Bloomberg separately reported that large public investors planned to shift more capital into private and less-liquid assets, citing an industry survey. Taken together, the reports describe a portfolio response to two pressures: the dominance of a narrow group of listed technology companies in public indices, and the need to finance data centres, power systems, private credit and infrastructure linked to AI demand.
Those pressures reinforce each other. If public equity indices become more concentrated in a handful of AI-linked companies, adding more listed exposure may increase rather than reduce portfolio risk. Private assets offer a route to different parts of the same investment cycle, from power supply and fibre networks to lending for data-centre construction.
The attraction is understandable. Public-market exposure to AI is concentrated in a small set of companies, while much of the physical build-out sits in assets that may never come to market as ordinary listed equity. A sovereign fund seeking long-duration returns can argue that infrastructure and private credit fit its mandate better than chasing the same public names at higher valuations.
The risk is also different. Listed equities can be sold daily at observable prices. Private infrastructure, private credit and venture-style AI exposure are marked less frequently and can become harder to exit when financing conditions tighten. A public investor can be right about the long-term need for data centres and still face a short-term liquidity problem if too much capital is locked into assets without easy buyers.
That matters because sovereign funds sit at the intersection of investment and public policy. Some are savings vehicles for future generations; others help stabilise budgets or manage commodity revenue. A liquidity mismatch is therefore not only a portfolio inconvenience. It can limit how quickly a state can draw on its wealth when fiscal conditions change.