AI spending raises earnings volatility concerns; high-quality 3–5-year bonds draw focus

Heavy AI investment and elevated borrowing costs are putting pressure on corporate profits, raising the prospect of sharper share-price swings around earnings releases. In fixed income, high-quality bonds with maturities of three to five years offer a way to capture yields while limiting sensitivity to interest-rate changes.

In an October 26 interview with David Lin, David Bush, chief investment officer at Trajan Wealth, said investors were demanding evidence of returns on AI spending. He also highlighted opportunities to secure yields on high-quality bonds as part of portfolio allocation.

Bush said rising corporate earnings were supporting equities despite the pressure higher interest rates place on valuations. Several companies reporting results had exceeded analyst estimates and delivered upbeat guidance, he said.

He identified the three- to five-year segment of the bond market as the most attractive, citing the greater price risk associated with maturities of 10 to 30 years. A 30-year Treasury with a duration of 20 to 25 could experience an estimated price move of 20% to 25% in the opposite direction for a one-percentage-point change in interest rates, he explained.

Bush said his firm was focusing on investment-grade corporate bonds, Treasuries and agency securities while avoiding riskier areas of the market. With yields at multi-decade highs, he argued that the portfolio mix of 60% stocks and 40% bonds had regained relevance: bonds could provide income and help cushion a substantial equity-market decline.

In AI-related equities, Bush said he was watching companies adopting the technology to improve profitability and efficiency. He described technology companies and chipmakers as the first phase of the AI trade, followed by infrastructure construction. That second phase has benefited utilities, materials and industrial companies, while the third would depend on how effectively businesses use AI.

For large technology companies, however, substantial capital spending creates earnings pressure. Bush said increased borrowing at higher interest rates was intensifying scrutiny of revenue growth and AI monetization. Companies that miss earnings estimates or cut guidance face swift market penalties, he said, making results announcements a source of volatility.

Bush nevertheless viewed infrastructure investment as necessary to compete globally over the longer term. Drawing on the dot-com era, he asked which AI competitors would become Google and which would become Netscape Navigator. Competition involving China means the eventual winners remain uncertain, he said.

A slowdown in AI infrastructure spending would also require a reassessment of the sectors benefiting from it, Bush said. Some traditional dividend-paying utilities, materials and industrial companies were behaving more like growth businesses. Without the AI spending impetus, their other projects and businesses would need closer examination.

He pointed to CoreWeave as an example of the credit risks surrounding the buildout. Comparing its bond yields with rating-based yield curves, Bush said they stood almost 500 basis points, or five percentage points, above even the CCC curve. He interpreted that gap as the market pricing in greater default risk.

Bush said equity-like bond returns come with equity-like risk. He described bond markets as an early indicator of corporate financial stress, underscoring that the high yields associated with AI infrastructure financing also reflect substantial credit risk.