Thesis overview

The TSEM thesis turns on three axes:

1. AI-photonics foundry-tier optionality — PH18 is one of two production-grade SiPh foundry nodes (the other is GlobalFoundries Fotonix). The Mar 11 2026 LWLG development agreement is the first publicly-disclosed photonics-customer name-and-date marker; the unit-economics question is whether 200mm PH18 can hold its own against 300mm Fotonix as polymer-modulator volumes ramp.
2. Specialty-analog cycle leverage — RF mobile, BCD power, image sensor, automotive analog, MEMS — Tower’s bread-and-butter end markets are mid-cycle and structurally less commoditized than leading-edge digital.
3. Standalone-public posture — the abandoned $5.4B Intel acquisition (Aug 2023, China antitrust) recast Tower as a re-rated specialty foundry rather than an Intel-internal asset. The standalone case rests on capital allocation, capex discipline, and Newport Beach Maxim 8” integration execution.

Sub-pages

• Bull case — primary drivers and confirmation paths
• Bear case — primary risks and what would invalidate the bull thesis
• Risks — formal risk register (process, customer, capex, geopolitical)
• Catalysts — dated near-term and long-term events
• Open questions — unresolved analytic items
• Cross-thesis implications — how TSEM news affects LWLG, GFS, MRVL, POET theses