So I'm trying to learn about credit markets. Was curious as to why AI buildout borrowing (est. $500B in 2026) has itself seen their credit spread has increased this year (lenders demanding more from them to borrow more as costs seem to get out of control), but that hasn't spilled over to credit spread increases for Genpop Joe's Tool & Die. Joe's credit spread hasn't budged an inch. Which would make one think that lenders really aren't all that worried. But then I did this:
📊 The Top Annual Net New Borrowers
Rank Country Annual Net New Borrowing (Est. 2026 Fiscal Deficit)
- 1 United States ~$1.9 to $2.0 Trillion
- 2 China ~$850 Billion
- ?? MSFT, META, GOOG, AMZN, ORCL ~$500 Billion
- 3 United Kingdom ~$110 to $130 Billion
- 4 France ~$100 to $115 Billion
- 5 India ~$95 to $105 Billion
- 6 Italy ~$80 to $95 Billion
- 7 Japan ~$75 to $90 Billion
- 8 Germany ~$60 to $75 Billion
- 9 Brazil ~$55 to $65 Billion
- 10 Canada ~$30 to $40 Billion
Of course this isn't impacting corporate borrowing spreads... Microsoft, Amazon, Google, Meta and Oracle (that's all that 500B is, btw, no one else) aren't competing with Joe's Tool & Die for funding... they're competing with the European Union! That's not going to show up in the corporate spread... that's going to show up in the underlying risk-free (treasury) yields. ie. EVERYONE's debt. US, EU, Joe's Tool & Die, and your fucking mortgage.
This hasn't been playing out for years. The debt portion of the AI buildout in 2025 was tiny on a world scale. The new country called AI-Buildout was morphed into existence and plopped into the center of world credit markets (more or less) in the last few months. It's the 3rd biggest country on the planet, and next year, will easily become the 2nd.
For a product that's OPEN SOURCE. And is gauranteed to get LOTS more hardware capacity coming online because the USA doesn't want to rely on a chip factory sitting on China's doorstep... EU doesn't want to rely on China or the US, etc. Lots more capacity kills margins.
Soooo... this pretty much has to blow up within the next year... if we're very lucky. Because the explosion doesn't scale linearly. JPM writing down 20% of their loans is far worse than 2x the effect of writing down 10%.
2008 wasn't caused by housing... it was caused by credit. And while it wasn't true to say it was the case in 2025, it's becoming the case in 2026, and will abso-fucking-lutely be the case in 2027.