The All-In Hosts Asked If the AI Trade Is Cracking. Here's What Income Investors Should Do

CoreWeave shares fell 22.88% over the past week to $61.90, while its 2032 bonds were cited as yielding 15% versus 5.09% for the 10-year Treasury. The company reported Q2 revenue growth of 112.3% to $2.575 billion and a $104 billion backlog, but $6.42 billion in capex drove free cash flow to -$5.74 billion and interest expense rose to $640 million. The article also highlights Procter & Gamble and ADP, whose fiscal 2026 free cash flow covered dividends about 1.58 and 2.0 times, respectively; both yields are below the Treasury rate. CoreWeave’s debt risk is balanced by bullish views on AI demand and a Goldman Sachs upgrade, while Meta’s results are flagged as a near-term indicator of hyperscaler spending.

Analysis

CoreWeave’s stress is a financing signal, not yet proof that AI demand is rolling over. The key second-order risk is that higher funding costs force leveraged compute providers to slow GPU purchases before end-customer demand weakens; that would transmit from credit spreads into equipment orders and, eventually, hyperscaler capacity availability. Conversely, longer useful GPU lives would support collateral values and ease refinancing pressure—but that remains an underwriting assumption until utilization, resale values, and cash conversion validate it.

The next 1–3 months hinge on whether credit weakness stays issuer-specific and whether Meta’s results confirm that hyperscaler capex is producing enough demand to sustain the buildout. A capex slowdown or weaker monetization would pressure the entire AI infrastructure chain; resilient spending with improving utilization would argue against extrapolating CoreWeave’s distress. Over 6–18 months, refinancing access and free-cash-flow conversion matter more for CoreWeave equity than headline backlog: contracted demand is not equivalent to cash available after investment and interest.

Contrarian point: after a steep equity decline, shorting CRWV outright risks a violent relief rally if lenders remain open and GPU economics hold. Also, PG and ADP may offer earnings stability, but yields below the cited Treasury rate make them poor near-term substitutes for cash income; their appeal is dividend growth and lower operating volatility, not current yield. No evidence here establishes that either stock is attractively valued.

Key Decisions for Investors

• Keep CRWV exposure small and avoid treating backlog as a proxy for debt-service capacity. Reassess after its next report: watch interest expense versus operating income, free-cash-flow trend, debt pricing/refinancing access, and evidence on GPU utilization and useful life. A further deterioration in bond prices or weaker cash conversion would falsify the stabilization case.
• Use Meta’s upcoming results as a near-term AI demand catalyst, focusing on capex plans alongside monetization and returns—not capex alone. A cut or weaker expected return on investment would favor reducing broad AI-infrastructure exposure; sustained spending with improving monetization would weaken the bear case.
• Prefer a relative-risk approach to an outright CRWV short: if credit stress broadens while Meta’s demand indicators remain firm, consider a small, risk-limited CRWV-underweight versus META position. The mismatch can persist or reverse sharply; exit if CRWV credit stabilizes and cash conversion improves, or if Meta signals weaker spending economics.
• For defensive income allocation, do not rotate from Treasury bills solely into PG or ADP for yield. Consider them only for equity exposure to dividend growth and operating resilience; ADP remains exposed to payroll volumes in a recession, while PG’s cost pressures could constrain near-term margin progress.