Switch in talks to raise funds at $50 billion-plus valuation, The Information reports

Switch is reportedly in talks to raise billions at a valuation of at least $50 billion, with Brookfield, KKR and other investors in the mix. The round could position the data center developer for an IPO as early as next year, supported by Goldman Sachs and JPMorgan. The news underscores rising AI-driven demand across the data center and server industry.

Analysis

The actionable read-through is not the headline valuation itself, but the signaling effect for capital allocation into AI infrastructure: private capital is still willing to mark up hard assets with long-duration, contracted demand even before public markets reprice the sector. That tends to compress the cost of capital for adjacent data-center owners, power contractors, and equipment suppliers, while raising the bar for any standalone cloud or colocation asset that cannot prove land, power, and interconnect scarcity.

The most important second-order effect is competitive: a large raise at this size could force regional data-center peers to either accelerate expansion or risk being locked out of the best power corridors. Over the next 6-18 months, the bottleneck is less compute demand than site readiness—grid interconnects, cooling, and permitting—so beneficiaries are the firms that control entitlement pipelines and financing, not just rack density. That favors diversified capital allocators over single-asset operators.

For public-market positioning, the near-term catalyst is the optionality around an IPO window, which can re-rate the whole AI infra basket if priced as a scarcity asset rather than a utility-like REIT. The contrarian risk is that $50B private marks are usually the point where execution risk shifts from construction to refinance risk: if rates stay sticky or demand growth normalizes, the market can quickly punish any asset with embedded capex and long payback periods. In that scenario, the next leg is not broad AI beta but selective dispersion between financed winners and over-levered builders.

Key Decisions for Investors

• Long BAM / KKR vs short high-beta data-center adjacencies for 3-6 months: own the capital providers with multiple ways to win from primary investment, fee growth, and follow-on deployment; avoid paying up for names whose returns depend on flawless build-out execution.
• Buy a basket of AI infrastructure enablers on dips over the next 1-2 weeks: GS and JPM benefit from financing/refi and M&A activity, while NVDA gets indirect support from capacity expansion; use this as a relative-value expression rather than outright sector leverage.
• Pair trade: long public alternatives managers / private capital platforms, short expensive REIT-like growth proxies that trade on AI narrative alone; the thesis is that private valuation marks will stay strongest where underwriting and control over capital are best.
• If an IPO process is confirmed, fade first-day enthusiasm in the issuer and use listed peers as the cleaner trade; the risk/reward favors buying post-lockup or on post-pricing weakness rather than chasing the initial print.
• Set a 30-60 day watchlist for power, cooling, and electrical gear names tied to data-center buildouts; those usually monetize the capex wave earlier than the data-center operator itself and have less multiple compression risk if public-market sentiment cools.