Sweden’s EQT clocks strong H1 as AI investments grow

The market should read this less as a one-quarter beat and more as evidence that private capital can still monetize the AI buildout without needing public-market euphoria. For listed alternatives managers, that matters because the fee pool is driven by scarce deployable assets, not just generic fundraising; EQT’s signal is strongest for firms with infrastructure, energy-transition, or data-center access points. The second-order beneficiaries are the adjacent suppliers with real capacity bottlenecks — power equipment, grid, cooling, and construction — but only if capital deployment keeps converting into projects rather than remaining “dry powder” headlines.

The main risk is that higher-for-longer rates keep exits and realizations weak even as AUM rises. That creates a trap where investors pay up for headline asset growth while fee-generating earnings lag for multiple quarters; the catalyst path is the next 1-2 quarters of fee conversion and the Q3 close/integration of Coller, not today’s print. If H2 fundraising slows or investment pace normalizes, the current AI infrastructure premium can compress quickly.

Contrarian view: consensus may be underestimating how crowded AI infrastructure capital is becoming. If every large private-markets platform crowds into the same data-center and power stack, entry valuations rise and future vintage returns deteriorate, which caps long-term carry even as current fee revenue looks strong. That argues for buying weakness rather than chasing momentum, and for treating NAV growth in the AI vehicle as a funding signal, not proof of durable compounding.