The bond market is pricing in a future Wall Street still hasn't admitted out loud — and it's got nothing to do with rate cuts.
The Summary
- Yields climbing while stocks try to hold ground — the old inflation playbook doesn't explain what's happening now, according to Bloomberg's market overview
- Citi strategist Kate Moore remains bullish on equities despite yield pressure, while Carlyle's Mark Jenkins flags "hidden AI credit risks" that most investors haven't priced in
- Google ships Gemini 4 Argon and Micron posts strong AI-related earnings — the infrastructure build continues regardless of macro headwinds
The Signal
Yields are rising and nobody's panicking about inflation anymore. That's the tell. When Kate Moore from Citi says she's still betting on stocks while 10-year treasuries climb, she's not ignoring the bond market. She's reading it differently than the consensus. The question isn't whether the Fed will cut. It's whether traditional macro frameworks still apply when capital is flooding into AI infrastructure faster than it ever moved into railroads or electricity.
Micron's earnings beat shows where that capital is going. Data centers need memory. Lots of it. The AI buildout isn't a tech story anymore, it's a balance sheet story. Google doesn't roll out Gemini 4 Argon because they think the AI hype is fading. They ship it because if they don't, someone else will eat their search business from the bottom up.
"Hidden AI credit risks" means someone borrowed billions assuming AI revenue would scale linearly, and it won't.
Here's what Mark Jenkins from Carlyle is actually saying: the credit markets are pricing AI companies like they're all NVIDIA. They're not. Some are building picks and shovels. Some are building the railroads. And some are building luxury hotels in ghost towns. The debt load is real. The revenue assumptions are heroic. When yields rise, the companies with the thinnest margin between their AI pitch deck and their actual business model get exposed first.
This isn't 2000 because the technology is real this time. But it's also not 1995 because everyone already knows the technology is real. We're in the awkward middle phase where:
- Capital allocation to AI is still growing
- Business models are still largely unproven at scale
- Debt is cheaper than it should be given the uncertainty
- Investors are treating "we use AI" and "we sell AI infrastructure" as the same risk profile
The Implication
Watch credit spreads on AI-adjacent companies, not just equity valuations. If Jenkins is right about hidden risks, the bond market will price it in before the stock market does. That's your early warning system. For builders and operators, this is actually good news. Rising yields flush out the tourists. The companies that survive the next 18 months won't be the ones with the best demos. They'll be the ones with actual unit economics.
If you're allocating capital or making build-versus-buy decisions, model for a world where your cost of capital keeps rising and your AI vendor's Series C never happens. Own the infrastructure you can't afford to lose access to.