Everspin (MRAM) Q3 2024 Earnings Transcript

Everspin reported Q3 revenue of $12.1 million, in line with guidance, and GAAP EPS of $0.10, well ahead of the guided loss, helped by $4 million of other income from a DoD award. Cash rose to $39.6 million and the company remains debt free, while new revenue opportunities emerged from Frontgrade, IBM FCM4, and Lucid Motors. Management guided Q4 revenue to $12 million-$13 million and EPS of breakeven to $0.05, but flagged ongoing weakness in Japan and Germany and only modest near-term product growth.

Analysis

The key incremental signal is not the quarter’s optics but the shape of the revenue bridge: this is becoming a “lumpy base + embedded option value” story, where defense/industrial qualification cycles and design wins now matter more than near-term product run-rate. The market should treat the DoD-related award and Frontgrade as evidence that Everspin’s technology is being used as a qualification platform for higher-value, longer-cycle programs; if those milestones convert, the operating leverage can surprise because the company is already carrying a relatively fixed cost base.

The real second-order winner is IBM’s FlashCore ecosystem: Everspin’s 1Gb supply continuity reduces qualification risk for a storage SKU that likely values supply reliability over cost, which should make IBM less sensitive to marginal memory pricing and more sensitive to execution consistency. In automotive, the Lucid win matters less for near-term revenue than for reference design credibility; if the Gravity program scales, it becomes a validation point for other OEMs and Tier-1s that want a proven, automotive-grade nonvolatile memory vendor, but the adoption curve is still gated by vehicle demand and platform volume.

The main risk is that management’s optimism about 2025 ramps is too dependent on long qualification windows while the current business still looks exposed to regional macro softness and fab fixed-cost absorption. Japan/Germany weakness can compress product utilization before the new wins mature, limiting margin expansion even if topline stabilizes. Contrarian take: the market may be underestimating the durability of the defense/space opportunity and overreacting to the softness in legacy product demand; if the non-core awards keep converting, Everspin’s earnings quality improves materially even without a step-change in product revenue.

For the next 1-2 quarters, the trade is less about owning beta to memory and more about owning the asymmetric conversion of backlog into recognized revenue. If management can show sequential improvement in gross margin despite flat products, the stock can rerate on proof of operating leverage; if not, this remains a range-trade with event-driven upside but limited fundamental momentum.

Key Decisions for Investors

• Long MRAM into Q4/Q1 print as a 3-6 month event-driven trade; target a rerating if management demonstrates sequential gross-margin stabilization and additional milestone recognition from defense programs. Risk: if product revenue stays flat and margin remains sub-50%, the stock likely stays range-bound.
• Pair trade: long MRAM / short a higher-beta memory hardware name with weaker balance-sheet flexibility over the next 1-2 quarters. MRAM’s cash position and defense-linked optionality give it better downside protection than a pure cyclical memory proxy.
• Buy MRAM Jan-2026 upside calls or call spreads if implied volatility remains modest; thesis is that 2025 qualification conversions create convex upside, while cash-burn risk is limited by the debt-free balance sheet. Use spreads to cap premium if revenue timing slips.
• Avoid chasing IBM on this headline alone; if anything, the incremental benefit is execution de-risking rather than material earnings delta. Use any weakness in IBM only if you want exposure to storage resilience, not as a direct MRAM lever.
• If MRAM breaks below the post-earnings support on no follow-through in defense milestones, reduce exposure quickly—this is a catalyst stock, and the downside case is a prolonged qualification trough rather than a balance-sheet event.