AEIS

Current Thesis

AEIS is mid-re-rating from a cyclical semi/industrial power-supply vendor into a picks-and-shovels AI-data-center power play. The fundamental narrative is genuinely ACCELERATING — DC Computing revenue doubled to a record $194.2M in Q1 (reported 2026-05-04) and the 2026-06-03 launch of the ADH 800V DC-DC converter series plugs AEIS directly into the next-gen 800V HVDC rack architecture NVIDIA is pushing for Rubin/Kyber-era 1MW+ racks. But the tradable leg is MATURING/consolidating, not a fresh breakout: sell-side already published (8 PT raises 2026-05-05/07), the stock corrected ~-25% from the ~$397 early-May high to an intraday $298.85 low (2026-06-02), and now trades $312 — still below the 50-day MA (~$346) with a fresh $1.15B convert-arb short overhang. This is a name to buy on a 50-day reclaim, not to chase in the chop.

Bull Case

• Data-center segment is compounding, not a one-quarter pop: DC Computing revenue +107% in 2025, then doubled again to record $194.2M in Q1 2026 (call 2026-05-04). FY26 DC-segment growth guided to ~mid-30%.
• Fresh product proof of the 800V thesis (2026-06-03): ADH series converts 800V→50V at 98.2% peak efficiency, 8kW peak / 6kW full load in a half-brick, >2,700 W/in³ peak density, liquid-cooled, parallelable. Paired with the NDQ 50V→12V NIBC + new Hot Swap Control module = a complete 800V DC solution spec'd for NVIDIA Rubin/Kyber-class racks. This is design-in territory, not vaporware.
• Guide was raised, not just met (2026-05-04): Q2 rev $520–560M (mid $540M vs $526.9M est), Adj EPS $1.93–$2.43 (mid $2.18 vs $2.04 est); FY26 total rev growth lifted to low-to-mid-20%.
• Structural uptrend intact: $312 (2026-06-02) sits above the 200-day MA (~$286); 52-wk range $112.72–$397.44 — a re-rating year, not a roundtrip.
• Smart-money accumulating the dip: Fisher Asset Mgmt added 104,661 shares (filed 2026-05-26), Wolverine opened a new stake (2026-05-06) — large allocators buying into the correction.
• Supply, not demand, is the constraint: >$2.5B revenue-generating output coming online (Malaysia/Philippines/Mexico) + $1B Thailand.

Bear Case

• The catch-up leg is largely spent: 8 PT raises in 3 days (BofA $430, Susquehanna $430, KeyBanc $415, Citi $410, Needham $400, Baird $380, even bears TD Cowen $350 / Wells $345) all dated 2026-05-05/07. Per playbook, clustered sell-side = narrative already mainstream.
• Price structure is broken [trade redacted] off the ~$397 high, undercut the May low with a fresh $298.85 print (2026-06-02), and sits below the 50-day MA (~$346)**. A momentum book wants price above the 50-day, not under it.
• Convert-arb overhang is mechanical and fresh: $1.15B 0% notes due 2031 priced 2026-05-13/14 at 50% premium (~$508.79 conversion). Arb desks short the equity to delta-hedge — structural sell pressure into any bounce.
• Second leg isn't firing: management flagged on the 2026-05-04 call that DC strength is masking soft semiconductor demand. If DC growth decelerates, there's no semi recovery to backstop the multiple.
• No catalyst in the tank: Q1 already printed; next report ~early August. Nothing dated in the next 30 days to re-ignite the tape.

Setup & Price Structure

• Last $312.28 (2026-06-02, +6.0% on the day); intraday range $298.85–$313.27. The +6% pop is a bounce off the fresh low, not yet a confirmed higher-low base.
• 50-day MA ~$346 (price ~10% below); 200-day MA ~$286 (price ~9% above). Mid-channel: corrected hard but the long-term uptrend is intact.
• The corrective sequence undercut the prior ~$309–$313 May lows with the $298.85 print — so this is a lower low + bounce, not a clean higher-low re-entry yet.
• Read: needs to either (a) reclaim the 50-day (~$346) on volume to confirm the pullback is over → momentum re-entry, or (b) build a higher low above [entry redacted] and base. The ~$397 early-May high is the swing-high to reclaim. Below $286 (200-day) = thesis-breaking.

Catalyst Calendar (next 30 days)

• ~2026-06-04 → 2026-07-04: no hard-dated binary catalyst. Q2 earnings not yet scheduled, expected ~early August (well outside 30d).
• Ongoing (post 2026-06-03 ADH launch): watch for 800V design-win / customer announcements or hyperscaler/NVIDIA reference mentions — these would be the next narrative accelerants and are unscheduled.
• Convert technicals: arb-hedging flow around the $1.15B 0% 2031 notes (priced 2026-05-13/14) continues to weigh; no fixed date.
• Industry reads: any NVIDIA / hyperscaler 800V HVDC roadmap commentary at June conferences is a sentiment driver for the whole industrial-power-AI cohort.

What Would Change Our Mind

• Upside (re-arm to APPROVE): weekly close back above the 50-day MA (~$346) on expanding volume, ideally after carving a higher low above [entry redacted] — that's the clean momentum re-entry on a MATURING name.
• Downside (kill thesis): weekly close below the 200-day MA (~$286), or loss of the $298 (2026-06-02) low with no reclaim — structural uptrend broken, becomes a value-trap; cut, don't average.
• Fundamental break: Q2 (~early Aug) DC Computing revenue printing a sequential decline vs $194.2M, or management cutting the FY26 ~mid-30% DC-growth guide — that ends the acceleration story regardless of price.
• Saturation flip: a fresh wave of mainstream/CNBC "AI power" cover-story coverage with the stock back near $397 and no new design wins = peak-narrative; fade strength, don't chase.

Correlation Notes

• Theme cohort (industrial-power-ai / AI-datacenter-power): trades with VRT (Vertiv), POWL, GEV, ETN, NVT, and to a degree the broader data-center power-and-cooling complex. Cluster confirmation = peers also reclaiming MAs; cluster failure = whole theme rolling over (de-risk together).
• Upstream driver = NVIDIA / hyperscaler capex: AEIS is a 2nd-derivative on the GB300→Rubin rack-power buildout. NVDA roadmap guidance and hyperscaler capex revisions move AEIS more than AEIS-specific news between prints.
• Semi-cap residual beta: legacy exposure to AMAT/LRCX/semi-equipment cycle still matters — a WFE downturn drags the semi segment that's already undergrowing.
• Rate sensitivity via the convert: as a high-multiple re-rating name with a fresh 0% convert, AEIS is sensitive to long-end rates and risk-appetite regime shifts; macro tightening hits the multiple and amplifies arb-short pressure.