New York: The corporate bond market is increasingly splitting into two camps, with investors showing greater caution towards debt issued by AI-related companies while continuing to enthusiastically back more traditional issuers such as financial and industrial firms.

According to portfolio managers and market analysts, concerns are not centred on the creditworthiness of major technology companies. Instead, investors are focused on the unprecedented scale and uncertainty of borrowing required to finance data centres, semiconductors and AI infrastructure.

Goldman Sachs estimates that gross debt issuance by hyperscalers could reach a record $420 billion in 2027, representing a 60 per cent increase compared with 2026 forecasts.

By comparison, overall US corporate bond issuance rose 30 per cent year-on-year to $1.9 trillion through August, according to data from the Securities Industry and Financial Markets Association.

"We're being very selective in terms of how we invest within hyperscaler debt," said Colby Stilson, Head of Fixed Income at Brown Advisory.

"Our degree of investment conviction needs to be very high because of the coming supply and because of the lack of visibility into that return on invested capital," he said.

Investors look beyond AI

Market participants say many investors are increasingly directing money towards sectors outside the AI ecosystem.

Loren Moran, fixed-income portfolio manager at Wellington Management, pointed to recent pharmaceutical and insurance acquisition financings that attracted strong investor demand with little or no pricing concessions.

While investors still have substantial cash available to deploy, many are seeking opportunities away from large technology issuers heavily exposed to AI spending.

The contrast was evident in recent debt offerings.

Google parent Alphabet reportedly had to offer significant pricing concessions during an August bond sale, while insurance broker Aon's $13.5 billion acquisition financing attracted approximately $65 billion in orders.

Strong demand enabled Aon to tighten pricing on its 30-year bonds by 35 basis points, highlighting investor appetite for less heavily supplied sectors.

AI bonds trade at wider spreads

As borrowing by hyperscalers, chipmakers and data-centre operators continues to increase, investors have demanded wider spreads to absorb the growing supply.

According to Goldman Sachs data, AI-related issuers are trading at spreads of around 115 basis points, compared with 78 basis points for the broader investment-grade market, according to ICE BofA figures.

Lon Erickson, portfolio manager at Thornburg Investment Management, said debt issued by companies such as Meta Platforms and Alphabet has consistently traded at wider levels than similarly rated peers, despite their strong balance sheets and cash generation.

The additional yield reflects expectations that these companies will continue to return to the bond market as AI-related capital expenditure grows.

"Investors are only able to digest so much, so fast," Erickson said.

Concentration concerns emerging

Beyond pricing, investors are becoming more conscious of concentration risk.

Some institutional investors are approaching internal limits on exposure to individual companies once debt issued through related structures, including data-centre financing vehicles, is combined with existing holdings.

Analysts say many investors also want flexibility in case enthusiasm for AI investments begins to cool.

Rather than taking large positions today, some are opting to preserve capital in anticipation of potentially higher yields in future issues.

Nick Elfner, co-head of research at Breckinridge Capital Advisors, noted that some hyperscaler deals have attracted lower levels of demand than investors typically expect from large, high-profile issuers.

He added that several transactions have traded poorly after issuance, contributing to investor caution.

Credit quality remains strong

Despite the growing selectivity, market participants emphasise that concerns are focused on supply rather than financial strength.

Russell Brownback, Deputy Chief Investment Officer for Global Fixed Income at BlackRock, said wider spreads among AI-related companies reflect basic supply-and-demand dynamics rather than deteriorating credit quality.

Companies continue to borrow because they believe returns from AI investments will exceed financing costs, while investors are receiving yields more commonly associated with lower-rated issuers despite lending to highly rated companies.

For now, bond investors appear less concerned about the ability of hyperscalers to repay debt and more focused on managing exposure amid an unprecedented wave of AI-related borrowing.

"There are a lot of investors that just want something other than hyperscaler debt for now," said Wellington's Moran. "The market is a bit starved for anything ex-hyperscaler."