AI boom: Tech companies finance investments with hidden debt

The AI ​​boom is consuming billions - and tech companies are increasingly financing it with debt. At the same time, previously barely visible obligations are rising to around $2.6 trillion, as a new study shows. Experts warn of growing risks to creditworthiness.

According to a study, the boom in artificial intelligence is increasingly being financed with hidden debt. The long-term debt of the eight leading US technology companies rose by 86 percent within a year, according to a published study by the credit insurer Allianz Trade. At the same time, their off-balance sheet obligations, especially for data centers, energy supply and AI infrastructure, rose from $573 billion to around $2.6 trillion within a year.

If these obligations, which have so far only been partially visible, are taken into account, the actual debt burden of companies increases on average by almost 150 percent. As a result, the credit quality of the corporations deteriorates mathematically by one or two rating levels.

“The tech industry has so far avoided the iceberg, but navigation is becoming more difficult,” said Alexander Hirt, an expert at Allianz Research. "However, the real development takes place below the surface of the water: Obligations that do not yet appear on the balance sheets today will gradually become actual liabilities in the coming years."

According to the study, the credit markets have already reacted. The risk premiums, the so-called spreads, for bonds from large technology companies have more than doubled within a year. However, Allianz Trade assumes that the risks have only been partially priced in. “Despite the increasing burdens, the immediate risk of failure for tech companies remains low,” explained Hirt. "The greater risk lies in a continued widening of spreads and thus a gradual reassessment of creditworthiness by the capital markets."

According to the study, the results are also relevant for the German economy. Many companies rely on the infrastructure of large US providers and are planning billion-dollar AI investments themselves. In addition to traditional debt, investors should also take long-term infrastructure and supply contracts into account when analyzing risk, Hirt advised.