Much has been written about the winners and losers from the artificial intelligence (AI) theme, but most of that attention remains focused on the equity market. We think investors should increasingly be looking to diversify. As companies commit vast sums to AI infrastructure, the story is becoming as much about financing as innovation. The next phase of the AI boom may not be determined by who builds the best model, but by who funds it. Investors appear to still be believin’ in the AI story, but perhaps it is time they paused to question how it is being paid for.
The scale of capital being deployed is remarkable. Hyperscalers, cloud providers and a growing list of adjacent businesses are undertaking capital expenditure programmes unlike anything the technology sector has attempted before, with capex forecast to exceed US$1.1 trillion in 2027.
What is notable is how much of this investment is increasingly financed through debt markets rather than internally generated cash flow. Companies historically defined by their cash generation, such as Alphabet, are turning to capital markets in size, and doing so specifically to protect credit ratings that until recently seemed unshakeable.
A growing share of investment-grade bond issuance is now tied, directly or indirectly, to the AI buildout. Concentration is building not only in equity indices, but in credit indices as well. For investors who still believe in the diversification benefits of holding equities alongside investment-grade credit, the reality is that they may increasingly own the same underlying theme in two different forms.
Our expectation is not a dramatic credit event, but a gradual repricing as the market comes to terms with the volume of financing required and the concentration of exposure now sitting within credit portfolios. Investors should be cautious about assuming that traditional fixed income allocations still offer the diversification benefits they once did.
We think selected convertible bond opportunities continue to offer attractive asymmetry, particularly in Asia and Europe where issuance has been less dominated by AI-related names. Emerging-market local currency debt continues to offer real yields that stand out in a world where inflation-adjusted returns are otherwise scarce. Meanwhile, in seeking consistent income, we looked to securitised credit, where floating-rate characteristics and attractive spreads continue to offer a differentiated source of return in an environment where broad credit looks expensive.
Within equities, we remain focused on regions and sectors trading well away from the crowded end of the market, including US healthcare, financials and industrials.
The AI story is not over. However, the way investors need to think about the theme is changing. It is no longer just about earnings growth and semiconductor demand. The AI story is increasingly about balance sheets, financing conditions and the choices companies make when the cost of building the future is greater than the cash flow they generate today.
Dorian Carrell, Head of Multi-Asset Income, Schroders
