BN · rank #2 · 2026-07-23
NVDA
NASDAQ · $5.14T (USD)
The exact numbers the algorithm saw.
| Composite score Z-score blend of the factors below; drives the rank. | 1.07 |
|---|---|
| Forward revenue growth Consensus forward revenue growth. | +42.5% |
| Forward net margin Consensus forward net margin. | +63.3% |
| Net margin TTM Trailing twelve month net margin. | +63.0% |
| Margin expansion Forward minus trailing net margin (percentage points). | +0.3% |
| Forward PEG Forward P/E to growth. Below 1 is cheap for the growth. | 0.57 |
| Debt / FCF Net debt relative to free cash flow. Lower is safer. | 0.11× |
| Analyst upside Spread between the consensus 12m target and the current price. | +42.8% |
| Last EPS surprise Most recent reported EPS versus consensus. | +6% |
| Market cap (USD) | $5.14T |
The AI research card
Independent qualitative review of each pick before the order is placed. On the rare day the research service is unavailable, the paper book trades on the quant ranking alone and no card appears here.
Summary
NVIDIA is the dominant supplier of AI datacenter GPUs and networking infrastructure, generating $81.6B in quarterly revenue with ~90% YoY datacenter growth driven by hyperscaler and enterprise AI factory buildouts.
Rationale
The bottleneck_solver thesis is directly reinforced — NVIDIA is the literal rate-limiting input to large-scale AI training and inference capacity, with a 0.57 forward PEG, 42% analyst upside, and 63% net margins confirming both structural scarcity pricing and durable profitability rather than sentiment-driven multiple expansion.
Material risks
- 1Hyperscalers (Google TPUs, Amazon Trainium, Microsoft Maia) are actively scaling in-house AI silicon, which is the canonical disqualifier for a bottleneck-solver thesis — if large cloud providers reduce GPU dependency, NVIDIA loses its sole-supplier pricing leverage with customers representing ~50% of datacenter revenue.
- 2U.S. export controls have already structurally excised China from NVIDIA's forward revenue model, capping the addressable market and creating binary regulatory risk if controls expand to additional geographies or chip classes.
AI verdict council
Each pick is reviewed independently by 3 models before any order. 2 of 3 voted to proceed.