BN · rank #17 · 2026-07-21
TSM
NYSE · $2.09T (USD)
The exact numbers the algorithm saw.
| Composite score Z-score blend of the factors below; drives the rank. | 0.23 |
|---|---|
| Forward revenue growth Consensus forward revenue growth. | +31.2% |
| Forward net margin Consensus forward net margin. | +47.9% |
| Net margin TTM Trailing twelve month net margin. | +49.9% |
| Margin expansion Forward minus trailing net margin (percentage points). | -2.0% |
| Forward PEG Forward P/E to growth. Below 1 is cheap for the growth. | 1.01 |
| Debt / FCF Net debt relative to free cash flow. Lower is safer. | 0.82× |
| Analyst upside Spread between the consensus 12m target and the current price. | +30.0% |
| Last EPS surprise Most recent reported EPS versus consensus. | +11% |
| Market cap (USD) | $2.09T |
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
TSMC is the world's dominant advanced-node contract foundry, manufacturing the majority of leading-edge AI, HPC, and smartphone chips for global fabless customers at 7nm and below.
Rationale
The bottleneck-solver thesis is directly reinforced by TSMC's sole-source status at leading-edge nodes, its 10x CoWoS capacity expansion to serve AI packaging demand, and Q2 2026 results showing 33%+ revenue growth, 67.7% gross margins, and a 77% beat on EPS — all confirming structural pricing power and order visibility rather than sentiment rotation.
Material risks
- 1Geopolitical escalation or expanded U.S./China export controls targeting Taiwan or advanced AI chip shipments could disrupt fab operations, freeze capital flows, and force customer re-sourcing — the single scenario that breaks sole-source status involuntarily.
- 2Hyperscalers or AI chip designers (e.g., Intel Foundry, Samsung) winning meaningful advanced-node or CoWoS share would erode TSMC's bottleneck pricing power, though no concrete displacement is evident in current data.
AI verdict council
Each pick is reviewed independently by 3 models before any order. 3 of 3 voted to proceed.