AIP

Current Thesis

Arteris is the hidden royalty-pick on the custom-silicon + automotive-AI build cycle. NoC interconnect IP sits inside every chip; 4B deployed chips (milestone 2026-02-12), +50% YoY variable royalties in Q4, and FY26 guide raised to $89–93M (+26–32% vs. +22% in FY25) mean the royalty rip is re-accelerating as hyperscaler + ADAS silicon ships. Non-GAAP op-profit targeted by Q4 2026 — that's the re-rate event. Q1 print 2026-04-30 is the 10-day binary.

NOTE: Seeded theme commodity-materials is WRONG — Arteris is a fabless semiconductor IP company (Nasdaq:AIP), not a materials name. Themes corrected below.

Bull Case

• Q4 2025 revenue $20.1M, +30% YoY — ACCELERATING from the FY25 +22% pace. Reported 2026-02-12.
• Variable royalties +50% YoY to $6.6M in FY25. Royalties are the lead indicator — they spike ~18–24 months after a design-win cycle. Cycle has turned.
• FY2026 guide: $89–93M (+26–32% YoY) — company guided ABOVE the 2025 growth rate; ACV+royalties projected $100–104M.
• ACV + royalties $83.6M (+28%), RPO $116.8M (+32%) — contracted revenue backlog growing faster than revenue = visibility.
• NXP expansion 2026-02-11 (FlexNoC, Ncore, CodaCache across edge-AI SoCs, NPUs, MCUs) + Renesas R-Car Gen 5 licensing 2026-03-24 (ADAS backbone) = automotive design-start flywheel.
• 92% non-GAAP gross margin, FCF turned positive in Q4 2025 ($3M), guiding $5–9M FY26 FCF, non-GAAP op-profit targeted Q4 2026.
• Cycuity acquisition contributes ~$7M to FY26, plugs in hardware security — directly relevant to automotive functional-safety SKUs where pricing power is highest.
• Small cap ($865M @ [entry redacted] 2026-04-09) = un-owned by mainstream AI-semi funds; retail hasn't found it.

Bear Case

• ~9.6x FY26 revenue at the midpoint — not cheap for a still-GAAP-unprofitable 30% grower. Any guide shortfall re-rates hard.
• License revenue is lumpy. Q4 was royalty-heavy; if Q1 prints light on new licenses (the portion reported on Apr 30), the "accelerating" narrative gets framed as one-quarter noise.
• Royalty concentration risk — a handful of reporters (auto Tier-1s, a couple of hyperscaler customers) drive most variable royalty. Any single-customer slow-down is outsize.
• Competes with Synopsys (ARC, foundation IP ecosystem) and NVIDIA's NVLink/Grace-internal fabrics for hyperscaler custom-silicon mindshare.
• Small float → gappy tape. Works both ways but a bad print is a -15–25% candle.
• GAAP still loss-making (TTM net income -$34.7M on $70.6M rev). No pricing power premium yet.

Setup & Price Structure

• Last-quoted $19.56 (2026-04-09) — we do NOT have fresh EOD price/MAs loaded; the rule engine will overlay real levels at trade time.
• Market cap ~$865M on ~44M shares — thin float, gappy around prints.
• Earnings 2026-04-30 is in 10 calendar days / ~7 trading days. We're currently INSIDE the 3-day pre-earnings blackout window for trim, but OUTSIDE the hard "no-new-entry" window (which kicks in on Apr 27 close).
• Operationally this means: ANY entry today must exit by 2026-04-27 close. That's a 5-trading-day hold into a binary. Not a narrative-momentum ride — that's an earnings gamble, which is explicitly off-playbook.
• Preferred plan: DEFER through the print. Re-evaluate 2026-05-01 intraday: (a) if Q1 revenue > $21.5M high-end AND royalty line accelerates, entry on the breakout-retest is a HIGH-conviction picks-and-shovels setup; (b) if in-line, wait for a pullback to post-print VWAP; (c) if below low end, SKIP and re-seed.

Catalyst Calendar (next 30 days)

• 2026-04-30 — Q1 2026 earnings print. Consensus Q1 revenue ~$21.0M (guide $20.5–21.5M). Watch: variable royalty growth rate (was +50%), ACV+royalties QoQ, any hyperscaler/custom-silicon customer callouts, reaffirmation of Q4 2026 non-GAAP op-profit target.
• ~Early-May 2026 — post-print conference circuit. Arteris historically presents at sell-side tech conferences post-print; design-win disclosures often arrive here.
• 2026-05 rolling — auto-OEM SoC tape-outs. Watch for Renesas R-Car Gen 5 and NXP edge-AI silicon sampling headlines — each one is a royalty-trigger leading indicator.
• No FDA, no macro Fed date inside the window that's name-specific.

What Would Change Our Mind

Upgrade to HIGH/SUPREME on fresh entry IF (post-Apr-30 print):

• Q1 revenue ≥ $21.5M AND variable royalty line ≥ +40% YoY (confirms the royalty rip is not one-quarter noise).
• FY26 guide raised (even $1M at top end is a tell).
• Stock closes > pre-print high on day-after with volume ≥ 2× 20-day avg.

Invalidate / SKIP / exit IF:

• Q1 revenue < $20.5M (below guide low end) — thesis of acceleration is dead.
• Variable royalties growth deceleration below +25% YoY — royalty-rip narrative broken.
• Weekly close below 20-EMA on any pre-print week without an obvious macro cause.
• Management pulls or softens the "non-GAAP op-profit by Q4 2026" milestone on the call.
• AI-semi theme flips SATURATED (e.g., hyperscaler capex guidance cut on any 2026 Q1 prints from NVDA/AVGO/ARM).

Trim triggers if long:

• 3 trading days before earnings (2026-04-27 close) — cut to zero, no exceptions.
• RSI > 75 on weekly + parabolic blow-off candle into print.
• Any single-day gap +15% without a name-specific catalyst = partial trim.

Correlation Notes

• Core correlation: semi-IP peers — SNPS, CDNS, ARM. When SNPS/CDNS rip on design-automation strength, AIP tends to follow with a lag given its smaller float.
• Secondary: custom-silicon / hyperscaler AI capex proxies — AVGO (custom-silicon king), MRVL, ALAB. Arteris royalty stream is effectively a lottery ticket on every AVGO/MRVL custom-ASIC tape-out.
• Automotive-AI exposure — MBLY, NXPI, ON, NVDA Drive. Renesas deal (private name, but similar theme) is the tell. Auto-SoC design-start cycle is the slower but stickier royalty lane.
• Anti-correlation / tell-tales — ARM weakness (royalty-model canary) bleeds through to AIP sentiment even if unit economics differ. Watch ARM royalty commentary as a cross-read.
• Not correlated with: commodity materials, lithium, rare earths. (Correcting the seeded theme tag.)