Alphabet’s $5.5 billion inaugural Australian bond issue has brought the growing financing needs of AI infrastructure into focus for Australian fixed income investors, according to Capital Group.
The company behind Google and Gemini has become the first hyperscaler to tap the Australian dollar debt market. It is also the first large-cap US tech firm to issue a “Kangaroo” bond – a foreign bond issued in Australian dollars by non-domestic entities – since Apple’s 2016 bonds, which reached maturity earlier this year.
The deal has highlighted the emergence of AI as a major new source of bond supply, with debt issuance rising sharply, particularly in the US.
Gross US investment-grade issuance was running 31 per cent above last year and at a record annual pace, while Capital Group estimates hyperscalers alone could issue US$100-150 billion of bonds annually over the next few years to finance new data centres.
Investment director Haran Karunakaran said the Alphabet deal showed the AI financing boom was extending beyond US credit markets.
“The Alphabet deal is significant because it shows the AI financing boom is no longer just a US credit-market story. The hyperscalers need enormous amounts of capital and they are increasingly looking across currencies and markets to raise it.”
Despite record issuance, broader credit markets remained resilient, with US investment-grade and high-yield spreads tightening over the 12 months to 31 July. Alphabet’s Australian issue reportedly attracted around $18 billion of demand.
However, Capital Group has been monitoring signs of ‘AI fatigue’, including weaker oversubscription for some recent deals and difficulties placing some AI-related financing in smaller credit markets.
“So far, investment-grade markets have digested the extraordinary amount of new supply remarkably well. The risk is that at some point investor appetite fails to keep pace with issuance. We are watching closely for signs that weakness in AI-related credit starts spilling into the broader market,” Karunakaran said.
The surge in supply has also pushed some hyperscaler debt from trading at a premium to the broader investment-grade market to a material discount, potentially creating opportunities in high-quality issuers.
“There is an interesting paradox developing. The sheer amount of borrowing required for AI is a risk for credit markets, but it is also creating opportunities to lend to some exceptionally strong companies at yields and spreads that are increasingly attractive,” Karunakaran said.
He said credit investors should focus on the quality and structure of individual opportunities rather than simply buying into the AI theme.
“For credit investors, this shouldn’t be about buying the AI theme. The important questions are who you are lending to, what the debt is financing, how the deal is structured and whether you are being adequately compensated for the downside risk.”