Obsessed with CoreWeave and the AI trade? You're missing out on some great stocks

Jim Cramer warned that CoreWeave’s heavy borrowing could constrain its data-center expansion and indirectly weigh on Nvidia, while arguing investors should not overlook opportunities beyond AI infrastructure. He cited J&J shares up 2.2%, Lilly up 3.5% on research suggesting potential GLP-1 pill earnings upside, and GE Aerospace orders up 49% despite shares falling 3.7%. He also highlighted 3M’s expanded Amazon relationship, including a warrant for up to 3.2 million 3M shares tied to $6 billion in cumulative sales, and noted Korean neocloud Elice raised $150 million in an equity offering before its shares fell 32%.

Analysis

The useful distinction is not “AI versus non-AI,” but funded capacity versus capacity dependent on continued access to credit. A fall in benchmark rates alone does not resolve neocloud risk: refinancing spreads, debt maturities, utilization, and contracted customer revenue matter more. If funding tightens, some demand may migrate to balance-sheet-rich hyperscalers rather than disappear, potentially supporting AMZN, GOOG, MSFT, and META while pressuring CRWV and other capital-intensive builders. That substitution could preserve aggregate GPU demand—and NVDA sales—so a neocloud failure is not automatically an Nvidia demand collapse; the key risk is whether customer concentration and financing stress impair deployments broadly.

The 3M–Amazon arrangement is strategically relevant but not yet proof of material earnings contribution. The warrant ties potential dilution to cumulative sales, making product qualification, ramp pace, and realized economics more informative than the announcement itself. Likewise, favorable commentary on JNJ, LLY, GE, and RTX is not a substitute for checking guidance, order conversion, and valuation: a strong order headline can disappoint if delivery schedules or margins lag. Over the next 1–3 months, monitor credit spreads and refinancing terms for neoclouds, hyperscaler capex commentary, and evidence that 3M’s data-center products move from agreement to revenue. Over 6–18 months, the likely divergence is between firms able to fund infrastructure internally and those reliant on leverage.

Key Decisions for Investors

• Keep CRWV on a financing-risk watchlist rather than treating lower Treasury yields as a buy signal. Reassess only after reviewing debt maturities, borrowing costs, collateral structure, contracted utilization, and liquidity; widening credit spreads, delayed capacity deployment, or weaker customer commitments would strengthen the downside case. A credible refinancing and sustained utilization would falsify it.
• Do not short NVDA mechanically against CRWV: a neocloud funding shock could shift workloads to hyperscalers and leave aggregate accelerator demand intact. Track NVDA order commentary alongside AMZN, GOOG, MSFT, and META capex guidance; broad deployment cuts, not one borrower’s stress, are the stronger bearish trigger for NVDA.
• Treat the 3M–Amazon relationship as a catalyst to verify, not an earnings estimate. Monitor product qualification, shipment timing, cumulative sales toward the warrant threshold, and dilution economics before adding MMM on this news; lack of disclosed ramp or meaningful revenue evidence would argue against paying for the narrative.
• For sector diversification, favor a measured review of JNJ, LLY, GE, and RTX on estimate revisions and valuation rather than assuming recent strength is durable. For GE and RTX specifically, test whether orders convert into deliveries and cash flow; slowing conversion or weaker guidance would invalidate the operational-strength thesis.