Earnings call transcript: Tower Semiconductor’s Q1 2026 earnings beat expectations

Tower Semiconductor beat Q1 2026 expectations with EPS of $0.65 versus $0.55 consensus and revenue of $414 million versus $408.38 million expected, while guiding Q2 revenue to a record $455 million ±5%. Revenue rose 15% year over year, gross profit jumped 52%, operating profit 96%, and the stock surged 15.36% pre-market to $254.76. Management highlighted strong momentum in silicon photonics and silicon germanium, but noted ongoing supply-chain and currency risks.

Analysis

TSEM is not just printing a beat; it is converting customer pull into balance-sheet-backed demand visibility, which changes the probability distribution for the next 12-18 months. The key second-order effect is that prepayments and multi-year capacity commitments de-risk the capex cycle and let management front-run expansion without leaning on external financing, which should compress perceived execution risk even if headline multiples stay optically expensive.

The market is likely underestimating the strategic value of the silicon photonics ramp relative to the current stock move. If the company can convert design wins into sustained yield and throughput at the new 300mm footprint, it creates a flywheel: more capacity enables more qualified customers, which improves pricing power on next-generation nodes and raises switching costs for competitors. That is structurally more important than the one-quarter EPS beat, because it positions TSEM as an infrastructure layer for the AI optical stack rather than a cyclical specialty foundry.

The real competitive tell is not TSEM vs. small photonics peers, but against larger integrated foundries that can bundle logic and packaging. TSEM’s advantage is that it appears to own the niche where optical innovation, not full-stack integration, matters most; that favors COHR as a design partner and puts pressure on alternative SiPho suppliers that lack comparable customer qualification breadth. The flip side is that if supply-chain bottlenecks in indium phosphide or Japanese expansion approvals slip by even 2-3 quarters, the market will have to re-rate the 2027/2028 visibility premium quickly.

Consensus may be too focused on the valuation headline and not focused enough on the fact that the company is buying time with capital discipline and customer advances. The move is probably directionally right but tactically extended after a 15% gap-up; the cleaner setup is whether management follows with raised long-term targets in the next 1-2 quarters. If that happens, the current premium can still expand; if not, the stock can retrace hard because the multiple already prices in uninterrupted execution.

Key Decisions for Investors

• Buy TSEM on pullbacks over the next 1-3 sessions rather than chasing the gap-up; use a 5-8% retracement as entry, with a 10-15% upside target into the next model update and a 6-7% stop if photonics volumes fail to confirm.
• Pair trade: long TSEM / short a basket of slower-moving specialty foundries for the next 1-2 quarters; the thesis is that customer prepayments and 300mm photonics capacity create a faster earnings revision cycle for TSEM than peers.
• Long COHR vs. short LWLG on a 3-6 month horizon; COHR benefits from commercialization and customer pull-through on optical components, while LWLG remains more of a science-risk story with lower near-term monetization visibility.
• Sell out-of-the-money TSEM calls against existing longs into implied-volatility strength; the stock has rerated faster than fundamental capacity can be proven, so covered-call income captures the near-term overextension risk.
• Add NVDA only as a second-order beneficiary on weakness, not strength; if the AI optical buildout is real, the supplier ecosystem should broaden, but the cleaner trade remains the optical component/foundry layer where operating leverage is more direct.