NVDA Nvidia
A real AI franchise priced for near-perfection; margins fading toward 60% leave little margin of safety.
The story
Nvidia is the dominant supplier of AI accelerated computing. Its moat is the CUDA software ecosystem plus full-stack systems (GPUs, NVLink, networking), which has produced 76% operating margins on roughly $300B of TTM revenue. It is a young-growth franchise at a cyclical peak in hyperscaler capex. Competition from custom ASICs (TPU, Trainium, MTIA) and AMD, together with customer concentration, should erode pricing over time but not end the franchise.
Growth starts at 35% because the data-center buildout is still accelerating, then decays as revenue reaches a scale where AI capex can only track the broader economy. Margins fade from 76% to 60% as custom silicon and hyperscaler bargaining power compress pricing, though the software and systems moat keeps them far above semiconductor norms. A 12-year horizon reflects a durable moat with a long reinvestment runway. Sales-to-capital stays near the observed 1.8.
Value drivers
| Revenue growth (Y1) | 35.0% |
| Terminal growth | 4.0% |
| Forecast horizon | 12y |
| Target operating margin | 60.0% |
| Years to target margin | 8 |
| Sales-to-capital | 1.80 |
| Beta | 1.30 |
| Failure probability | 2.0% |
| Cost of capital (WACC) | 11.1% |
| Terminal WACC | 9.8% |
Valuation bridge
| PV of explicit FCFF | 3.57T |
| PV of terminal value | 4.77T |
| Equity value | 8.18T |
| ÷ shares → per share | $338.57 |
Projected free cash flow to the firm
| Yr | Revenue | Growth | Margin | NOPAT | Reinvest | FCFF | PV |
|---|---|---|---|---|---|---|---|
| 1 | 409.01B | 35.0% | 74.0% | 256.86B | 58.91B | 197.95B | 178.14B |
| 2 | 540.63B | 32.2% | 72.0% | 330.36B | 73.13B | 257.23B | 208.33B |
| 3 | 699.38B | 29.4% | 70.0% | 415.50B | 88.19B | 327.31B | 238.56B |
| 4 | 885.04B | 26.5% | 68.0% | 510.79B | 103.14B | 407.65B | 267.38B |
| 5 | 1.10T | 23.7% | 66.0% | 613.41B | 116.66B | 496.75B | 293.22B |
| 6 | 1.32T | 20.9% | 64.0% | 719.22B | 127.20B | 592.02B | 314.49B |
| 7 | 1.56T | 18.1% | 62.0% | 822.81B | 133.07B | 689.75B | 329.74B |
| 8 | 1.80T | 15.3% | 60.0% | 917.91B | 132.66B | 785.25B | 337.83B |
| 9 | 2.03T | 12.5% | 60.0% | 1.03T | 124.71B | 907.53B | 351.37B |
| 10 | 2.22T | 9.6% | 60.0% | 1.13T | 108.50B | 1.02T | 356.52B |
| 11 | 2.37T | 6.8% | 60.0% | 1.21T | 84.17B | 1.12T | 352.67B |
| 12 | 2.47T | 4.0% | 60.0% | 1.26T | 52.75B | 1.20T | 339.89B |
Key risks
- AI capex digestion or a cyclical downturn as hyperscalers' ROI on AI spend disappoints
- Custom ASICs and AMD eroding share and gross margins
- Export controls and geopolitical exposure (China, Taiwan/TSMC supply concentration)
Catalysts
- Rubin platform ramp and sustained sovereign and enterprise AI demand
- Growth in software, networking and inference revenue that diversifies the business beyond training GPUs