Betsy Cohen sees SPACs as the on-ramp for AI data center companies going public

Betsy Cohen argues that SPACs offer a viable pathway for AI data center companies to access public markets more efficiently than traditional IPOs. However, investors must carefully evaluate execution risks and the historical underperformance of SPAC-backed companies before committing capital.

SPACs have emerged as an alternative capital-raising mechanism for infrastructure-heavy businesses seeking rapid public market access. AI data center operators face substantial upfront costs for compute hardware, real estate, and power infrastructure, making traditional IPO processes—which demand multi-year profitability metrics and conservative projections—restrictive. SPAC mergers compress timelines and allow companies to present growth narratives that standard IPO roadshows might scrutinize more heavily. Cohen's perspective reflects growing recognition that infrastructure underpinning the AI boom requires flexible financing options.

The SPAC landscape has matured significantly since 2020-2021 when blank-check acquisitions dominated headlines. Sponsor quality, regulatory oversight, and investor sophistication have increased, though underperformance remains endemic to the sector. AI data center firms present a structurally different proposition than typical SPAC targets: they offer tangible assets, contracted revenue from hyperscalers, and transparent unit economics. These fundamentals differ markedly from software or biotech SPACs that often faced execution shortfalls.

For investors, this represents both opportunity and caution. Data center valuations remain elevated as AI workload demand intensifies, but SPAC mechanics introduce dilution from sponsor promotes and warrant structures that reduce returns. The market impact hinges on whether AI data center SPACs attract institutional capital that enforces accountability or retail-driven bubbles that inflate valuations beyond fundamentals. Regulatory scrutiny of SPAC disclosures has tightened, potentially reducing fraud risk but also increasing compliance costs for companies and sponsors.

• →SPACs compress go-to-market timelines for capital-intensive AI data center infrastructure compared to traditional IPOs.
• →Data center operators possess contracted revenues and tangible assets that differentiate them from problematic SPAC targets of prior cycles.
• →SPAC dilution mechanisms including sponsor promotes and warrants reduce investor returns despite faster market access.
• →Institutional investor participation and regulatory oversight quality will determine whether AI data center SPACs create sustainable value.
• →Investor due diligence on sponsor track record, power supply contracts, and hyperscaler commitments remains critical.