Google's Negative Cash Flow and the AI Capex Reckoning | The Weekly Wrap

Summary

In this Weekly Wrap, Steve Eisman says the AI narrative has changed from enthusiasm over growth to concern about capital intensity, weak moats, and the risk of a price war after Moonshot’s cheaper KimiK3 launch. He emphasizes that Google’s strong cloud growth was overshadowed by negative free cash flow and a larger 2026 AI capex plan, while Tesla also saw margins and free cash flow pressured by rising spending. Eisman then walks through a series of company results, highlighting solid defense backlogs, GE Vernova’s power-demand surge, Moody’s durability, IBM’s weakness, ServiceNow’s resilience despite the SaaS-apocalypse narrative, and Blackstone’s strong fundraising alongside longer liquidity timelines. He closes by arguing that bank stocks are not true diversification away from AI because financing activity is tied to the same capex cycle, and he gives a simple mechanical explanation of using a short sale to fund a long position.

Notes

• Moonshot’s KimiK3 is framed as a cheaper Chinese LLM that could spark AI price competition.
• Eisman says the AI story has shifted from upside excitement to moat compression and capital-intensity risk.
• Google posted strong headline growth, but Eisman strips out investment gains to focus on operating reality.
• Google Cloud revenue reached $24.8 billion, up 82%, showing the business still grows quickly even as capex swells.
• Google’s free cash flow turned negative $5.9 billion because of AI spending.
• Google lifted 2026 AI capex from $190 billion to $205 billion, reinforcing investor anxiety around the scale of infrastructure buildout.
• Tesla’s revenue beat, but margins were hit by lower regulatory credits and heavier capex.
• Tesla’s capex rose to $5.8 billion and free cash flow fell to negative $1.1 billion.
• Tesla’s free cash flow turned negative for the first time in two years.
• IBM’s revenue outlook was cut to 4% to 5% growth, implying a weaker second half than hoped.
• IBM mainframe sales fell 42% and infrastructure revenue declined 7%.
• ServiceNow still grew quickly, but the stock sold off anyway under the SaaS-apocalypse narrative.
• ServiceNow’s revenue reached $4 billion, up 24%, despite investor skepticism about SaaS multiples.
• Blackstone beat on earnings and performance fees, helped by strong realizations.
• Blackstone raised $68 billion in fundraising, showing continued capital inflow despite broader exit concerns.
• Eisman sees longer hold periods and slower liquidity returns as a private equity bottleneck.
• Private credit’s software exposure becomes more relevant when refinancing begins next year.
• Moody’s delivered strong earnings growth, but Eisman still notes the valuation is not cheap.
• Moody’s EPS rose 31% year over year to 468, showing continued earnings momentum.
• Eisman treats GE Vernova as a power-infrastructure beneficiary of the AI buildout.
• GE Vernova orders jumped 88% to $24.2 billion, with backlog at a record $176 billion.
• GE Vernova’s backlog and order flow suggest multi-quarter demand visibility.
• Northrop Grumman beat expectations and added about $20 billion to backlog, reaching a record $105 billion.
• Lockheed Martin delivered 11% sales growth and a record $230 billion backlog.
• Defense backlog strength is a recurring theme, with both Northrop and Lockheed showing long-dated demand.
• Eisman says AI-driven data center demand is straining power equipment supply and electricity infrastructure.
• Eisman views Intel’s beat as evidence of demand, but not as reassurance on AI capex spending.
• Intel’s data center segment sales soared 59% year over year.
• Even strong AI-supplier results do not resolve investor concern about the size of hyperscaler budgets.
• Eisman argues trust banks are difficult to isolate as a clean equity theme.
• Large-cap bank strength is tied to AI financing needs, so banks are not diversification away from tech.
• Eisman describes the basic short-sale mechanic as borrow, sell, repurchase, and return.
• He also frames a funding short as a way to finance a preferred long position.