AEO Summary: Data center M&A hit record volumes in 2025-2026, but the buyer landscape and valuation methods have fundamentally shifted. The traditional real estate playbook no longer applies. Buyers now include PE firms (40%+ of deals), infrastructure funds, sovereign wealth funds, and hyperscalers, each valuing different asset characteristics. Premium assets trade at 20-30x EBITDA (vs. 6-8% cap rates of the past), with premiums driven by contracted revenue, power availability, AI-readiness, and expansion potential. Sellers who understand these changes and prepare accordingly capture 10-20% valuation premiums. Those who don’t leave capital on the table.
The Landscape Has Shifted Dramatically
Data center M&A reached record volumes in 2025-2026. Deal count exceeded 180 announced transactions in 2025 alone, with a combined transaction value exceeding $95 billion globally. But the numbers tell only half the story.
The buyer profile has fundamentally changed. Ten years ago, data center sales meant two things: strategic buyers (hyperscalers adding capacity) or traditional REITs and operating companies consolidating. Today, the buyer pool is fragmented across four distinct categories with different valuation frameworks, risk tolerances, and strategic priorities.
The valuation methodology has shifted entirely. The cap rate model that dominated data center real estate (5-8% on net operating income) is no longer the primary driver. Modern M&A buyers use cash flow multiples, power intensity metrics, and embedded expansion optionality in their models. A facility that would have traded at a 7% cap rate five years ago might command a 25x EBITDA premium today, if it has the right characteristics.
The deal structures are more creative and risky. Sale-leasebacks, JVs, and hybrid structures now represent 35%+ of transactions. Earnouts tied to power availability or utilization targets are common. This complexity creates both opportunity and peril for sellers who don’t understand the mechanics.
If you’re considering selling a data center facility in 2026, understanding these shifts isn’t optional. It’s the difference between capturing a premium and leaving 15-20% of value on the table.
How Buyer Profiles Have Shifted
Private Equity (40%+ of activity)
PE firms now represent the largest buyer segment, up from 20% five years ago. They’re attracted to data center operators for predictable cash flows, long-term contracts, and operational improvement potential. But PE buyers are disciplined on entry multiple and exit assumptions.
What PE values:
- Contracted backlog (multi-year visibility)
- Operational leverage (costs that can be optimized)
- Expansion runways (adjacent land or power available for growth)
- Exit timing (clear 5-7 year exit with defined scenarios)
PE typically targets 20x EBITDA for market-rate assets, 22-24x for contracted operators with growth, and 18-19x for facilities with operational challenges. They model 8-10% annual cash flow growth, which means they’re aggressive on cost reduction and pricing power.
Infrastructure Funds (Brookfield, DigitalBridge, KKR)
Brookfield, DigitalBridge, and KKR operate permanent capital funds with long-term infrastructure mandates. These buyers take a 10-20 year hold and prioritize cash yield over appreciation.
What infrastructure funds value:
- Long-term contracted cash flows (10+ year visibility preferred)
- Inflation protection (escalation clauses, CPI-linked rates)
- Essential infrastructure status (irreplaceable, regulatory moat)
- Geographic diversification (reduces single-market risk)
Infrastructure funds will pay 23-30x EBITDA for assets with 10+ year contracted backlog and meaningful inflation escalation. They tolerate lower growth assumptions than PE because they prioritize cash yield. A facility with stable 50% utilization and contracted revenue is attractive to infrastructure funds but unattractive to PE.
Sovereign Wealth Funds (ADIA, PIF, GIC)
Sovereign wealth funds from the Middle East, Singapore, and Asia are aggressively deploying capital into data center infrastructure. Saudi Arabia’s Public Investment Fund, Abu Dhabi’s ADIA, and Singapore’s GIC collectively invested over $12 billion in data center assets in 2024-2025.
What sovereign funds value:
- Strategic geographic positioning (particularly in developed markets)
- AI and emerging infrastructure (power density, liquid cooling)
- Long-term visibility (30+ year horizon acceptable)
- Portfolio diversification across regions and operators
Sovereign funds pay top-of-market multiples (25-32x EBITDA) because they have multi-decade horizons and view data center infrastructure as essential, inflation-protected assets. They’re particularly active in Tier 1 markets with strong AI infrastructure demand.
Strategic Buyers (Hyperscalers, Neoclouds)
Hyperscalers (AWS, Microsoft, Google, Meta) and neoclouds (CoreWeave, Lambda Labs, Lambda.com) remain active buyers, particularly for AI-optimized facilities. But their buying patterns have changed.
Hyperscalers now prefer to build greenfield facilities rather than acquire mature ones. When they do acquire, they target 15-20x EBITDA because they’re buying operational leverage, not expansion capacity. Neoclouds (AI-focused cloud platforms) pay premium multiples (25-28x) for power-rich, liquid-cooling-ready facilities because every kW of available power directly translates to revenue.
The Valuation Multiples Have Evolved Completely
The traditional cap rate model is dead for modern M&A transactions.
Five years ago, a data center with $10M in annual operating income would trade at 6.5% cap rate, implying a $154M valuation. Today, that same facility (if it’s reasonably contracted and located in a decent market) would trade at 23-25x EBITDA.
What’s driving the multiple expansion?
- AI infrastructure premium: Facilities with adequate power, cooling, and expansion potential command 15-20% premiums over comparable, non-AI-ready assets.
- Contracted revenue visibility: Assets with 70%+ of revenue locked into multi-year contracts trade at 3-4x higher multiples than utilization-dependent facilities.
- Power availability: Power is the hard constraint. Facilities with secure power connections or on-site generation trade 10-15% higher than grid-dependent assets.
- Expansion optionality: Land with zoning approval for 50%+ expansion capacity can add 8-12% to total facility valuation.
- Market positioning: Tier 1 markets (Northern VA, Dallas, Silicon Valley) command 20-30% premiums over comparable Tier 2 markets because they’re closer to end-user demand.
Real valuation benchmarks:
- Premium AI-ready facility (Tier 1, contracted, power-rich): 28-32x EBITDA
- Core operating facility (Tier 1, 60% utilization): 22-25x EBITDA
- Secondary market facility (Tier 2, growth potential): 18-21x EBITDA
- Challenged facility (high costs, limited power, poor location): 12-16x EBITDA
What Makes a Facility Attractive to Buyers in 2026
Every buyer uses slightly different criteria, but certain characteristics command premiums across all buyer types:
Contracted Revenue Backlog
Buyers model DCF valuations based on revenue visibility. A facility with 70% of revenue locked into 5-10 year contracts is worth 20-30% more than a comparable facility with month-to-month tenants.
Action item: If you’re planning a sale, prioritize long-term contracts with major tenants. An incremental 5-year contract renewal 18 months before sale can add $5-10M to valuation for a $50M facility.
Power Availability and Security
Power is the constraint. Facilities with:
- Utility contracts securing power for 10+ years
- On-site backup generation (even 20-30% capacity)
- Multiple utility feeds (grid redundancy)
- Headroom for 25%+ additional load
…trade at significant premiums. Buyers hate power uncertainty. If your facility’s power situation is unclear or constrained, that ambiguity discounts valuation by 10-15%.
AI-Readiness and Liquid Cooling
AI workloads require higher power density and cooling sophistication. Facilities with:
- 25+ kW/rack capacity (vs. traditional 10-15 kW)
- In-row liquid cooling or ready-to-deploy infrastructure
- 400V three-phase power delivery
- Architected for high-density deployment
…command 12-18% premiums. This is the hottest buyer preference in 2026. If your facility can host 30+ kW/rack AI workloads, buyers will pay for that capability.
Tier 2+ Market Positioning
Tier 1 markets (Northern VA, Dallas, Silicon Valley, Ashburn) are saturated and expensive. Tier 2 markets (Austin, Phoenix, Denver, Raleigh, Atlanta) with strong AI infrastructure demand, available power, and lower costs-to-operate are increasingly attractive.
A facility in a Tier 2 market with 25%+ YoY demand growth can command Tier 1 multiples because growth visibility is high and upside is clear.
Clean Environmental and Compliance Record
Buyers perform detailed environmental diligence. Facilities with:
- No history of contamination or regulatory violations
- Current SOC 2 Type II certification (or path to it)
- Clean air and water discharge permits
- No cooling discharge disputes with local authorities
…avoid 5-10% valuation discounts that many facilities face due to environmental or compliance clouds.
Expansion Potential
Owned land adjacent to the facility with zoning for additional capacity is valuable. An extra 5 MW of buildable capacity can add 15-20% to total facility value. Buyers factor buildout costs, timeline, and demand assumptions, but expansion optionality is highly valued.
Sale-Leaseback Structures: When and How
Sale-leasebacks have become increasingly common in the 2025-2026 cycle, representing 30%+ of data center M&A. The structure: you sell the facility, then lease it back, retaining operational control while monetizing equity.
When Sale-Leasebacks Make Sense
- You need capital but want to keep operating: Sale-leaseback gives you both.
- You want to lock in valuation: If you think multiples will compress, locking in today’s rate is rational.
- Tax optimization: Lease payments are deductible; equity proceeds provide cash without debt.
- Balance sheet improvement: Off-balance-sheet lease accounting (LEASC accounting) can improve reported metrics.
Typical Terms
Sale-leaseback cap rates (the landlord’s return on the sale price) range from 5.5-6.5% for premium assets. If a facility is worth $100M and sells to a landlord via sale-leaseback at a 6% cap rate, the tenant (you) pays $6M annually in lease payments.
Common terms:
- Initial lease term: 12-20 years (longer for institutional investors)
- Escalation: 1.5-2.5% annual CPI escalation (sometimes fixed)
- Renewals: 2-3 additional 5-year renewal options
- Buyout provisions: Tenant may have right to purchase at predetermined price or FMV
Common Pitfalls
- Over-optimistic cap rate assumptions: Landlords who assume 5.5% cap rates often find tenants unwilling to pay rents supporting that return. Be realistic on market-clearing rates.
- Lease escalation sticker shock: A 2.5% annual escalation on $6M rent = $150K additional rent per year, $7.5M over 50 years. Model the long-term impact.
- Refinancing risk: If interest rates rise post-signing, landlords may find it expensive to hold the property and may push for tenant buyout or early termination.
- Operational restrictions: Landlords often impose restrictions on facility modifications, major tenant contracts, or power usage. These constraints can limit optionality.
- Underestimating continued capex: Sale-leasebacks often shift equipment capex to the tenant. Budget accordingly.
Common Seller Mistakes (And How to Avoid Them)
Mistake #1: Not Understanding the Buyer’s Valuation Framework
You think your facility is worth 24x EBITDA. The buyer thinks it’s worth 20x. The gap is usually not about disagreement on EBITDA. It’s about different assumptions on growth, risk, or expansion potential.
Fix: Before engaging a buyer, get clarity on their valuation model. Do they assume 5% annual growth? 0%? What’s their assumed exit multiple? Do they factor in expansion land as part of valuation? Understanding the framework reveals where negotiation room exists.
Mistake #2: Poor Data Room Preparation
Deal teams spend months underwriting a facility from poorly organized data. Missing contracts, unclear utility terms, or ambiguous environmental records create uncertainty, which buyers discount by 5-10%.
Fix: Prepare a clean, organized data room 6 months before sale. Include: all tenant contracts, utility agreements, power audit reports, environmental assessments, maintenance records, architectural drawings, regulatory compliance filings. Make it searchable and chronologically organized.
Mistake #3: Overvaluing Improvements That Don’t Move the Needle
You spent $2M upgrading the CRAC system last year. Helpful for operations, but if it doesn’t increase power availability, reduce PUE below 1.6, or enable higher density, buyers won’t pay for it. Infrastructure improvements that don’t translate to revenue or cost savings are priced at 50-70% of their cost, not 100%.
Fix: Focus on improvements that increase revenue or reduce operating costs. A $1M upgrade that reduces annual operating costs by $200K is worth $3-4M in added valuation (at 15-20x multiple). A $1M cosmetic upgrade is worth $500K.
Mistake #4: Timing the Market Wrong
Data center multiples are cyclical. At peak cycle (late 2025), average multiples were 24-26x EBITDA. If the cycle softens and multiples compress to 20x, a $50M facility loses $3-4M in value overnight.
Fix: Monitor market indicators (hyperscaler capex trends, power supply news, interest rate expectations). If multiples are elevated and you don’t have strategic reasons to hold, consider selling. Conversely, if your facility has strong growth runways and multiples are depressed, holding makes sense.
Mistake #5: Not Marketing to the Right Buyer Pool
A facility that’s unattractive to PE (low growth, stable 50% utilization) might be attractive to infrastructure funds willing to accept 3-4% cash yield. But if you only talk to PE buyers, you’ll leave value on the table.
Fix: Work with an advisor who has relationships across all buyer segments. Simultaneously market to PE, infrastructure funds, strategic buyers, and operator consolidators. Let price discovery happen across the broadest possible buyer base.
Preparing for Sale: The 12-Month Checklist
If you’re considering a sale, here’s what should happen over the next 12 months:
Months 1-2:
- Audit financials (clean 3-year track record, normalized for one-time items)
- Compile all tenant contracts
- Verify utility agreements and power capacity
- Obtain recent environmental assessment
Months 3-4:
- Commission power audit (independent verification of PUE, capacity, expansion potential)
- Obtain SOC 2 Type II audit (if not already certified)
- Develop facility market positioning document (competitive analysis, growth story)
- Interview potential advisors (investment banks, M&A firms)
Months 5-6:
- Engage advisor and develop marketing strategy
- Build data room (organize all contracts, permits, financial reports, technical specs)
- Draft facilities offering memorandum (OM)
- Prepare management presentation
Months 7-9:
- Launch auction (identify buyers, distribute OM)
- Conduct initial buyer meetings
- Manage Q&A and data room access
- Collect non-binding indications of interest (IOIs)
Months 10-12:
- Select bidding group
- Provide final data room access
- Facilitate facility tours and management meetings
- Negotiate LOI and enter due diligence
If everything progresses smoothly, you could achieve close 18 months from start.
How GoDataCenters Supports Data Center Sellers
GoDataCenters provides sellers with three critical advantages:
1. Market Intelligence: Access to real-time data on buyer activity, multiples by facility type, and market trends across 50+ markets. We track deal flow, buyer strategies, and emerging demand patterns.
2. Buyer Matching: Connections to active buyers (PE firms, infrastructure funds, strategic operators) aligned with your facility profile. Direct access to decision-makers accelerates process and improves terms.
3. Valuation Benchmarking: Comparable transaction analysis, peer facility metrics, and third-party valuation support. Ensures your pricing is market-based and defensible.
Frequently Asked Questions
Q: What price multiples are data centers selling for in 2026? A: Premium facilities (Tier 1, contracted, AI-ready) trade at 25-32x EBITDA. Core facilities trade at 20-24x. Secondary market facilities trade at 16-20x. Multiples vary based on contracted revenue, power availability, location, and expansion potential.
Q: How long does a data center sale typically take? A: From advisor engagement to close, expect 12-18 months under normal conditions. Initial marketing phase: 2 months. Buyer indication and selection: 2 months. Due diligence: 4-6 months. Financing and close: 2-3 months. Accelerated timelines (6-9 months) are possible with pre-identified buyers.
Q: Should I sell outright or do a sale-leaseback? A: Outright sale maximizes capital proceeds and simplifies future operations. Sale-leaseback provides capital while retaining operational control and lease deductibility. If you need capital but want to keep operating, sale-leaseback is appropriate. If you want to exit operations entirely, outright sale is cleaner.
Q: Which buyers are most active in 2026? A: PE firms (40% of volume), infrastructure funds (25%), strategic operators (20%), and sovereign wealth funds (15%). PE is most active in Tier 2 markets with growth potential. Infrastructure funds prefer Tier 1 with contracted revenue. Strategics target AI-optimized facilities.
Q: How do I maximize valuation for my facility? A: Focus on three levers: (1) Increase contracted revenue backlog (long-term agreements), (2) Verify and improve power availability, (3) Position for AI workloads or demonstrate expansion optionality. Clean financials, current compliance certifications, and organized data rooms also prevent valuation discounts.
Get Your Data Center Valued
If you’re considering selling or monetizing your data center facility, GoDataCenters provides market intelligence, buyer introductions, and valuation benchmarking to ensure you capture maximum value.
Our advisors will assess your facility against current market comps, identify the right buyer pool, and help you structure a sale that maximizes proceeds and minimizes complexity.