AIP · Arteris, Inc. · Stock research

Current Thesis

Arteris licenses the network-on-chip (NoC) interconnect IP embedded in SoCs and collects a variable royalty each time a licensed chip ships. Royalties lag design wins by roughly 18–24 months, so the 2023–24 custom-silicon and ADAS win wave is only now converting to revenue. That inflection is real and confirmed: the Q1 2026 print (2026-05-12) showed revenue $22.9M (+39% YoY, ~8% above the ~$21M consensus), variable royalties +67% YoY, ACV+royalties $92.8M (+39%), RPO $118.3M (+33%), and a raised FY26 guide of $91–95M with positive free cash flow.

The complication is entirely in the tape. Price rode that inflection vertically clearing the $38.99 prior ATH on a +11.9% session (2026-06-12), extending to a $44.94 intraday high (6/15), and topping at a $48.59 closing ATH on 2026-06-30. Then it reversed hard: a -10.7% session to $39.11, a -16.4% intraday flush to $36.63, and a slide to ~$34.46 by 2026-07-08 roughly -29% in a week. The move sliced back through the ~$39 breakout shelf, so the breakout the prior structure was built on has failed. A late-week bounce to ~$36.61 (2026-07-10) reads as short-covering and dip-testing rather than a repaired base. The royalty narrative is fundamentally accelerating; the stock's own momentum leg is broken. Fresh entries into a post-blowoff reversal 3+ weeks ahead of the next print carry no edge. The correct stance is to stand aside until the name builds a higher low and reclaims structure. That rates a low-conviction, watch-for-re-base setup, not a buy here.

Bullish and bearish views on Arteris, Inc.

The model's bull view on Arteris, Inc. (AIP), in brief: Q1 2026 (2026-05-12): revenue $22.9M, +39% YoY, ~8% above the ~$21M consensus, EPS -$0.03 vs -$0.09 expected a clean beat and an acceleration from the FY25 +22% pace. The bear view: The breakout failed. After a $48.59 closing ATH (2026-06-30), price collapsed ~29% in a week and closed back below the ~$39 shelf it had broken out from. A failed breakout that loses its base is a broken momentum structure, not a pullback to support. The top was a single-week… Both cases follow in full.

Bull Case

• Q1 2026 (2026-05-12): revenue $22.9M, +39% YoY, ~8% above the ~$21M consensus, EPS -$0.03 vs -$0.09 expected a clean beat and an acceleration from the FY25 +22% pace.
• Variable royalties +67% YoY in Q1 the lead KPI accelerated hard, the strongest evidence that prior design wins are converting to shipped chip volume, and the metric that carries the whole thesis.
• ACV + royalties $92.8M (+39%) and RPO $118.3M (+33%) contracted backlog is growing in line with revenue, giving the acceleration forward visibility rather than a one-quarter spike.
• FY26 guide raised to $91–95M with positive free cash flow; Q2 guide $23–24M a beat-and-raise quarter, guiding to continued sequential growth.
• Fresh automotive wins keep landing: SiEngine Technology licensed FlexNoC IP for a next-gen automotive SoC (intelligent cockpit, ADAS, AI cockpit-drive fusion) in late June 2026, stacking on the June Li Auto partnership (L9 Livis SUV) and the Q1-call wins (a global hyperscaler, a top HBM supplier, a MIPS physical-AI collaboration, NoC IP on NASA Artemis).
• Analyst cluster re-rated up in Q2: TD Cowen $40 (from $22), Rosenblatt $38 Buy (from $20), Northland $38, Jefferies $35 (from $16); consensus is Buy with a $37.75 average target the sell-side narrative went mainstream, and price at ~$36.61 has now fallen back to the middle of that band.

Bear Case

• The breakout failed. After a $48.59 closing ATH (2026-06-30), price collapsed ~29% in a week and closed back below the ~$39 shelf it had broken out from. A failed breakout that loses its base is a broken momentum structure, not a pullback to support.
• The top was a single-week blowoff. A +11.9% spike to fresh highs (6/12) followed two weeks later by -10.7% and -16.4% sessions is parabolic exhaustion; chasing the former high after the reversal is the anchor-to-old-price trap.
• Insider selling into the move: CEO K. Charles Janac filed a Form 4 (early July 2026) for 70,000 shares sold (~$2.44M), retaining >9.18M shares trimming into strength, not accumulating the dip.
• Valuation is still rich for a name mid-reversal: ~17x forward revenue (~$1.6B cap on ~44M shares against the $91–95M FY26 guide), down from ~23x at the $48.59 peak a de-rate that has room to run if the next guide wobbles.
• Still GAAP loss-making: Q1 carried an operating loss (inflated by Cycuity-related items); this is a cash-flow-inflection story, not a profitability story, so multiple support is thin on any disappointment.
• Management transition overhang: CFO Nicholas Hawkins retires 2026-08-31 (announced 2026-05-08); no named successor yet raises the stakes on the Q2 print and guide.

Setup & Price Structure

The chart is a completed parabolic top and early reversal. Structure ran ~$19.56 (early April) → $38.99 ATH break (6/12) → $48.59 closing peak (6/30), then reversed to a ~$32–34 intraday low zone (7/8) before bouncing to ~$36.61 (7/10). The ~$39 prior-ATH shelf that acted as the breakout pivot is now overhead resistance rather than support the single most important level flip on the chart. Price sits below the 20-day and is mean-reverting toward longer MAs after trading ~50% stretched above them at the peak. The July bounce lacks a confirmed higher low. A constructive re-base would need price to hold above the ~$32–34 reaction low, build a higher low, and reclaim ~$39 on volume before any clean continuation setup exists. Small float (~44M shares) means gappy, high-beta tape in both directions the reason the round trip was this violent and the reason a re-entry should be sized conservatively even if structure repairs.

Catalyst Calendar (next 30 days)

• ~2026-08-04 (est.) Q2 2026 earnings. The binary. Consensus tracks the $23–24M Q2 guide; the read that matters is the variable-royalty line (+67% YoY in Q1). Some calendars list 2026-08-11 date unconfirmed. Any pre-announcement blackout begins ~3 trading days prior.
• Ongoing automotive/hyperscaler design-win disclosures. SiEngine (late June) and Li Auto (June) show a live cadence; a further named win before the print would support a re-base thesis.
• 2026-08-31 CFO Nicholas Hawkins retirement (announced 2026-05-08); a successor announcement can land any time and is a sentiment swing factor into the print.
• No other hard-dated company catalyst inside the next 30 days.

What Would Change Our Mind

The bear-on-the-tape read flips constructive if price stops reversing and rebuilds: holds the ~$32–34 July low, prints a higher low, and reclaims the ~$39 shelf on expanding volume that re-establishes the clean continuation setup the June breakout briefly offered, and a fresh entry becomes valid at that retest. On fundamentals, a Q2 print (~Aug 4) that keeps variable royalties compounding above +50% YoY with another guide raise would confirm the royalty annuity is still inflecting and justify paying up again. Conversely, a weekly close below $33 breaks the early-July reaction low and converts the blowoff reversal into a full trend break, and a royalty line decelerating below +25% YoY on the print would break the thesis outright either one keeps the name a stand-aside.

Correlation Notes

AIP levers the same AI-semiconductor beta as the large-cap complex (TSMC, NVIDIA, Broadcom, Marvell, HBM suppliers) its royalty base rises and falls with the same custom-silicon and data-center build cycle, so it adds exposure to that theme rather than diversifying it. It is more volatile than the megacaps on that shared beta because of its ~44M-share float, which amplifies both the June melt-up and the July flush. The automotive-AI silicon angle (SiEngine, Li Auto, ADAS wins) is a partial diversifier away from pure data-center demand, but still rides the broad AI-chip narrative. Within a book already carrying AI-semi exposure, this is a concentrated, higher-beta expression of the same trade, not an independent one.

Notes on Theme State

Theme (ai-chip-infra-memory) remains ACCELERATING at the sector level; the individual name has shifted into a corrective, post-blowoff phase where its own price structure is broken despite intact fundamentals. That divergence accelerating story, broken chart is the whole read: the story is not the problem, the entry location is.