CoreWeave paid $846M in interest last quarter.
Its adjusted operating income was $379M.
That gap is why the stock and the bonds are telling two different stories.
The stock side:
• Revenue $3.66B, up 168% in a year
• $132.5B backlog of signed contracts
• Full-year guidance raised to $13.0–13.3B
The stock fell about 47% in the month before earnings. Then it jumped about 20% the day after.
The bond side:
• Its 9.625% bonds due 2032 yielded over 16% the week before earnings. A 10-year Treasury pays about 5%.
• Interest expense: $846M, up from $310M a year ago
• Next quarter's guide: $1.05–1.15B in interest
• Capex plan for this year: $37–39B, and a lot of it is borrowed
Why they disagree:
A stockholder needs the AI buildout to keep going. If that backlog turns into cash, the upside is big.
A bondholder's best case is getting paid back. Their worry is refinancing: CoreWeave borrows to buy GPUs, and those GPUs have to earn enough before the next debt comes due.
And Microsoft was about two-thirds of 2025 revenue.
So 16% isn't a bet that CoreWeave goes bust. It's the price of lending to a company that has to keep borrowing to grow.
What to watch:
→ The ~$9B in customer prepayments management expects over the next 3 months
→ Whether the bond yield drifts back toward mid-October's ~14.5%
→ Q4 interest landing near $1.1B
My take: the business is real. The balance sheet is the risk, and bond traders are pricing that more honestly than the stock chart.
Stock or the 16% bond, which would you rather own?
Its adjusted operating income was $379M.
That gap is why the stock and the bonds are telling two different stories.
The stock side:
• Revenue $3.66B, up 168% in a year
• $132.5B backlog of signed contracts
• Full-year guidance raised to $13.0–13.3B
The stock fell about 47% in the month before earnings. Then it jumped about 20% the day after.
The bond side:
• Its 9.625% bonds due 2032 yielded over 16% the week before earnings. A 10-year Treasury pays about 5%.
• Interest expense: $846M, up from $310M a year ago
• Next quarter's guide: $1.05–1.15B in interest
• Capex plan for this year: $37–39B, and a lot of it is borrowed
Why they disagree:
A stockholder needs the AI buildout to keep going. If that backlog turns into cash, the upside is big.
A bondholder's best case is getting paid back. Their worry is refinancing: CoreWeave borrows to buy GPUs, and those GPUs have to earn enough before the next debt comes due.
And Microsoft was about two-thirds of 2025 revenue.
So 16% isn't a bet that CoreWeave goes bust. It's the price of lending to a company that has to keep borrowing to grow.
What to watch:
→ The ~$9B in customer prepayments management expects over the next 3 months
→ Whether the bond yield drifts back toward mid-October's ~14.5%
→ Q4 interest landing near $1.1B
My take: the business is real. The balance sheet is the risk, and bond traders are pricing that more honestly than the stock chart.
Stock or the 16% bond, which would you rather own?
