J.P. Morgan says CoreWeave debt selloff poses limited risk to global banks
J.P. Morgan estimates most global banks have limited exposure to CoreWeave, even after its stocks and bonds slumped last week on debt worries.
- published on Newsmetry
- original source published
the public page opens no earlier than 15 minutes after we detect the story. approved stories may reach the api sooner. connect the api →
what happened
J.P. Morgan estimates that most global banks have limited exposure to CoreWeave. Large US investment banks and Japan's MUFG could hold roughly $1 billion to $3 billion of exposure, while most European and Asian lenders have under $500 million. Much of it comes through delayed draw term loans, which are mostly held in special-purpose vehicles and released as projects reach milestones.
Last week CoreWeave's stocks and bonds fell on investor worries over debt financing. The yield on its notes due 2031 briefly reached almost 16%, and the cost of insuring its debt rose to around 1,000 basis points. The company has about $50 billion in debt.
J.P. Morgan said its figures rest on simplified assumptions, since banks disclose little about their exposure to the AI investment cycle. The main risk, it said, is the mismatch between upfront AI infrastructure spending and revenue received over time.
why it matters
Debt loads at AI infrastructure companies are increasingly a concern for credit markets. Investors will watch how quickly CoreWeave can reach positive free cash flow, which J.P. Morgan expects to stay negative until 2030.
market impact
Banks and high-yield AI infrastructure debt: direct risk seems limited, but sensitivity to debt concerns remains
tags
instruments
sources
- CoreWeave debt selloff poses limited risk to global banks, J.P. Morgan says — Reuters, 20 October 2026 at 07:20 UTC
prepared by the newsmetry newsroom with ai assistance from the sources listed below. not investment advice.
the public page opens no earlier than 15 minutes after we detect the story. approved stories may reach the api sooner.
get stories in real time ↗