SoftBank’s DigitalBridge Deal: Terms, Strategy & AI Infrastructure Impacts

On December 29, 2025, SoftBank Group struck a definitive agreement to acquire DigitalBridge, a global digital infrastructure asset management firm, in an all-cash deal valued at approximately US$4.0 billion, inclusive of debt. The per-share offer of $16.00 implies a ~15% premium to DigitalBridge’s closing share price just before the announcement and ~50% above its unaffected 52-week trailing average as of early December 2025. DigitalBridge manages approximately US$108 billion in assets across data centers, fiber networks, cell towers, small cells, and edge infrastructure, with operations spanning North America, Europe, the Middle East, and Asia.

SoftBank’s acquisition is clearly aligned with its intensified push into AI infrastructure—its “physical AI” ambitions. Key supporting pillars include its recent divestment of its Nvidia stake (≈ US$5.8 billion), rising investments in OpenAI, and involvement in the Stargate initiative with Oracle, which aims to deliver large-scale infrastructure and compute capabilities. DigitalBridge fits this thesis by owning and managing assets fundamental to the growth and deployment of AI workloads: power, connectivity, real estate, and edge. SoftBank’s leadership under Masayoshi Son frames this move as part of its long-term AI platform strategy.

For investors, until the deal closes—expected in the second half of 2026 pending regulatory and shareholder approvals—DigitalBridge shares trade in a narrow band below the offer price. This creates a merger arbitrage scenario: buying at prices such as ~$15.30–$15.40 implies a spread of about $0.60–$0.70 per share (~3.8-4.4% absolute return), which annualizes to roughly 6-10% for a 6- to 9-month holding period.

However, several strategic and execution risks are non-trivial. The transaction must pass customary regulatory hurdles, including U.S. antitrust and possibly CFIUS scrutiny, given the critical infrastructure nature of DigitalBridge’s assets and foreign company involvement. Once the offer is public, the stock’s upside is capped at that price, which shifts investor risk/reward toward deal execution rather than organic growth. Also, the energy and capital intensity associated with scaling data centers and infrastructure could strain returns if demand or pricing for AI infrastructure weakens or competition intensifies. The timeline to closing remains uncertain but is currently projected for H2 2026.

Strategic implications go beyond the immediate transaction. SoftBank secures ownership over a core asset base essential to AI’s “power problem”—availability of electricity, land, fiber, and real estate capacity. It also gains access to DigitalBridge’s LP relationships and fee-earning infrastructure, diversifying its income beyond equity returns. For DigitalBridge, going private under a well-capitalized owner removes public market pressures and supports longer-duration capital commitments. For the broader digital infrastructure space, this deal further underscores a consolidation trend: players controlling power and compute capacity are being priced at significant premiums, and merger arbitrage becomes a common investment consideration in the sector.

Open questions: Will regulatory authorities impose conditions (e.g. around asset ownership, national security)? What is the risk of overcapacity in AI infrastructure, particularly in power supply and cooling? Can SoftBank sustain value chain control economically—for example, balancing third-party capital vs balance sheet exposure? Finally, the return for arbitrageurs depends materially on deal timing—any delay or complication could erode returns significantly.