The Stock Market Is Totally Wrong About AI Stocks
Speakers
Summary
- The S&P 500 has shed $2 trillion in market cap in 48 hours, driven by a panic over AI Capital Expenditure (CapEx).
- Google (Alphabet) is targeting $185 billion in CapEx for 2026 alone (vs. analyst estimates of $120 billion), effectively doubling its spend to chase scaling laws.
- A massive divergence exists: "Mag 7" companies are reporting record earnings and cloud growth (Google Search +17%, Gemini 750M users), but stock prices are crashing (Microsoft -21%, Nvidia -11%) due to margin compression fears.
- The speakers argue the market is "totally wrong"; the $680 billion combined CapEx from hyperscalers is justified by demand and scaling laws ("dollars in, intelligence out").
- Apple is the short-term outlier (+7%), outperforming solely because it has zero AI CapEx exposure, though speakers view it as "washed up" for missing the innovation cycle.
- Coding agents (ClaudeCode, OpenAI Codex) are viewed as deflationary threats to traditional SaaS and Law firms, with the SaaS index down 10-15% on fears of automation.
Trade Ideas
Google reported record earnings (Search +17%, Cloud up) but the stock is crashing because they announced $185 billion in CapEx for 2026. Microsoft is down 21% and Amazon is increasing CapEx despite beating expectations. The market interprets high CapEx as a risk to margins/free cash flow. However, the speakers argue this is a mispricing: scaling laws are holding, and demand for compute is infinite (Sundar Pichai cannot meet demand). The massive spend is necessary to secure future dominance, and current fundamentals (revenue/users) are stronger than the price action suggests. LONG. The market is "wrong" about the AI bubble popping; this is a buying opportunity on high-quality compounders aggressively investing in their future. Scaling laws hit a wall (diminishing returns on "dollars in, intelligence out") or supply chain gluts occur.
Hyperscalers (Google, Microsoft, etc.) have committed to ~$680 billion in combined CapEx for 2026. Nvidia stock is down 11% alongside the broader tech sell-off. One company's expense is another's revenue. Jensen Huang's projection of "half a trillion dollars of revenue" is directly validated by the hyperscalers' CapEx announcements. The $185 billion Google is spending is flowing primarily to GPUs and infrastructure. LONG. The panic selling in Nvidia ignores the confirmed massive inflow of capital from its largest customers. Regulatory intervention or a sudden halt in hyperscaler spending if ROI proves elusive.
Apple is up 7% this week while the rest of the "Mag 7" are down double digits. Investors are using Apple as a defensive haven because it has "no exposure" to the AI CapEx war (spending $0 relative to Google's $185B). It cannot be hurt by margin compression if it isn't playing the game. WATCH. While it is the current "winner" of price action, the speakers view this as a negative long-term signal ("washed up," "failed to participate"), implying the outperformance is purely defensive positioning rather than fundamental growth. The market realizes Apple has missed the AI innovation cycle completely, leading to long-term irrelevance.
