Currents: Ep323 – Data Center Financings: Part II

Podcast by Norton Rose Fulbright | November 6, 2025

Main Theme & Purpose

This episode features an expert panel discussion on the evolving landscape of data center financings. Drawing from a recent client webinar, industry leaders from Rabobank, DC Blocks, SocGen, and Rowan Digital Infrastructure unpack current trends, lender structures, capital recycling, and whether the surge in data center growth constitutes a “bubble.” The conversation is moderated by Norton Rose Fulbright partners Christine Brzezinski and Christy Rivera.

Panelists & Moderator Introductions

• Klaus Hertel (Rabobank): Head of Energy & Core Infrastructure Coverage for North America, overseeing power, renewables, and digital infra (esp. data centers).
• Mellie Elleri (DC Blocks): Chief Investment Officer, specializing in hyperscale data centers in the U.S. Southeast.
• Tim McGuire (Rowan Digital Infrastructure): VP, Capital Markets; manages investment evaluation, risk, and deal structuring for hyperscale developments.
• Mike Johnson (SocGen): Director, TMT Finance and Advisory; covers construction, leveraged finance, ABS/CMBS, and private placements in digital infra.
• Moderators: Christine Brzezinski & Christy Rivera, Project Finance Partners, Norton Rose Fulbright New York.

Detailed Breakdown of Discussion

1. Market Volatility and Financing Environment

(05:00–08:00)

• Volatility mostly affects renewables and regulatory risk, while the data center space remains “fairly stable.”
• Deal flow is at record highs: Investment-grade characteristics (including ratings) draw a broader set of investors—including banks, institutional capital, and private credit.
• Financing costs: Remain “fairly stable” (low to mid-200s over SOFR).
• Biggest challenge: How much capacity markets and lenders can absorb as deal size and frequency increase.
• Capital recycling noted: Established projects refinancing into ABS/CMBS will free up lender balance sheets.

Quote:
"My question is, how does the market absorb all this paper? ... Portfolios of projects can be refinanced to the ABS market or the RBS market. So that can free up balance sheet capacity for existing lenders."
— Klaus Hertel (07:10)

2. Hyperscaler Activity & AI CapEx

(08:07–11:30)

• No pullback on AI-related data center investments; if anything, demand is accelerating.
• Large recent Meta/Blue Owl bond deal ($20bn+) well received by the market.
• Lender pool is broadening, with banks and institutions comfortable with the “core infrastructure” status of hyperscale data centers.

Quote:
"The appetite from the hyperscalers for projects that can deliver and energize by 2027 is just very, very strong."
— Tim McGuire (08:37)

• Deal size escalation: Previously $600m transactions, now billion+ handled without disruption.
• Cautious portfolio management: Lenders monitor concentrated exposures as deal sizes grow.

3. Current Lender Mix and Deal Structure

(11:30–14:25)

• More pension, infra, and sovereign wealth funds entering the space.
• Early ratings for construction financings are unlocking new investor classes earlier in project life cycles.
• Still a hybrid between real estate and project finance structures, but project finance “discipline” dominates for major new builds.

Quote:
"Construction financing [is] double over the past couple of years... as additional projects get bigger and more come to market, we're going to need to broaden out space as well."
— Mike Johnson (12:30)

4. The Role of Completion Guarantees

(14:25–19:33)

• Completion guarantees rooted in real estate legacy: Now “inherited” by project finance for data centers, even if the asset class is less risky than traditional infra (like power).
• Guarantees provide comfort for banks, especially since data center market has limited long-term loss data and “fully wrapped” contracts are not always present.
• Trend expected: Guarantees may decline as market matures and lenders grow more comfortable.

Quote:
"It's more of a historical thing than anything else on the merits of the, the, you know, the completion guarantee thrown in on a typically non-recourse project finance structure."
— Mellie Elleri (15:40)

5. Non-Recourse and Land Acquisition Financing

(19:33–22:51)

• Standard: post-completion financing is fully non-recourse to the sponsor, except for discrete triggers in mixed/warehouse facilities.
• Land acquisition inclusion depends on timing. If land and pre-development work are complete, often included in the financing budget. Early land acquisition sometimes needs separate (often expensive) facilities.

6. Real Estate vs. Project Finance Classification

(22:51–25:08)

• Shift to project finance: 90% of major data center construction now using PF approach.
• Still flexible: Borrowing base or DEVCO loans can support asset pools, strategic land, and power/infrastructure needs, reminiscent of corporate real estate methods.

7. Financing Early-Stage Development & Power

(25:08–28:07)

• Early-stage capital (equipment, power deposits) is unlocked via DEVCO structures or revolving facilities—critical as power is often the development bottleneck.
• Lenders want to see power agreements and a clear path to tenanted leases.
• Market focus: Tier-1 or strong growth markets get preference.

8. Average Deal Size and Tenor

(28:07–32:28)

• Deal sizes escalating: Underwritten bank commitments often $200–$500m for lead banks, $100–$150m for other leaders, and $15–$75m for final “buy-and-hold” participants.
• Typical tenor: 4+1 years (with extensions tied to performance/completion), but varies by facility type and project stage (supply, development, warehouse, PF, capital markets).

Quote:
"There's a lot of people out there who want to write checks in the, call it 15 to $75 million range... At the end of the day someone's actually got to want to hold this paper."
— Tim McGuire (29:55)

9. Hyperscale vs. Colocation: Financing Considerations

(32:28–36:50)

• Colo assets: Benefit from location, more users (customer diversification), but get lower leverage and price at a premium.
• Hyperscale: Driven by single, investment-grade tenants on 15–20yr leases, looked at as “core infrastructure” with robust, predictable cash flows.

Quote:
"Are you financing an infrastructure asset or a niche real estate asset? ... In the case of the hyperscale customer ... those cash flows are pretty close to bulletproof once the development's been completed."
— Tim McGuire (35:20)

10. Data Center vs. Renewables: Pricing Differentials

(36:50–40:48)

• Why price wider than renewables? Mainly the “sheer amount of capital” needed for data centers (e.g., $20bn deals), requiring more banks (sometimes up to 100 participants).
• Renewables long seen as more mature, but contain complex risks (e.g., tax equity, curtailment) not relevant to data centers.
• Data center tenants can support higher pricing for speed and reliability of delivery.

11. Supply Chain, Leases, and Payment Structures

(40:48–47:10)

• Developers usually supply equipment (unless it’s a powered shell, in which the tenant may bring their own).
• Planning horizon: 2-year construction timeline typical; aligns with when tenants begin pre-leasing.
• Lease structures:
  • Power shell: Tenant fits out, triple net lease (tenant pays for OPEX, taxes, insurance, utilities).
  • Turnkey: Developer invests in and operates mechanical/electrical plant—can follow triple, double, or modified gross net structures.

12. Debt Sizing and Lease Renewal Assumptions

(47:10–51:28)

• Lenders may grant credit for 're-leasing' (e.g., assuming renewal of 5-year options after a 15-year base lease)—especially for strong tenants in top markets.
• Holistic view: Tenant credit, market tier, and structural mitigants all factor into how much future (uncontracted) revenue counts in debt sizing.

13. Are We in a Bubble? (Round Robin)

(51:28–58:00)

• Unanimous panel response: “Not a bubble.”
  • Demand growth is underpinned by essential everyday digital activity and only partly from AI (currently <15% of demand but rising fast).
  • Data center tenants for AI are large, profitable, diversified companies.
  • Cloud and infrastructure is “integrated into everything we do” and under intense, ongoing demand pressure.
  • Some admit there could be “bubblicious pockets,” but fundamentals are strong and developer/bank financial commitments reinforce confidence.

Quotes:
"No. The everyday use of compute is essential ... AI is just starting to scratch the surface."
— Mike Johnson (52:18)

"There might be some bubblicious pockets ... The cloud services business is really booming ... the backlog of unfulfilled revenue ... is in the hundreds of billions."
— Tim McGuire (54:00)

"We are putting our money where our mouth is. ... We believe in this growth. ... Our actions are backing that up."
— Mellie Elleri (56:40)

"AI is not going anywhere. ... I just don't see a substitute right now for the foreseeable future. ... I don't see this as a bubble at all."
— Klaus Hertel (57:49)

Memorable Moments

• [08:25] Tim McGuire: On Meta’s $20bn bond and the ongoing “megadeal” appetite.
• [13:45] Klaus Hertel: On rating agencies starting to assign IG ratings prior to project completion.
• [22:51] Mellie Elleri: On 90%+ of data center construction liquidity trending into PF structures.
• [32:41] Mike Johnson: On the nuanced differences in risk and capital pool between colo and hyperscale assets.
• [51:42] Mike Johnson: On AI’s share of data center demand and market resilience.

Key Takeaways

• Data center project finance is rapidly maturing, attracting a broader and more diverse investor base.
• Megadeals and institutional buy-in are driving deal sizes and liquidity to new highs, but with prudent risk management.
• Leasing, credit support, and contract structuring are all evolving, and banks are getting comfortable with more forward-looking credit assumptions.
• Market fundamentals appear strong, with ongoing demand growth for both traditional compute and AI; most panelists vocally reject the “bubble” narrative.

This summary reflects the language, themes, and tone of Ep323 of Currents by Norton Rose Fulbright, providing a comprehensive synthesis for industry professionals seeking insight into state-of-the-art data center financing trends.