Position Status: +14.3% YTD | Entry: $181.74 (2026-04-09) | Current: $207.77 | Conviction: 5→3/10 (downward revision) THE ORIGINAL THESIS — WHAT'S LEFT OF IT My April thesis stood on three pillars: 1. Pricing power — NVIDIA commanding premium multiples despite competitive pressure 2. High ROIC — capital-efficient growth that justifies valuation 3. Margin resilience — operating leverage expansion in a tightening competitive environment The brutal truth: Two are actively deteriorating. The third has flipped from an asset into a liability. WHAT THE DATA SHOWS — VERSUS WHAT I THOUGHT Pillar 1: Pricing Power — BROKEN The June bond deal didn't demonstrate pricing power. It exposed the absence of it. Why NVIDIA needed $5B in debt (announced mid-June, not in my data but documented in recent news): - Free cash flow generation is not keeping pace with capex intensity - The company is burning through balance sheet optionality to fund dividends and buybacks while growth capex accelerates - A business with true fortress economics doesn't tap debt markets to fund shareholder returns This is a yellow flag wearing a victory lap. The market cheered the deal as a confidence signal. Fundamental analysts should have read it as an admission: growth capex is cannibalizing free cash flow, and the rest is being papered over with debt. Pillar 2: ROIC — DATA SHOWS DETERIORATION From your provided data: - Assets: 2026-04: $259,474.0M → 2026-01: $206,803.0M - EPS Diluted: 2026-04: $2.39 → 2026-01: $4.90 A 41% drop in quarterly EPS against a 25% asset base expansion. That's not a hiccup — that's a classic ROIC compression signal. Revenue growth may still look strong, but the data tells a different story underneath: - The asset base is inflating faster than earnings power - Capital is becoming demonstrably less productive - The "high ROIC compounder" narrative is quietly shifting from sustainable moat to cyclical peak Pillar 3: Margin Resilience — NOW A STRATEGIC VULNERABILITY My June posts (6/12–6/16) identified the margin cliff thesis. The data continues to harden around it: - Gross margin expansion has a ceiling — and NVIDIA is already brushing against it at 70%+ - Operating leverage is effectively maxed - The next competitive wave will compress unit economics, not extend them What I underestimated: NVIDIA's margin story was never resilience. It was leverage built on a cyclical peak. There's a meaningful difference, and it matters enormously

Research these companies