Major institutional investors, including pension and sovereign wealth funds with billions to invest, are finding that the traditional diversification strategies they once relied on are becoming an illusion due to the widespread integration of artificial intelligence across all asset classes. Monte Tarbox, Chief Investment Officer of the $327 billion New York City Retirement Systems (NYCRS), recently passed on a promising private equity fund due to its heavy AI holdings, not out of existential dread, but because his primary job is diversification, and he sees AI exposure everywhere he looks.
This challenge stems from AI's deep penetration into various market segments. For instance, Goldman Sachs reports that AI infrastructure companies account for approximately 40% of the S&P 500's total market capitalization, and shares of just three semiconductor manufacturers make up over a quarter of the emerging markets index. Apollo Global Management notes that AI contributes to nearly half of all investment-grade bond issuance this year and 87% of venture capital funding. An additional risk is posed by "circular financing schemes" where large tech companies finance each other, creating interconnected dependencies.
Assessing AI exposure is complex because there's no single definition of what constitutes AI-related risk; it can range from dedicated chip manufacturers to companies merely integrating the technology. This makes it difficult for funds to track their true AI concentration. Invesco's survey of 90 sovereign wealth funds revealed that over half cited market concentration as the main risk associated with AI investments. Lisa Shalett, CIO at Morgan Stanley Wealth Management, cautions that the current AI investment cycle will inevitably end, similar to past tech booms, leaving investors exposed if not properly diversified.
Some funds are trying new approaches, like the Finnish Elo Mutual Pension Insurance, which uses AI-based tools to identify companies sensitive to AI developments within its portfolio. Despite concerns, investments in AI have been highly profitable, with the Bloomberg Intelligence index of AI-related companies outperforming the broader market by an average of 11 percentage points annually. However, this success means that the largest pension funds, on average, have an AI exposure close to zero, although some individual funds, like the California Public Employees' Retirement System (CalPERS), show significantly higher exposure, particularly through private investments. This has led some institutions to adopt a "Total Portfolio Approach" to gain a holistic view of their assets and better assess aggregate risk factors like AI.
