Aeluma, Inc. (ALMU) is a $450M compound semiconductor company in Goleta, CA that integrates III-V materials on silicon for photonics applications. The stock has rallied +87% since March, pulled along by NVIDIA's $2B+ investments in named photonics suppliers (Lumentum, Coherent, Marvell, Corning) and the InP shortage out of OFC 2026.

What the filing says

CEO Jonathan Klamkin, asked directly about customer qualification status: "We have not been qualified by a specific customer at this time." After five years of operation, all ≈30 commercial engagements remain in evaluation/feasibility. There is no named customer, no LOI, no purchase order.

The mobile SWIR engagement — ALMU's largest commercial TAM and the most-discussed opportunity on prior calls — regressed in disclosure specificity. Q2 FY2026 management mentioned conversations "shifting to pricing and delivery." When an analyst on the Q3 call asked directly for the pricing/delivery update, the CEO reverted to generic "evaluation and feasibility" language and declined to commit to a timeline.

Full-year revenue guidance was cut 23% on the top end: $4-6M → $4.2-4.6M, attributed to government contracting delays. Q3 revenue was $1.2M, all government. The government pipeline did recover (6 new contracts >$5M signed in H2 meeting the company's 3-7 goal), but recognition is deferred to FY27.

The $50M ATM facility filed March 20 is untouched. The CFO guided that spending will "increase." Cash is $37.8M against a quarterly burn of $792K that management says will accelerate.

What the market thinks

At $24.83, ALMU trades at ≈100x FY2026 guided revenue. The two analysts (Craig-Hallum $26, Benchmark $25) imply commercialization is imminent and material — and Craig-Hallum is the ATM agent. Short interest is 27.5% with 1.8 days to cover. Call IV exceeds put IV by 7.8% — unusual skew, market positioned for upside spike. Beta is -0.09; this is a pure-idio name moving on its own narrative.

Why the gap exists

This is sector-narrative pull-through colliding with company-specific reality. NVIDIA's $2B+ photonics investments and the InP shortage are public and real, but ALMU is not Lumentum or Coherent — it is a pre-revenue device supplier that has not yet been qualified by any of them. The Q-over-Q mobile SWIR disclosure regression is a careful filing-read that retail does not parse. The May 1 CEO share gift (1,570,000 shares, ≈9% of float, ≈$38.9M) plus same-day $499K cash sale is the largest insider distribution event to date, and the Form 4 has not yet propagated through coverage. The material weakness in disclosure controls (Dec 31, 2025 10-Q) went unmentioned on the call.

Risks (ranked)

1. Squeeze leg 2. 27.5% SI + 1.8 days cover + call-heavy positioning = asymmetric upside tail. Break of $25.88 52w high triggers gamma cover.
2. Commercial breakthrough. Named Tier 1 customer in 8-K Item 1.01 would invalidate.
3. Sector narrative duration. If photonics rally extends, ALMU rides pull-through indefinitely.

Catalysts

• May 14-31, 2026: Insider window reopens post-earnings; additional distribution probable.
• ~Sept 2026: Q4 FY2026 earnings + 10-K — revised guidance test, material weakness status.
• Any time before Q4 10-K: ATM first tap disclosure.
• Dec 31, 2026: Deadline for named commercial customer qualification.

What would change our mind

• Named Tier 1 customer in 8-K Item 1.01 (LOI or PO, not press release)
• First commercial purchase order >$500K
• ATM deployment <$2M cumulative through Sept 2026 (suggests internal capital sufficiency)
• Material weakness remediated in next 10-K
• CEO Form 4 open-market purchase (Code P) >$200K — would invert the distribution pattern

Evidence

Source: Q3 FY2026 earnings transcript (2026-05-13), supporting 10-Q (FY2026 Q2), Form 4 filings (May 1, 2026), market data (May 14, 2026).