Coverage decision: Add to optical-supply-chain.json. Do not add to ai-power.json (weak fit — the metallurgy/equipment-tier framing in the source row doesn't apply; AEHR is test-equipment downstream of power-semi makers, not infrastructure). Do not add to memory.json (over-tagging — HBM is one of many markets AEHR serves, not its primary thesis). Retain in monster-discoveries.json (no harm, valid as a momentum holder).

Thesis (one line): Aehr is a small-mid-cap wafer-level burn-in (WLBI) equipment maker whose historical Big Customer was silicon-carbide EV-power, transitioning into CPO photonics + HBM burn-in. The stock has already done +1,591% in 1y / +917% in 3m on the promise of that transition; fundamentals show revenue collapsing from $88M (FY24-Q4, one-time SiC spike) to ~$10M/quarter today with operating losses. Coverage gap is real; the trade window is hard to time. Conviction: medium-low because valuation has already absorbed the transition narrative; the bull case requires a CPO/HBM-burn-in order materializing in primary filings, not in retrospect.

Setup

Prices as of 2026-05-18 close (full price history back to 1994; scan refresh 2026-05-23).

• Last: $95.02
• 52w high: $108.75 (company profile figure; the full price-history file shows $99.50 — see source notes)
• 52w low: $9.14
• 3M change: +917% — from ~$10 to ~$95
• 1Y change: +1,591% — from ~$5.57 to ~$95
• SMA50: $33.91 — price 180% above
• SMA200: $24.55 — price 287% above
• From 52w high: -12.6%
• Market cap: $2.99B (company profile figure, 30.7M shares diluted × $95)
• Avg volume: 2.96M; 10d avg 2.65M
• Watchlists currently: monster-discoveries only

Entry zone: No clean entry geometry as of 2026-05-18. Price is 180% above SMA50 and 287% above SMA200; any conventional zone is either far below (pullback to $50–60 = -37% to -47% from here) or a breakout above $108.75 (52w high). For a coverage-watch, this is fine — for a trade, defer until either (a) a clean pullback to $60–70 holds (SMA50 catch-up zone), or (b) a fresh CPO/HBM-burn-in order disclosed in an 8-K or earnings release re-rates the multiple structurally.

Stop: N/A for watchlist coverage. If/when entered, structural stop would be $45 (below SMA50 and the early-2026 breakout shelf). Anything closing below that means the SiC-EV-cooling-without-replacement story is winning.

Target: N/A for watchlist coverage. The thesis has no clean price target because the bull case is "AEHR's TAM transitions from SiC to CPO+HBM and revenue scales 5-10× over 3 years" — those are operating outcomes, not chart levels. A re-rate to peak fundamentals (FY24-Q4 annualized = ~$350M run-rate, never repeatable) at peer multiples would price somewhere $130–180; a re-rate to FY27 consensus revenue if CPO/HBM materializes (rough $80–120M) at 20× P/S = $1.6–2.4B mcap, which is $50–80/share, i.e., below current.

What Aehr actually does (the burn-in layer)

Per the company profile and FOX-XP/FOX-NP/FOX-CP product line:

"Aehr Test Systems provides test solutions for testing, burning-in, and stabilizing semiconductor devices in wafer level, singulated die, package part form, and installed systems."

Markets named in their own description: silicon carbide-based and other power semiconductors, 2D and 3D sensors, memory semiconductors, processors, microcontrollers, SoCs, and photonics and integrated optical devices.

The product is wafer-level burn-in (WLBI) — running un-diced wafers at elevated temperature/voltage to surface latent defects (infant mortality) before the dies are singulated and packaged. Two reasons WLBI exists vs final-test-only:

1. Cost of failure scales with downstream packaging. Once a die is bonded into an HBM stack or co-packaged with an optical engine, a latent defect scraps the entire assembly, not just one die. WLBI catches defects before that cost is incurred. The same physics that makes ATEYY's KGSD (known-good-stacked-die) test indispensable (see 2026-05-25-ateyy-deep-dive-hbm-test-equipment.md:60) make WLBI indispensable upstream of KGSD.
2. High-reliability applications need accelerated reliability screening. SiC power MOSFETs in EV inverters need ~15-year field reliability. CPO photonic ICs run at high optical-power density continuously. Both need accelerated-aging burn-in to certify reliability.

Layer placement in the test stack (synthesized from ATEYY deep-dive + AEHR product description + FORM's wafer-probe role):

AEHR is structurally Layer-2 — distinct from Layer-1 (FORM probes) and Layer-3 (ATEYY/TER final test). This is the @bullofbritain claim from 2026-05-03: "wafer-level burn-in is a missing layer in Goldman's CPO supply-chain map." The claim verifies — none of the existing optical-supply-chain.json members occupies this layer.

Why the coverage gap is real (the @bullofbritain claim)

Per reference_bullofbritain_substantive_class.md (substantive class, not skip-class), the 2026-05-03 tweet flagged 7 layers Goldman's CPO map missed. Six were already covered (SOI/IQE/AIXA/TSEM/LPK/SIVE); AEHR was the lone genuine coverage gap. Cross-corroboration from two other substantive-class sources:

• @aleabit photonics scorecard (2026-04-11): AEHR called out at 2x+ YTD as one of seven photonics names where aleabit's "going long" historically preceded big moves (see inputs/tweets/2026-04-11-aleabitoreddit-photonics-scorecard.md). AEHR was tracked but not in any thesis watchlist at that point.
• @aleabit track-record method (2026-04-25): "The moment I go long on $AEHR, it almost triples." Self-victory-lap, but the call-to-move sequencing is verifiable on the OHLC tape (max.json: $25.64 on 2026-01-02 → $95.02 on 2026-05-18 = +271%).
• @Speculator_io leaderboard (2026-04-14): AEHR ranked #6 by YTD return at +253.85%, alongside AAOI/AXTI/SNDK/LWLG/ICHR — all currently in optical-supply-chain.json or ai-infrastructure.json. AEHR was the odd one out at "monster-discoveries" only.

Three substantive-class sources independently surface AEHR as a photonics/optical-supply-chain name; none surface it as an ai-power name. That's the placement signal.

Bull case

1. Wafer-level burn-in is the structurally missing layer in our optical-supply-chain coverage. The Goldman CPO supply-chain critique by @bullofbritain held up against cross-coverage — 6/7 named tickers in coverage, AEHR the one gap. Adding AEHR closes a real layer-2 hole; not closing it leaves the test-equipment-for-CPO bucket incomplete. This is the layer-completeness argument, independent of the trade.

2. CPO + HBM both scale WLBI demand simultaneously. CPO photonic ICs are reliability-screening sensitive (high-power-density optical I/O, datacenter 5-10yr lifecycle), and HBM stacks need pre-bond burn-in to avoid stacking known-bad dies. Both end-markets are in mid-supercycle (HBM contract prices +58-63% QoQ per 2026-05-25-hbm-memory-supercycle-blind-test-verdict.md; CPO transition documented across optical-supercycle perspective). A WLBI vendor servicing both has structurally accelerating TAM — if the technical fit is real.

3. Existing customer base proves the addressable-market shape. AEHR's FY24-Q4 $88M revenue (largely SiC-EV-power) demonstrates the company can book a $300M+ run-rate quarter when an end-market lines up. The technology platform isn't new; the question is whether the next end-market (CPO + HBM-WLBI) lines up the same way. The shape is precedented.

4. Balance sheet is debt-light, dilution-disciplined. Total liabilities $18M against $157M assets and $139M equity. Shares outstanding +7.3% over 2.5 years (28.6M → 30.7M) — not the dilution-machine pattern (KOPN-class). Recent ATM-style financing raised ~$20M over the past two quarters but at $95 share price, that's <0.5% dilution per raise. The company isn't going to need to dilute existentially to survive the SiC→CPO transition.

5. Substantive-class consensus. Three independent substantive-class sources (@bullofbritain physics-first, @aleabit photonics scorecard, @Speculator_io supply-chain leaderboard) converge on AEHR as photonics-supply-chain. None of the skip-class messengers (@RealJGBanks/@MrMikeInvesting/@asklivermore-class) push this name. Source-quality signal is clean — the underlying thesis is what's being validated, not a recycler-rotation.

Bear case

1. Valuation has already absorbed the transition narrative. $2.99B mcap on $40M annualized revenue = 75x P/S with operating losses. The bull case priced in is not "AEHR survives the transition" but "AEHR re-ramps to $300M revenue at high-margin multiples within 18-24 months." Both clauses have to verify in primary filings, not in narrative. A coverage-add at this multiple is a watchlist decision, not a buy.

2. Revenue has collapsed and the recovery isn't yet in the numbers. Quarterly revenue trajectory (calendar order):

   • FY24-Q1 (Aug 2023): $20.6M, +$0.16 EPS
   • FY24-Q2 (Nov 2023): $21.4M, +$0.21 EPS
   • FY24-Q3 (Feb 2024): $7.6M, -$0.05 EPS — SiC EV cooling visible
   • FY24-Q4 (May 2024): $88.5M, +$1.36 EPS — one-time SiC EV order recognition
   • FY25-Q2 (Nov 2024): $13.5M, -$0.03 EPS
   • FY25-Q3 (Feb 2025): $18.3M, -$0.02 EPS
   • FY26-Q2 (Nov 2025): $9.9M, -$0.11 EPS
   • FY26-Q3 (Feb 2026): $10.3M, -$0.10 EPS

   The recovery the stock is pricing has not shown up in revenue yet. Quarterly losses have widened (-$0.02 → -$0.10 over the past year) on a TTM-revenue base that's fallen (~$60M → ~$40M).

3. The competitive landscape is wider than the thesis admits. ATEYY (Advantest) and TER (Teradyne) both offer burn-in capability — AEHR's edge is the wafer-level specialization, but neither incumbent is structurally locked out. If CPO/HBM-burn-in becomes a real TAM, ATEYY's customer relationships with SK hynix, Samsung, and Micron (per ATEYY deep-dive 2026-05-25-ateyy-deep-dive-hbm-test-equipment.md:62) give it the inside track. AEHR is the niche player, not the cohort leader.

4. No primary-filing confirmation of the CPO/HBM transition yet. The bull case rests on press releases, sell-side notes, and analyst tweets — none of which require GAAP disclosure. A primary-source verification path would be: (a) AEHR's next 10-Q discloses a major new customer in the photonics/HBM segment with a multi-million-dollar PO; (b) a customer-side filing (NVDA, AVGO, MRVL, MU, SK hynix) names AEHR as a qualified WLBI supplier; (c) AEHR's gross margin guide steps up as the mix shifts away from SiC. None of these is in the public record as of 2026-05-26.

5. The chart is already at the level you'd expect on success. Adding AEHR to a watchlist after +1,591% in 1y is the watchlist-add equivalent of writing the eulogy after the funeral. The coverage gap is real, but the trade gap was identified by aleabit at $30 and by the @Speculator_io leaderboard at ~$25 — at which point AEHR was already up 5x from the May 2025 $9 low. By the time we're filing the coverage decision, the trade window has been open for 6+ months and is at -12.6% from ATH.

Catalysts

• AEHR FY26-Q4 earnings (~late July 2026). First catalyst with primary-filing potential. Watch for: (a) revenue inflection or continued $10M-base; (b) backlog disclosure with CPO/HBM customer named or hinted; (c) gross margin trajectory; (d) FY27 guide. If revenue holds at $10M and the guide doesn't show step-up to $20M+ within FY27, the bull case starts to fail. If a CPO customer is named, the thesis hardens.
• Customer-side filings citing AEHR as a qualified WLBI supplier. NVDA / AVGO / MRVL CPO program filings; SK hynix / Samsung / Micron HBM-WLBI qualification disclosures. Any of these is a primary-source confirmation worth a thesis-shift entry.
• OFC 2027 / GTC 2027 product announcements. AEHR has historically used industry conferences to disclose customer engagements; aleabit's method explicitly notes "I time my longs around catalysts like $NVDA GTC or OFC" (per inputs/tweets/2026-04-25-aleabitoreddit-track-record-method.md:19).
• AEHR conference call language shift. Watch for the company itself to formally rebrand from "SiC test leader" to "CPO/HBM burn-in leader" in their own earnings transcripts. The narrative shift on the company side is the verifiable language signal; absence of that shift is itself a tell.

Coverage decision rationale

Recommendation: add to optical-supply-chain.json as wafer-level burn-in layer-2.

Three reasons:

1. Layer completeness. Goldman's CPO supply-chain map is the implicit framework optical-supply-chain.json is built around (LITE/COHR/CIEN at laser, ALAB/SMTC at silicon-photonics, FORM at probe, TSEM at foundry, etc.). The map is missing Layer-2 (WLBI). AEHR fills it. Not filling it leaves the chain incomplete; a future trace-back across "who tests CPO wafers" would still surface this gap.

2. Three substantive-class sources point here. @bullofbritain, @aleabit, @Speculator_io all framed AEHR as photonics/optical-supply-chain. None framed it as ai-power. Source-quality signal is what places it.

3. ai-power doesn't fit the metallurgy/equipment-tier framing. The source row guessed AEHR might fit the "metallurgy/equipment tier" of ai-power alongside HWM/ATI. That's the materials layer (single-crystal turbine-blade casting, specialty alloys for jet engines). AEHR is test-equipment for power-semi devices, two levels downstream of the metallurgy layer and on a different supply chain entirely (semi-equipment, not aerospace-grade casting). The placement is wrong.

Not adding to memory.json — AEHR services HBM-WLBI as one of many markets per their own product description; the deep memory-supercycle exposure is in MU/STX/WDC/DRAM/HY9H.F, not in test-equipment vendors. (For symmetry: ATEYY isn't in memory.json either despite being more memory-pure than AEHR; ATEYY sits in supply-chain-traces.)

Retain in monster-discoveries.json — still a valid +1,591% momentum-holder population member. Adding to optical-supply-chain doesn't require removal.

Sources

• info — current price ($95.02), 52w range, market cap ($2.99B), business description, sector.
• income-statements — 8 quarters of revenue/margin trajectory, EPS history, share count over time.
• balance-sheets — debt-light, equity-heavy balance ($139M equity, $18M liabilities).
• cash-flow-statements — recent $20M financing inflow, ~$5M operating burn — disciplined dilution.
• max — full price history (1,683 rows back to 1994); 52w low $5.57, 52w high $99.50; 1y return +1,591%, 3m +917%.
• inputs/tweets/2026-04-11-aleabitoreddit-photonics-scorecard.md — AEHR called out as 2x+ YTD photonics name.
• inputs/tweets/2026-04-14-speculator-io-ai-supply-chain-leaderboard.md — AEHR ranked #6 at +253.85% YTD, in a 12-name optical-supply-chain leaderboard.
• inputs/tweets/2026-04-25-aleabitoreddit-track-record-method.md — aleabit's verifiable long-AEHR-then-tripled pattern; method note for catalyst timing (NVDA GTC, OFC).
• 2026-05-25-ateyy-deep-dive-hbm-test-equipment — Layer-3 KGSD-test peer; structural test-stack analogy used to place AEHR at Layer-2.
• 2026-05-25-hbm-memory-supercycle-blind-test-verdict — HBM supercycle structural read; demand-side context for HBM-WLBI angle.
• ~/.claude/projects/-Users-janzheng-Desktop-Projects---active-ticker-tape/memory/reference_bullofbritain_substantive_class.md — substantive-class source confirmation; AEHR named as the one genuine coverage gap from the 7-ticker Goldman-CPO-map critique.
• TASKS-RESEARCH.md:137 — source row that surfaced this coverage decision.

Carrier notes

• Price/OHLC data is current as of 2026-05-18 close (full price history, refreshed 2026-05-23). The company-profile snapshot shows a slightly different 52w high ($108.75) than the full price-history data (max 1y close $99.50) — likely because the profile's 52w high uses intraday highs while the daily series records closes. Both are sourced; the $108.75 is the official 52w-high figure.
• Financials from Massive are current through FY26-Q3 (period end 2026-02-27, filed 2026-04-08). FY26-Q4 earnings (~late-July 2026) is the first forward catalyst that has not yet landed.
• No fresh fetch was run for this deep-dive — analysis uses cached data per producer-only constraint.