BN · rank #3 · 2026-10-22
TSEM
NASDAQ · $21.52B (USD)
The exact numbers the algorithm saw.
| Composite score Z-score blend of the factors below; drives the rank. | 0.91 |
|---|---|
| Forward revenue growth Consensus forward revenue growth. | +37.9% |
| Forward net margin Consensus forward net margin. | +21.2% |
| Net margin TTM Trailing twelve month net margin. | +16.9% |
| Margin expansion Forward minus trailing net margin (percentage points). | +4.2% |
| Forward PEG Forward P/E to growth. Below 1 is cheap for the growth. | 3.43 |
| Debt / FCF Net debt relative to free cash flow. Lower is safer. | 0.47× |
| Analyst upside Spread between the consensus 12m target and the current price. | +65.1% |
| Last EPS surprise Most recent reported EPS versus consensus. | +14% |
| Market cap (USD) | $21.52B |
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
Tower Semiconductor is a specialty analog/mixed-signal foundry whose silicon photonics and SiGe processes are directly enabling the 800G/1.6T optical-interconnect buildout powering AI data centers.
Rationale
The bottleneck-solver thesis is concretely reinforced: SiPho revenue surged from a $180M to $680M annualized run rate in four quarters, Semtech locked in dedicated multi-fab capacity, and NewPhotonics is already shipping high-volume PICs — this is a capacity-constrained ramp, not sentiment rotation.
Material risks
- 1Tower is not a confirmed sole-source supplier — Semtech's dedicated-capacity deal provides demand visibility but leaves open dual-sourcing or customer insourcing risk if hyperscalers vertically integrate silicon photonics (Intel, TSMC, and GlobalFoundries all have competing SiPho platforms).
- 2The fwd PEG of 3.4 prices in near-flawless execution on the >$1B SiPho annualized revenue target by Q4 2026; any miss on that ramp — from Japan fab delays, METI grant complications, or 3.2T standard timing slippage — compresses the multiple sharply given 65% analyst upside already embedded in consensus.
AI verdict council
Each pick is reviewed independently by 3 models before any order. 2 of 3 voted to proceed.