# 2455.TW - VPEC (Visual Photonics Epitaxy)
## Date: 2026-04-12
## Score: 69/100

### Thesis (2-3 sentences)
VPEC is the closest thing in Taiwan's III-V epitaxy layer to a current production choke point for AI optical demand. Public materials and DIGITIMES coverage tie the company directly to data-center optical demand, silicon-photonics-related optical growth, and better order visibility than a typical upstream epi house. The limitation is that VPEC is still a mixed III-V platform spanning wireless and photonics, so the stock is not a pure-play optical bottleneck.

### Source Trail
| Topic | Source | Date / Period |
|-------|--------|---------------|
| Stock code / listing | VPEC investor relations stock information | Accessed 2026-04-12 |
| Business / product stack | VPEC annual report | 2024 |
| Data-center order visibility | DIGITIMES Asia | 2024-03-25 |
| SiPh / optical growth angle | DIGITIMES Asia | 2025-03-17 |

### Key Metrics
| Metric | Value | Notes |
|--------|-------|-------|
| Exchange | 2455.TW | Reported by company |
| Core process | MOCVD III-V epitaxy | Reported by company |
| Product scope | HBT, EEL / VCSEL epi, FP / DFB LD epi | Reported by annual report |
| End markets | Wireless, optical communications, sensing | Reported by annual report |
| Public demand signal | Data-center orders with visibility into 2025 | Reported by DIGITIMES |

### Value Chain Position
VPEC sits in the upstream epi-wafer layer, above device and module vendors but directly exposed to optical component demand. Relative to other III-V suppliers in the repo, VPEC appears closer to current production demand rather than only platform optionality: the public read-through is that AI data-center optical demand is already flowing into its order book.

### Moat Analysis
**Rating: Strong (4/5)**

- **Current-production positioning**: among the three-company comparison you provided, VPEC reads most like the supplier already on the active volume ramp.
- **Relevant optical mix**: the company sells the right laser and optical epi products for datacom / optical networks, not just generic compound-semiconductor capacity.
- **Still mixed, not pure**: VPEC also serves microelectronics / wireless markets, so it is not as clean a photonics expression as a dedicated datacom-only supplier.
- **Not a monopoly**: the choke point is best understood as supply responsiveness and current qualification position rather than sole-source control.

### Catalysts
1. **Brisk data-center order visibility**: DIGITIMES explicitly tied VPEC's order visibility to the data-center sector.
2. **SiPh-linked optical upside**: DIGITIMES later linked VPEC's growth plans to silicon-photonics advancements and optical-component demand.
3. **Optical mix improvement**: if data-center optics becomes a larger part of the mix, valuation should look more like an AI optical supplier than a traditional III-V epi house.

### Risks
1. **Mixed end-market exposure**: wireless / RF softness can still dilute the optical thesis.
2. **Competitive layer**: this is a strong position inside III-V epi, not an uncontested bottleneck.
3. **Valuation framing risk**: if the market keeps treating VPEC as a cyclical compound-semi name instead of an AI optical beneficiary, rerating can stall.

### CHIPS Act Update (2026-04-12)
Status: `NONE`.

I did not identify a direct US CHIPS award that changes the investment framing here. The thesis is commercial AI optical demand and production positioning, not subsidy support.

### Verdict: HOLD
VPEC looks like the best current-production expression of the three-way comparison with IQE and IntelliEPI. It deserves inclusion because it appears closer to the live AI optical volume ramp than IQE, and more operationally proven than IntelliEPI. I am keeping it at `HOLD` rather than `BUY` because public valuation work is still thin and the business mix is broader than a pure optical epi asset.

**Final Score: 69/100**

### Update 2026-08-17 — 經濟日報 InP hike

Bloomberg 17 Aug: NT$417, mcap NT$77.1bn (~$2.4bn), TTM P/E 101x, P/S 20.4x, GM 36.1%, OM 19.8%, TTM sales NT$3.38bn / NI NT$548m. 1D +9.9%, 1M +47.9%, YTD +175%. 14 analysts, rec 4.79, BEST TP 430 (stock is at the target).

FY27E BEST: sales NT$4.37bn, EPS 5.37 (4wk −0.09). FY28E sales NT$6.15bn, EPS 9.13. Implied FY27E P/E ~78x.

Jul-26 monthly NT$422m (+29.5% MoM / +45% YoY, fifth consecutive monthly rise) independently confirms the article’s “Q3 逐月走升.” The 10% Q4 epi hike remains a 經濟日報 channel check, not a company 公告.

Verdict unchanged: HOLD. Current-production epi is real; the stock is no longer undiscovered versus consensus TP.