolio approach can enhance performance and outcomes for your shareholders?

Yes. Well, thank you. I'd like to frame that sort of answer with just the enormity of the opportunity that sits in front of us. I think people don't really get it contextually how big what's happening right now from an infrastructure perspective. And listening to my friend, Steve Vondran, right before you and I got on stage, this is a multi-decade opportunity. You can't frame this in the prism that we frame 2G or 5G or public cloud. This is really a transformational moment in time.

And so to get the infrastructure to perform at that level, Michael, requires an investment level that we've never seen and it requires different parts of the ecosystem to come together, but it also requires us to think about how to address problems in delivering that infrastructure, which has created some adjacent swim lanes, which really works out well for DigitalBridge, and I'll come back to the framing of the thesis of why DigitalBridge is probably the most unique TMT stock you could possibly own.

We see AI CapEx as a $7.2 trillion opportunity. We see the adjacencies of power as another $1.3 trillion opportunity in CapEx so when you put the AI ecosystem together with power, you're talking about $8.6 trillion in CapEx. We used to -- even I used to sit at these conferences a decade ago and talk about, gee, a great CapEx year was if a customer spent $100 million in CapEx. So just in a period of 10 years of how far we've come from a sector perspective and DigitalBridge is 12 years old now, it's staggering the amount of CapEx that's required to do what we need to do to support our customers. I've been doing this, you and I have been friends almost over close to 30 years. You know I got my start in the '90s like you. I've never seen anything framed like this in my career.

And the #1 thing that is required to deliver for shareholders today is your ability to form capital. If you cannot form capital, Michael, you're out of the equation. Again, I'm not picking on Steve, but we're good friends. Vondran and I've known each other a really long time. He owns CoreSite. CoreSite is not building 200 to 400-megawatt campuses. CoreSite is not in the room in a 1 gigawatt conversation with the customer. I'm not only in that room. I'm the guy making it happen. That's a very different position. And it all revolves around this notion of being asset-light versus being asset heavy. I made a decision to de-REIT this business 2 years ago because I saw what was coming.

Just like I saw what was coming in the '90s, like I saw what was coming out of the dot-com crash. I've been able to identify cycles very early. And this battle, this opportunity is just different. And what you had on the stage previously was the right solution for the last 30 years in digital infrastructure. What you see sitting in front of you right now is the solution for the next 30 years. The recognition of being asset-light and understanding how to form capital is what gives you the seat at the table to go have these conversations not just in data centers, but in towers and in fiber and in power.

A great example of that in towers was American Tower didn't have a shot on goal on Verizon. We were the preferred partner. We weren't the highest bidder but we won those Verizon Towers because I have a 30-year relationship with Verizon of trust and building network for them and not gouging them on MLAs. And again, you talk about the ecosystem. The second part of your question is how does DigitalBridge show up. We have 57 companies around the world. We understand how all the pieces of the ecosystem tied together, whether it's connectivity, dark fiber, interconnection, edge compute, hyperscale, private cloud, public cloud, our portfolio companies wake up every day, and we address thousands of customers, tens of thousands of customers.

So I look at it and say, you can go as an investor, and you can go buy stocks and take out a single arrow and shoot at a target like American Tower and you're probably going do really well. You'll do well over a decade performance. Or you can say, look, I want to be with the guy that has multiple arrows in his quiver and can shoot at multiple targets to create bigger returns and bigger outcomes.

And in that asset-light model where we're forming capital and we're doing them in permanent capital vehicles that have no end of fund life or infrastructure funds that are typically 11 to 13 years, our management fees, our revenues are actually longer than a fiber company. They're same if longer than a data center operator. And our weighted average tenor, believe it or not, is actually longer than most of the public tower companies. So if you pull it back and you look at DigitalBridge as a digital infrastructure owner and operator, we have the most durable and steady revenue streams, and we have a higher concentration of investment-grade revenues because guess where our revenues come from sovereign wealth funds and pension funds around the world and insurance companies. So we have investment-grade counterparties that anchor our earnings.

And at the end of the day, whether it's tower cash flow NOI or an our metric FRE, where I'm gauged against Blackstone and KKR. My FRE growth is 23% year-over-year, okay? That's not what American Tower and Digital Realty and Cogent are doing. I can assure you of that. So if you put us in that ecosystem and you put us in that prism, no one's growing faster with the durability and the cadence that DigitalBridge is growing and nobody has a front row seat to all of the opportunities, not just one swim lane.

So we've got to get out there and we've got to tell that story with a little more precision because when you frame it that way, if you don't own DigitalBridge and your digital infrastructure portfolio, you're missing a huge part of the opportunity. And so that's why I'm in New York today, that's where I'm sitting with you is so we can have that conversation, and we can bring the relevance of what's happening to the forefront and investors have to make a decision. They have to choose. They have to decide where they're going to put their capital and my bet is I want to be diversified. And then I'd also want to seat at the table around power because power is changing just as fast as digital infrastructure is changing, the power industry is changing, and we're rewriting some of those rules.

The thing you mentioned just the quantum of CapEx and capital that's going to get put in the ground. I think the thing that surprises me the most is how much of the capital is coming from private sources versus public. When I look at our last 30 years of covering these industries, is that sustainable? And how do you look at your opportunity to go after these private investment opportunities?

Well, look, for us, we have to recycle capital. I mean, that's really important. And again, I'm not picking on Steve. He talked about his dividend as the way he recycles his capital. He pushes his money back to his constituents. We do the same thing. We've returned over $9 billion of capital to investors in the last 2 years. That's called DPI. And at the same time, we've raised tens of billions of capital. So what's interesting is, as I returned $10 million, I'm generally tacking on another $20 million to $25 million of AUM every year. As you know, out of the quarter, we were at $106 billion of assets under management, which actually if you gauge Crown American and Digital Realty and Equinix on that metric, we would be the biggest by a wide margin.

And again, we're growing that AUM by about 20% per year. To do that, you've got to form capital. But the key to private capital markets is the return of capital. And we said that last year in our earnings presentation, we put a little more focus on DPI. And look for public shareholders, DPI is not very popular because you lose management fees. You take a step back. And there was a moment in time last year in the fourth quarter where we took a little step back. But guess what, in the last -- in the third quarter last year. But in the fourth quarter, first quarter and second quarter, what we've done is we're taking many steps forward. So sometimes in our business, you return the capital and that opens the door and widens the aperture to go raise more capital. So with the consequence of having 2 years of good DPI was it really enabled us to grow our private credit business, which I'm really excited about. It will enable us to go into the power business, which we talked about last quarter. It enabled us to go out and raise more capital in our third flagship fund, which is performing extremely well.

And so I think for us, private capital for DigitalBridge will remain open provided that we're on this normal cadence of kind of 2 steps forward, 1 step back or you could say 1 step back, 2 steps forward, depending on how you look at it. But allocators we know 2 things. When I walk into a meeting with an allocator, I know 2 things that are absolutely true. One, they are underexposed to alts, like flat out insurance companies, pension funds, sovereign wealth funds. We know the minute I walk in the door to pitch them about DigitalBridge. We know they're underexposed to private alternative assets.

The second thing we know as they are underexposed to digital infrastructure. And so I walk in with a big advantage when I'm asking for a pension or a sovereign or an insurance company or a family office to give us capital, which is we know right away that as they are reshaping their global portfolio away from private equity, away from real estate, where are they filling those holes? Digital, power infrastructure, alts, those are all very safe swim lanes for us to grow. So we have a couple of things going really well for us, which is we're returning capital. We know allocators are underallocated to what we do. And we know our story is incredibly unique and differentiated, and there's not even a close second place in terms of AUM, in terms of specific digital infrastructure manager like us. So we got to take that lead. We've got to grow it. We've got to press our competitive advantage and then we have to press on our comparative advantage, which is this cross-pollinization of our portfolio companies.

When a customer says, hey, we have a really complicated build. And we really need you to address this compute capacity. We need x amount of pairs of dark fiber. We need redundancy. And oh by the way, we need interconnection. And we can say, no problem. We've got all 3 or like in the state of Louisiana where they said, we have this opportunity and we have this big grant, but it requires towers and fiber. So what did we do? We showed up with Vertical Bridge since they are only one, nobody else could address the RFP. We have this enormous flexibility, Michael, to go anywhere, show up and manifest ourselves in a way that is different and differentiated from a DLR or an American Tower or a Blackstone or a KKR, that's really the uniqueness of the brand and the power of what we've built in the last 12 years.

And you mentioned earlier the 23% FRE growth. When you look at the growth of the assets that you're investing in, can you give us a portfolio update in terms of what you're seeing?

Yes. Well, as we say, you try to love all your children the same, but of course, you can't. You have to love some children a little more than others. Let's take data centers as a good place to start. Data center growth has been fantastic. If you looked across the 9 different platforms we own, all of them are growing double digit. Most of them are growing greater than 20%. So this is really a unique moment in time. We highlighted it on our last quarterly call, just around the depth of our pipeline. If you look at the year-over-year growth in our bookings and our pipeline, both those metrics are greater than 50%. That's stunning, not even in the heyday of towers, did you see that kind of growth. So our backlog today across our 9 different platforms is over 9 gigawatts of opportunity in terms of whether it's an inquiry lease application or draft lease or lease that's about to be signed. And if you go back a year ago, that pipeline was at about 5 gigawatts.

So we are really seeing an enormous explosion and opportunity. Now once you get through that opportunity, you've got to parse it, right, private cloud, public cloud, large language model training, inferencing, all of this stuff is so specific and the ability to have our platforms show up with that specificity. So if I have a big sovereign cloud opportunity, you know where it's going, it's going to Switch, Tier 5, highly secure and really the inventor of private cloud. If it's a big hyperscale opportunity in the U.S. and a customer says, we need a 1-gigawatt giga campus to be built, we tap a real advantage. And you saw that a couple of weeks ago at the press release about what they're doing in Texas. And Vantage is just a juggernaut in terms of its capabilities and what it can do for the hyperscalers.

And then the stuff that Raul is doing in DataBank is incredible. He actually has the fastest 5-year CAGR growth of all of our portfolio companies. He's been averaging 25% CAGR growth for 5 years. And I'll -- again, I hope Steve doesn't think I'm picking on him. Edge compute is going like crazy, actually. It just happens at DataBank, we have a lot of capital and we can move faster. So what's happening at the Edge right now is these kind of 10-megawatt to 60-megawatt mini campuses where you initially you're seeing the proliferation of edge cloud or edge applications. But now what's happening is inferencing.

And so instead of an inferencing workload is very different from a large language model workload. And those workloads are between 5 megawatts and 60 megawatts, but the ability to have a tethered solution where you're not sitting in a Napa, in a downtown Napa, but you're outside in Plano, Texas or you're in Overland Park, Kansas or you're in Bluffdale, Utah, where we have between 50 to 100 megawatts, we can show up and we can deliver that inferencing workload for Grok, for example, in Houston and the Woodlands, and deploy for them very quickly because we know what they need, what do they need? They need interconnection, they need multipath fiber pass. They need sort of not too diverse pass. They need 4 diverse pass. They don't need 2 pairs of dark fiber, and they need 28 to 56 and our ability to show up with interconnection, fiber in the facilities is why Grok uses us. It's why CoreWeave uses us. It's why NVIDIA uses DataBank.

DataBank's ability to deliver 80 unique locations in 29 markets is actually a bigger portfolio footprint than Equinix and what it has that Equinix doesn't have is it has that tethered solution and it has capital and has capital to go fast. And so the edge is happening. Sorry, Steve Vondran, but it's happening. And if you're not involved in it, you're going to miss it. Because right now, we're building those public cloud workloads on the edge, but now we're building those inferencing workloads. And Steve did touch on a little bit about machine-to-machine connectivity. That's where inferencing really hits and connects with the mobile network, and that's where we get really excited about towers.

So data centers are going quite well. Again, if you're focused on public cloud, private cloud, hyperscale, large language models and inferencing, and edge, if you've really tethered across those opportunities, you're going to see a lot of high growth. I think in fiber, it's kind of the tale of sort of 3 cities, right, which is maybe even 4 cities, I would offer to you. Start with the good, the bad and the ugly. The good is what's working right now. We really like this idea of building a backbone to multifamily, we call the MDU space. So we have two investments that do that really well. We have one in Europe called Netomnia. We have another one called Fibernow in South Florida, 10-year contracts, 15-year contracts with HOAs. You take no execution risk in the subs. You have 100% penetration day 1. We like businesses that are infrastructure-like and when you have those MDU businesses, it's different than competing with the cable company because you have the exclusivity with the landlord and you can go fast.

Resi fiber, we just made an announcement in WideOpenWest. So we do believe in that story. We believe in cable plant, transforming it into broadband plant. But you got to do it at like 4x to 5x EBITDA. You can't do it at 16x, 18x, you're just too far behind the curve. So I would say that's not the bad, but if you buy it wrong, it can be very bad. If you buy it right, it can be very good. So I think we're very price disciplined around residential fiber.

The other two pieces of the ecosystem are really enterprise fiber and what I call true infrastructure, transport. The commercial side has been fine. Enterprise has been good at Zayo. Bookings have been steady at 4% to 5%. Churn has been under 200 basis points. So you're a net 3% to 4% grower. It's not terribly sexy, but it works. Cogent is a good example of that. If you've got a great sales team, you've got people tethered to the buildings and you can really penetrate, you can achieve good growth rates. I think the area that's working the best in fiber today is transport. Traditional infrastructure, which is fiber to AI, fiber to inferencing, fiber to data centers, long-haul transport routes where you're connecting to NAPS like we have a couple of routes, new routes at Zayo that are performing really well. Those are great because we're getting 15- to 30-year commitments from our customers. And most of those builds are happening at where our customers are putting up 50% of the CapEx. We're taking very little execution risk. And the paybacks are now back to inside of 36 months.

Now go back 6, 7, 8 years ago before we bought Zayo and Caruso was talking about paybacks inside of 10 years, that just doesn't work. Those are bad ideas. So we're making our customers pick up 50% of the CapEx. We're very disciplined for paybacks inside of 48 months. We're now seeing those paybacks inside of 36 months. And what's more important is we're getting the second and third carrier on those routes a lot faster. So you look at routes, for example, like Northern Virginia to Dublin, Ohio, or you look at Atlanta, Georgia, to the Infomart in Dallas, you look at One Wilshire all the way up to Mae West and extended into Beaverton and Seattle, where we built new routes, we're now on to the 6 or 7 tenant on those routes because we built those not with 12 pairs, not with 24 pairs, we're building them with 412, 524, I mean, high, high, high paired routes, a lot of dark fiber.

So that you think about these things, I know Jay Brown always to say, think of fiber as a tower put on its side. Well, if you only have 2 to 4 pairs of dark fiber, there's not a lot of RAD centers for you to co-locate to. But if you have 500 pairs of fiber, the transport business is really good. So we're seeing extraordinary returns right now at Zayo in terms of our transport bookings. So we're very excited about that. So fiber is a little bit of a kind of a mixed bag, but you really as an investor, you have to think carefully about where you're investing.

And then the last place that I'll go to is mobile infrastructure. And in mobile infrastructure here in the U.S., it's been -- I think I said this to one of your competitors' conferences a few weeks gone in Utah. I haven't seen leasing like this in 13 years to be precise. We see, as Steve said, this is the first quarter at Vertical Bridge, our U.S. cell tower REIT, where colos outnumbered amendments for the first time since 2013. So we've been very patient. We've waited 12 to 13 years, and now colos are back.

And you have to ask yourself as a guy who understands RF engineering, why is that happening? It's happening because of the amount of pressure on the device, when you think about applications and you think about AI, it's putting enormous download pressure on the network.

Now history lesson. We go back to 2013 and why was that happening? It was happening because public cloud was being introduced to wireless devices. So mobile data traffic moved from 2011 to 2015 up 10x. That's just hard empirical data. That's the fact. In the last 6 months, I've heard Jensen, and I've heard Masa-san say the same thing, which is they believe mobile data traffic will increase 3x to 5x and the carriers' behavior is manifesting itself. And so all 3 bands are being tested and you were talking about the AT&T merger, we can talk about that in a second. But we see 14% organic growth in U.S. towers this year so far. And the amount of colos that the customers are wanting is because of carriers don't like the word densification, but it is densification. It's the network. As you turn up more capacity, you create holes in the network and you got to fill in those holes. And so we're seeing a lot of that.

And we've had really incredible leasing, huge demand on BTS. Our BTS backlog in the U.S. is up over now 5,000 towers. We have agreements in place with all 3 customers. We're the preferred partner to Verizon. We're a preferred provider to AT&T and T-Mobile. And there's not a single tower company in the world that's building 1,000-plus towers in the U.S. other than us, so advantage us. I think last year, when you accumulated SBA Crown and American, they built 200 towers. We're building 5x with the 3 of them have combined are building. And this is a great DigitalBridge investment.

We're very bullish on U.S. towers. I believe the fastest growth area in AI will be machine-to-machine connectivity because once machines start talking to each other, there is no off switch. We can always turn our phones off. We can turn an application off. The minute you have an IoT sensor talking to Tesla's mainframe, that's a conversation. The minute you have public safety, interagency conversations where the devices are talking to each other for early detection. Those large language models are powerful, and they're building. Every day, they're getting bigger and bigger and bigger. We know this from Switch because of some of the data sovereignty work that they're doing for the U.S. government. This is a huge opportunity. I don't think people are thinking about it correctly.

Vondran was correct. The demand for towers over the next decade is going to amplify. And if the FCC doesn't move on spectrum, it's completely the advantage to the tower operator and our ability to keep building is a huge advantage for Vertical Bridge. We got to keep introducing new locations. Having these 7,000 Verizon towers are really helpful. We just moved to being the #3 operator in the U.S. We just passed SBA in terms of tower capacity and tower locations. So I'm naturally a tower guy, so I'm always bullish. I feel like I'm the cheer leader for the tower industry, but this is a moment in time where you really have to get excited about what's happening in mobile infrastructure. And small cells are coming back. We see renewed interest in small cells. It may not be the small cell you used to know. You got to be thinking in multi-architecture like mini macros, smaller nodes with one antenna and one RRU. Our customers are getting a lot more creative. This notion that a small cell looks like that. That's not true.

Small cells are taking different forms. We've seen explosive growth at Boingo. Boingo keeps signing up new deals, has a great opportunity with the U.S. military. We have all the U.S. military bases for fiber and WiFi. And we see small cells coming back in '26, '27 and '28. I know it's kind of a bit of a dirty word in mobile infrastructure parlance, but carriers need it. They need that solution and they need macros as well. So this is -- I could tell you it's the golden era of like 1996 of towers, but there's this whole new investment cycle coming into towers right now, and I think investors underappreciate it.

So you mentioned the leasing at Vertical Bridge. The tower companies have been talking about leasing getting better as well. And I think what's happened over the last couple of weeks is with the possibility of carriers getting spectrum where they could just like plug and play it. And it's some spectrum. But there is a concern that's risen that maybe this then slows that densification ramp or pushes it out a few years because in the formula of capacity, they can use spectrum, they can use cell sites, they could use technology and here, you're amping up that potentially that spectrum factor. So is that something that at least could put the brakes on or hit a speed bump in terms of like this multiyear path or do you think just given what you're seeing in the near and the medium term, the spectrum maybe reshuffle is not going to be significant for tower leasing.

Well, I think, look, first of all, each carrier has a very different spectrum position. So I think what we've always prided ourselves in is having very specific conversations with specific customers. So our big 3 customers here in the U.S. want very different things. It's the same thing in Germany. We have the largest tower portfolio in Germany, and we've had 8% organic growth there, which is almost double what Cellnex has. And people go, why is that happening? And I said, well, we have great locations. We have towers that are really big and we also now really acutely understand the problem of Telefonica, Vodafone and Drillisch. Having these very detailed conversations with your customers where you're trusted, and you're not perceived as the enemy because of 30 years of bad MLA negotiations is a real advantage right now.

So again, when you walk in the room, advantage DigitalBridge. Why? We don't have the taint of having these multigenerational MLAs and SLAs that Cellnex may have or Steve may have or Crown or SBA may have. And so I'm having these conversations with customers. Our CEOs are -- they're incredibly direct. And what I would tell you is you've hit the nail on the head. I think they are concerned about spectrum because some of them have them, some of them don't. Even T-Mobile that's very spectrum-rich is very active right now, extremely active because their network needs to get stronger as they've added more subs.

But when you have this paradigm where you add new subs, new devices and a completely new frame of reference from a technology perspective, which is AI T-Mobile's network needs help. AT&T's network needs help, which is why they went out and they got DISH. And Verizon still needs help. Hans and Lynn Cox are busy, they're building in Kyle, and they have a great network. They have a network advantage but the way Hans thinks and Kyle thinks is they're always saying, I want to be ahead. I want to be one step ahead. That's the mentality of Verizon. They're always planning ahead.

So all these carriers are very different, Michael. But I think they are thinking about how are they deploying their O-RAN or C-RAN, most of them are C-RAN now. And they're thinking also about how do they deploy that radio technology at the edge, how do they interface to interconnection, how do they take more revenue from the hyperscalers this time around because I think they feel like they got, they didn't get the full value of the bargain in cloud. I think they will get the full value of the bargain in AI. And at the same time, they're counterbalancing that with their spectrum needs and ultimately, their site needs.

I see all 3 carriers for the first time in the U.S. as being healthy, well capitalized and growing, and they're growing ARPU for the first time in a decade. Those are really good healthy metrics. Now how do they keep doing that? They do it through the network. They do it by having network advantage and keeping customers loyal. And as we all know, we're on our devices. We're on our mobile devices. Someone said recently like 60% of the day, I find that really hard to believe that's not me. But we do really -- we are very tethered to our mobile devices in more ways than not.

And the last thing I would say about towers that gives me a lot of optimism is, remember, Michael, today, we'll finish 2025 to 30 billion connected devices. That's up from 26 billion last year. Most people don't know that we're going to 59 billion connected devices by 2033. So imagine, if today, we have 30 billion devices, and you're doubling that, what do you think has to happen in the network? Spectrum only gets you so far. C-RAN only gets you so far. At the end of the day, it is physics. There is a physicality to how radio spectrum waves propagate. And those propagation characteristics define ultimately how the radio works. I'm trying to distill this at a very simple level and that's why we have a lot of optimism around towers.

Very helpful. I want to go back to the data center conversation, the power conversation. You talked about your pipeline going from 5 billion to 9 billion. How do we think about the win rate for your portfolio companies and the annual opportunity to turn that pipeline into bookings and revenue.

Well, I think, look, this year, our global footprint, when we looked at the 5 gigawatts last year, we converted about 1.9 gigawatts globally. So we felt pretty good about that. That's a good win ratio. That's a 40% win ratio. I look at our pipeline to have 9 gigawatts I think it's not totally unrealistic when you own companies like Vantage and Switch and Scala, I mean, in DataBank, these are big companies that are effectively their business plans are measured in gigawatts now.

So we'd like to get that 40% win ratio. I mean it would be great if we could do another 4 to 5 gigawatts of bookings this year, not sure where we'll land. But if we did land at 4 to 5 gigawatts booking, it would be multiples higher than Equinix and DLR. What it took to win today versus what it took to win 6 years ago has completely changed. And I think our -- having those 9 companies that can be very surgical in those conversations and go for those wins is really important. But at the end of the day, it's power.

Do you have the land? Do you have the entitlements? Do you have the power and it's not good enough to say you have a will serve letter. What's your connection date? And what's your certainty on that connection date. Customers are really sophisticated. They have massive power teams. Google, Meta, Oracle now, Amazon, Microsoft, all these guys have huge dedicated teams, hundreds of people thinking about power, the same way we do. We have a huge team that wakes up every day and thinks about power. And so we're trying to get out ahead. We said on our call, we have a 22 gigawatt power bank and that's big. That's the biggest power bank of any data center operator in the world. We'll lease through that power bank easily in the next 5 years, easily.

Now the question is, how do I keep building more power? And how do I bring more power online. And there's a couple of ways we can do that. We can build power adjacent to our substation through a microgrid, tether that to our substation and supplement the grid power with external power. We're doing that very successfully. And then what we do is we sell that power through to our customers on much like a tower lease or a data center lease, it looks like a 10- to 15-year PPA. It's a power purchase agreement.

And then as we think about new sites and we go out into communities and the power company says, yes, look, we have no power left for you. We said, what if we brought our own power and we turn that power up. And then during peak hours, we sell some power to you. And off-peak, we bring it back and we buy it from you between 1 a.m. and 6 a.m. We're starting to learn how to trade power for the first time. So we have 2 active microgrids, 1 in Brazil and 1 in the U.S. that we built ourselves. We're actually trading in the power. In fact, in Brazil, we sell power to our competitors. I sell power to DLR and Equinix. No problem. Happy to sell power to my friends. We make money off that. And what we're finding is the returns in these microgrids and these independent generation capabilities that we possess are producing similar if not better returns in data centers. I never expected that. But now we know how the utility companies make their money. They trade power all day long.

So we figured that out, we understand now how to trade actively trade power. We have excess power. I think the key part of the conversation point that a lot of people aren't talking about that I've been saying for the last 2 years is, you have to go back to FERC and you have to look at baseload. And you have to look at the different grids in the U.S. and you have to see where we have grid imbalance. And the fact, for example, that take, for example, what happened at Susquehanna, FERC came in and shut down Amazon for that 800 megawatts. Amazon only got to that first 100 megawatts of nuke power. Why? Nobody did a baseload study. Nobody figured out why that nuclear power plant, how much baseload it gave to the Mid-Atlantic, you have to do that work.

So for example, when we go to a place like [ Rhino ] and we have a massive microgrid there serving the SUPERNAP of Switch, we're actually a net contributor to the grid in that microgrid. I'm putting power back into baseload. And when you start putting power back in a baseload, your conversation with FERC changes really fast. They don't look at you as the enemy. They look at you as actually helping to solve the problem. That's what we're doing at DigitalBridge. We've got a massive pipeline of power projects. We're super excited. We announced our first power project with ArcLight. We have other projects that we're going to bring to the market shortly. They're great guys. They're like the DigitalBridge of power. All they've been doing for 30 years is building power.

And so we have again, we've gone back to our old routes, which is go back to the drawing board, go build something, go solve the problem. And by the way, it's okay to use solar. It's okay to use wind. It's okay to use hydro. I know it's not super-chic to be talking about green power. But look, those power sources are abundant. They're adjacent to where we have microgrids. And when we can be a good corporate citizen, we're going to do that. We're going to make the right decision.

And we're going to make the right decision around water in terms of what we're doing with our recirculating cooling capabilities with our liquid cooling that we're doing at Switch and EVO. That EVO capability is stunning. I mean Jensen said it best, [ Mike Intrator ] said it best at [ CoreWeave ]. He did his entire IPO roadshow in a Switch data center. And when you see the ability to not waste one drop of water and you bring cooling at the level that we can do at Switch, which is the preferred location for the Blackwell chip for NVIDIA. It's the preferred location for CoreWeave, that's comparative advantage.

When you have an engineering capability where you can not only perform for your customer, but you can perform for the planet, I know that sounds a little folksy in today's environment, but I'm going to stand by that. I'm going to stand by my conviction that we have to be thinking long and hard about power and water.

Marc, thank you so much for your time today. Thank you.

Thanks, Mike. Appreciate it. Thanks.