UBS worst-case AI disruption scenarios:
– Private credit: 13% defaults
– Leveraged loans: 8%
– High-yield bonds: 4%
PC most exposed because that’s where 2021-2022 vintage software deals got funded when banks couldn’t syndicate.
The concentration is now becoming the vulnerability.
UBS worst-case AI disruption scenarios:
– Private credit: 13% defaults
– Leveraged loans: 8%
– High-yield bonds: 4%PC most exposed because that's where 2021-2022 vintage software deals got funded when banks couldn't syndicate.
The concentration is now becoming the… pic.twitter.com/HTMt6M0vw0
— junkbondinvestor (@junkbondinvest) February 2, 2026
jeroen blokland
@jsblokland
Keep a close eye on this chart in the years ahead. Private debt is back.
Artificial intelligence is real. So is the pressure to participate. Companies are being pushed, by markets, investors, and a wide range of stakeholders, to commit vast sums to AI. Much of that spending is being financed through newly issued debt.
And this is not limited to the hyperscalers. Judging by the frequency with which AI is mentioned in earnings calls, every company now faces the same binary choice. Invest in AI, or risk being left behind.
During the dot-com era, investment enthusiasm was widespread. Profits were not. Not every company that invested heavily in technology was able to translate those investments into durable returns. This time won’t be different.
As a result, investor attention is likely to broaden. Today, the focus is still overwhelmingly on government debt. Over the next year, or perhaps the one after that, private debt will increasingly move into the spotlight as well.
There is a crucial distinction. Unlike sovereign governments, companies do not issue a currency, and do not possess a printing press in the form of a central bank (central bank independence is a myth). If you conclude here that this makes public debt less of a problem, you are dead wrong. Private debt, however, will lead to a financial crisis more quickly.
I do not believe we are in an AI bubble. At least not yet. What we are witnessing, however, is the early stage of an AI-driven debt bubble.
When investor focus shifts from exuberance to balance sheets, companies that finance AI investments largely out of operating cash flows while maintaining conservative debt levels will outperform.
Low leverage is a defining characteristic of scarce quality companies. The debt accumulation taking place today is likely a crucial source of under- or outperformance tomorrow.
Keep a close eye on this chart in the years ahead. Private debt is back.
Artificial intelligence is real. So is the pressure to participate. Companies are being pushed, by markets, investors, and a wide range of stakeholders, to commit vast sums to AI. Much of that spending is… pic.twitter.com/PN36Qo4nbk
— jeroen blokland (@jsblokland) February 3, 2026