CoreWeave’s Bonds Flash Red as Strong Earnings Fall Flat

Dear readers,

Yesterday’s question was who has the pricing power to defend a valuation. Today’s is harder: who can still pay for the buildout, and at what price? The first serious crack in the AI trade has appeared in credit rather than demand. The Nasdaq lost nearly 2 percent on Wednesday, and the bond market is now asking about balance sheets. Bank, agrochemical and consumer-goods earnings point the same way. Good results alone no longer carry a stock.

CoreWeave: The AI Boom Has a Financing Problem

CoreWeave closed Wednesday at $65.78, down 15 percent on a little more than triple its normal volume. In premarket trading Thursday it slipped another 6 percent or so. That puts the year’s low of $60.55, set on July 29, back within reach. Over twelve months the stock is down a little more than 50 percent.

The real alarm is coming from the bond market. The yield on CoreWeave’s notes due 2031 climbed above 16 percent, and the notes due 2032 fell below 80 cents on the dollar.

Credit default protection now costs roughly 1,000 basis points, which implies a default probability of about 63 percent over five years. A year ago the figure was around 30 percent.

Behind those numbers sits a business carrying roughly $50 billion of debt. CoreWeave expects only about $13 billion of revenue in 2026 but plans capital spending of $35 billion to $39 billion.

The Gartner warning we cited yesterday sharpens the picture. The research firm sees more than $3.5 trillion of AI write-offs by 2030, and over 80 percent of new cloud providers failing or being acquired.

Demand is not the problem. Anthropic reported third-quarter revenue of $16.8 billion and reached profitability on an adjusted basis. Analysts are split on CoreWeave itself: Deutsche Bank rates it a buy with a $150 price target, while Bernstein stays at Underperform. Nvidia owns about 10 percent of CoreWeave and is directly exposed. Nebius fell more than 4 percent in premarket trading as well.

Anyone holding AI infrastructure should now check bond yields and leverage alongside order books.

Banks: Strong Numbers, Weak Shares

Most of the big US banks beat profit expectations, and their shares fell anyway. Equities trading is booming while fixed income lags. Morgan Stanley lifted equities trading revenue 41 percent to $5.80 billion, while bond trading fell 7.3 percent. Earnings per share came in at $3.03 against expectations of $2.91.

Bank of America was weighed down by a $1.2 billion penalty from the US Treasury. Reported earnings per share were $0.91, or $1.08 adjusted, just under expectations.

The stock lost 4.8 percent on Wednesday to $51.39 and is down 14 percent over 30 days. RBC cut its price target from $65 to $60 but kept its Outperform rating. Baird went the other way, upgrading the stock from Neutral to Outperform with a $62 target.

Goldman Sachs beat expectations by a wide margin. Investors fixated instead on its provisions for credit losses, which more than tripled from $102 million to $368 million. JPMorgan missed profit expectations for the first time since the fourth quarter of 2023.

The market is looking past trading profits and asking how clean the loan books are. That matters most if AI financing starts to wobble.

Should investors sell immediately? Or is it worth buying CoreWeave?

FMC: When Inventory Becomes a Balance-Sheet Problem

FMC plunged between 35 and 46 percent on Thursday, depending on the time of trading. The agrochemical group slashed its third-quarter guidance. Instead of $840 million to $900 million, it now expects revenue of just $65 million to $125 million, excluding its India business. Adjusted EBITDA is projected at minus $565 million to minus $510 million, against a previous plan of plus $120 million to $140 million.

The cause is charges of $625 million to $675 million. FMC is cutting prices in Brazil, Argentina, the US and Canada to work down expensive inventory held by distributors. A possible impairment of intangible assets looms on top. The company says it met its loan covenants as of September 30 and is negotiating adjustments with its banks.

Full results are due October 29, when we will learn whether clearing the channel is enough or the balance sheet needs repair. FMC is a lesson for investors: a revenue hole from destocking can quickly turn into a financing problem when debt is high.

Consumer: Better Than Feared, Cautious Outlook

The consumer-goods giants are beating expectations but failing to convince on outlook. PepsiCo posted third-quarter adjusted earnings of $2.34 per share against expectations of $2.29, and revenue rose 5.6 percent to $25.27 billion. Yet it lowered its guidance.

Adjusted earnings per share are now expected to grow just 2.5 to 3.5 percent in 2026. The company had previously aimed for the low end of a 5 to 7 percent range. North American beverage volume fell 2 percent, and the turnaround there is taking longer than planned. The stock was still up about 2 percent in premarket trading.

Pernod Ricard reported an organic revenue decline of 0.9 percent to €2.23 billion in its first fiscal quarter. Analysts had expected a drop of 1.7 percent. Revenue in China collapsed 25 percent and fell 10 percent in the US, while India grew 19 percent.

According to the CFO, destocking in China and the US will likely last the entire fiscal year. The shares trade at just over €66, little changed from the previous day. Pernod expects to reach only the low end of its medium-term growth target of 3 to 6 percent.

Kimberly-Clark confirmed its full-year outlook after preliminary figures. Disruptions on social media in China are holding back organic growth by about 100 basis points. Final results arrive November 5.

For now, consumer staples offer stability but little upside.

Rocket Lab: A Delay That Costs Price Targets

Rocket Lab is pushing the first launch of its Neutron rocket from late 2026 to 2027. The cause is manufacturing defects in the first-stage fuel tanks. The company says the problem is fixed, but the investigation cost time.

KeyBanc cut its price target from $135 to $100 and kept its Overweight rating. Cantor Fitzgerald reaffirmed Overweight with a $122 target. The stock recently traded around $59, far below its 52-week high of $151. For growth stocks without profits, every schedule slip feeds straight into the valuation.

The Takeaway

The coming weeks will show whether credit worries stay confined to CoreWeave or spread to other AI providers. The next checkpoints are the Fed meeting on October 28, FMC’s results on October 29 and Kimberly-Clark’s on November 5. The market’s test has moved from pricing power to balance sheets, and it will favor companies where earnings and debt tell the same story.

Best regards,
The StocksToday.com Editorial