Everspin (MRAM) Q3 2026 Earnings Call Transcript

Everspin reported record Q3 2026 revenue of $20.6 million, up 47% year over year and at the high end of guidance; MRAM product sales rose 34% to $16.9 million, while non-GAAP EPS doubled to $0.12. The company guided Q4 revenue to $21.5 million–$22.5 million and non-GAAP EPS to $0.13–$0.18, and said price increases are being implemented to offset higher manufacturing costs. GAAP net loss was $7.2 million, including $6 million in litigation costs and a $2.7 million provision for an anticipated BIS export-control settlement. Management also highlighted a Teledyne HiRel defense partnership and CXL MRAM demonstrations, while noting that litigation costs should decline significantly in 2027.

Analysis

The key distinction is between demand momentum and earnings quality. Product growth appears broader than a single end market, but reported profitability is getting support from high-margin subcontract revenue that is recognized over time, while that stream is finite. If it fades before product margins recover, headline revenue can keep growing while gross-profit conversion disappoints. The announced price increases are the near-term test: their success depends on customers accepting higher prices, not merely on management implementing them.

The NOR shortage creates a substitution window for MRAM, but it is not yet evidence of durable share gains. If capacity returns or customers qualify alternative suppliers, distributor pull could reverse. Similarly, the CXL demo, JEDEC work, Microchip process transfer and Teledyne channel agreement provide longer-dated options, not near-term earnings proof. Standardization and qualification timelines make these poor bases for extrapolating current multiples. Teledyne’s benefit is likely customer access and portfolio breadth; no material consolidated contribution is established here. Microchip’s role is enabling manufacturing capability, with qualification still ahead.

Near term, litigation and export-control resolution keep GAAP earnings noisy; the provision is an estimate, not a settled liability. Over 1–3 months, watch product gross margin, price realization and whether guidance growth is product-led. Over 6–18 months, qualification progress and repeat orders matter more than demos or design-win counts. The constructive view is that customers value persistence, endurance and reliable supply; the contrarian risk is that a small supplier may struggle to convert interest into scaled, profitable volume.

Key Decisions for Investors

• Consider a small, staged long in MRAM after the initial post-results price discovery, sized as a high-volatility growth position rather than a core holding. Add only if the next report shows product-margin improvement alongside continued product revenue growth. Upside depends on pricing sticking and substitution converting to repeat demand; downside is meaningful if margins compress or growth relies increasingly on nonproduct revenue.
• Do not value CXL, JEDEC, UNISYST or defense opportunities as current earnings. Treat customer qualification, production orders and disclosed revenue contribution as confirmation triggers; delays or continued demo-stage commentary would weaken the longer-term thesis.
• Monitor litigation and the BIS matter separately from operating performance. Reassess if litigation costs do not decline as management expects in 2027, the final regulatory settlement materially exceeds the provision, or cash generation weakens while operating expenses remain elevated.
• Falsification checklist: product gross margin fails to improve after price increases; product revenue growth stalls as NOR availability normalizes; or guidance growth is sustained mainly by subcontract revenue. In those cases, avoid adding to MRAM and revisit the long thesis rather than assuming technology interest will translate into earnings.