Nvidia’s share price has fallen nearly 30% since mid-October, but the price of its bonds maturing in 2036 has barely budged—the bond and stock markets are pricing AI risk as if they were worlds apart.
Over the past week or so, credit spreads on bonds issued by cloud giants like Alphabet, Amazon, and Meta have narrowed, as have spreads on high-yield data center bonds. The bond market’s logic is simple: if data center construction slows in the future and companies shift their focus to preserving cash flow, that could actually make existing assets and debt safer.
This kind of divergence between stocks and bonds is nothing new. It happened during the dot-com bubble in 2000, too: stocks plunged, while investment-grade bonds stayed rock-solid. Creditors care about whether a company can repay its debt; shareholders care about the story of future growth. When the market starts questioning that growth narrative, stocks often crack first, while bonds can hold up for a while longer.
Here’s where things stand now: the AI infrastructure boom may have peaked, but the cash flows and balance sheets of these big tech companies are still healthy. The bond market is saying: even if the AI story falls apart, these companies won’t be unable to repay their debt. The stock market is saying: but if growth expectations are gone, valuations need to be reassessed.
As the old saying goes, the bond market is smart money, while the stock market is hot money. Right now, the bond market is reminding everyone that the AI bubble may be bursting, but we haven’t reached the point of systemic risk yet.
Over the past week or so, credit spreads on bonds issued by cloud giants like Alphabet, Amazon, and Meta have narrowed, as have spreads on high-yield data center bonds. The bond market’s logic is simple: if data center construction slows in the future and companies shift their focus to preserving cash flow, that could actually make existing assets and debt safer.
This kind of divergence between stocks and bonds is nothing new. It happened during the dot-com bubble in 2000, too: stocks plunged, while investment-grade bonds stayed rock-solid. Creditors care about whether a company can repay its debt; shareholders care about the story of future growth. When the market starts questioning that growth narrative, stocks often crack first, while bonds can hold up for a while longer.
Here’s where things stand now: the AI infrastructure boom may have peaked, but the cash flows and balance sheets of these big tech companies are still healthy. The bond market is saying: even if the AI story falls apart, these companies won’t be unable to repay their debt. The stock market is saying: but if growth expectations are gone, valuations need to be reassessed.
As the old saying goes, the bond market is smart money, while the stock market is hot money. Right now, the bond market is reminding everyone that the AI bubble may be bursting, but we haven’t reached the point of systemic risk yet.