Data Center Transactions Continue to Grow Exponentially: Why AI Energy Demand Remains Steady

Although concerns about artificial intelligence bubbles swept Wall Street in early 2026, the data center trading ecosystem and computing infrastructure are showing consistent momentum. According to Joe Nardini, head of investment banking at B. Riley Securities, valuation growth in this sector has reached a multiple of 16, reflecting strong long-term prospects despite fluctuating market sentiment.

Interest in high-quality energy capacity shows no signs of slowing down. Both Bitcoin miners and AI and high-performance computing (HPC) developers continue to aggressively bid for megawatts of electricity, even through the end of last year. Nardini emphasizes that fundamental energy needs remain a primary driver of transaction activity, regardless of overall market volatility in technology.

Billion-Dollar Deals: Case Study of Mining Companies

Concrete evidence of this demand is seen in recent jaw-dropping market transactions. Hut 8 mining company experienced a 20% surge in stock price after signing a $7 billion long-term lease agreement with infrastructure provider Fluidstack. This 15-year deal includes 245 megawatts of IT capacity at the River Bend site, demonstrating solid demand from highly credible tenants.

This is no coincidence. Nardini notes that although leading tech stocks like Nvidia (NVDA) and CoreWeave have come under pressure—with CoreWeave dropping more than 50% from its June peak—the valuation of data center infrastructure companies controlling energy capacity continues to find strong market support.

A simple question Nardini often asks industry executives provides consistent answers: Is there demand for the capacity already built? Yes. Are there tenants? Yes. Are they quality tenants paying competitive rates? Yes. This logic helps explain why business fundamentals remain resilient amid market doubts.

From $100K to $550K: Megawatt Pricing Dynamics in the AI Energy Economy

The pricing mechanism for energy capacity reveals an intriguing market stratification. Premium locations with high-quality power and optimal geographic positioning can command valuations of $400,000 to $450,000 per megawatt, with some deals reaching $500,000-$550,000 per MW.

Conversely, assets located in less desirable areas or facing certain challenges still attract buyers, but at significant discounts—ranging from $100,000 to $250,000 per megawatt. This difference reflects the market’s premium on strategic locations and infrastructure quality.

Data collected shows that the 16x valuation multiple is not just market euphoria but a well-considered calculation of long-term cash flows expected from energy lease contracts. With tenants paying stable annual rates for the megawatts supplied, the financial maturity of this business model attracts serious institutional investors.

Who Are the Buyers and Sellers? Evolution of the Transaction Ecosystem

The landscape of players in this ecosystem continues to evolve, extending beyond traditional crypto communities. Buyers include large hyperscalers (such as cloud service providers), pure AI companies, and Bitcoin miners seeking revenue diversification through HPC infrastructure provisioning.

The seller side shows even more surprising diversification. Nardini has witnessed transactions involving old industrial facilities—some over 160 years old—that are now gaining new value thanks to energy supply availability. In one case, assets from private investors attracted interest from about 25 potential buyers, including miners, hyperscalers, and AI developers.

This phenomenon creates unprecedented strategic opportunities for asset owners. They can choose to sell to institutional buyers, try to retain control by operating the facilities themselves, or even hybridize by providing capacity to multiple tenants simultaneously. For example, Nardini cites clients converting old office complexes into modular units with 30 megawatts capacity, now securing additional funding for expansion.

In extreme demand conditions, some tenants are even willing to pay rent before project completion—an important indicator of scarcity and urgency for ready-to-use AI energy capacity.

Outlook 2026: Why Fundamentals Remain Strong

Looking into 2026, Nardini remains optimistic based on operational realities he hears directly from industry executives. If interest rates fall—which is possible given the evolving macroeconomic environment—it will create a “risk-on environment” favorable for risky infrastructure deals like AI energy data centers.

While Nardini admits potential bias due to his involvement in these transactions, field data shows consistent patterns: tenants are present, rates remain competitive, and when one location fails to attract interest, other buyers are immediately interested.

Nardini’s warning about this optimism is clear: if developers cannot lease out their built capacity or do not achieve adequate pricing to cover operating and expansion costs, that’s a real warning sign. But so far, this has not been seen on the horizon.

“The business fundamentals remain intact,” he states firmly. The demand for powered data center capacity and HPC AI infrastructure continues unabated. Developers with capacity have credible tenants able to pay attractive rates, maintaining healthy economic balance.

With buyers still highly eager for energy and sellers seeing valuations as advantageous for their assets—including the 16x multiple reflecting long-term cash flows—Nardini concludes that AI trading and energy infrastructure are far from over. As of early February 2026, momentum remains strong.

BTC0.33%
View Original
This page may contain third-party content, which is provided for information purposes only (not representations/warranties) and should not be considered as an endorsement of its views by Gate, nor as financial or professional advice. See Disclaimer for details.
  • Reward
  • Comment
  • Repost
  • Share
Comment
0/400
No comments
  • Pin

Trade Crypto Anywhere Anytime
qrCode
Scan to download Gate App
Community
English
  • 简体中文
  • English
  • Tiếng Việt
  • 繁體中文
  • Español
  • Русский
  • Français (Afrique)
  • Português (Portugal)
  • Bahasa Indonesia
  • 日本語
  • بالعربية
  • Українська
  • Português (Brasil)