ACMR · ACM Research, Inc. · Stock research

Current Thesis

The order-book leg an investor is buying here got stronger, and the tape stopped paying for it. On 2026-08-07 ACM Research reported Q2 2026 revenue of $292.9M, +36% YoY, with non-GAAP diluted EPS of $0.61 against a $0.34 consensus, and raised the FY2026 revenue floor to $1.125B from $1.080B (top end unchanged at $1.175B). On the call, CEO David Wang put first-half 2026 new orders at +105% YoY — an acceleration from the +65% reported at the 2026-05-07 Q1 print. Against that, price round-tripped: the 52-week high of $126.89 was set somewhere inside the 2026-07-26 to 2026-08-14 window that brackets the print, and the last completed close was $80.42 on 2026-08-14, -36.6% from that high, RSI(14) 50.5. The $90–94 shelf flagged in the 2026-07-26 note — the level that framed the June breakout — has been lost, and the 2026-06-18 record close of $109.87 is now overhead supply. That combination is the whole read: a compounding order curve and a broken price structure, with nothing scheduled to arbitrate before the Q3 print in early November.

Bullish and bearish views on ACM Research, Inc.

The model's bull view on ACM Research, Inc. (ACMR), in brief: Orders doubled, not merely grew. First-half 2026 new orders +105% YoY (CEO commentary, Q2 call 2026-08-07), versus +65% YoY at the Q1 print on 2026-05-07. Orders lead the revenue line by several quarters in wafer-fab equipment. Beat and raise, both sides. Q2 revenue $292.9M… The bear view: A beat-and-raise that could not hold a bid. Both cases follow in full.

Bull Case

• Orders doubled, not merely grew. First-half 2026 new orders +105% YoY (CEO commentary, Q2 call 2026-08-07), versus +65% YoY at the Q1 print on 2026-05-07. Orders lead the revenue line by several quarters in wafer-fab equipment.
• Beat and raise, both sides. Q2 revenue $292.9M (+36% YoY) versus $269.3M consensus; non-GAAP diluted EPS $0.61 versus $0.34 consensus; FY2026 guidance lifted to $1.125B–$1.175B, framed by management as 25–30% annual growth (2026-08-07).
• 2027 pre-loaded. CFO Mark McKechnie described 2027 as "starting to shape up" as a good growth year, citing unfulfilled 2026 orders spilling into the next period plus new platform contributions (Q2 call, 2026-08-07). Ultra C Tahoe platform expansion was announced alongside the print.
• Addressable market framed by management. Wang projected China's domestic WFE market above $80B by 2029 on the 2026-08-07 call — the localization leg the multiple has been paying for.
• Targets sit well above the tape. Craig-Hallum raised its target to $100 from $81 on 2026-08-10; the average published target stood at $115.14 against the $80.42 close of 2026-08-14 (stockanalysis.com). Roth carried $125 from its 2026-06-17 step-up.
• Multiple compressed against rising estimates. $5.60B market capitalisation on 69.65M shares outstanding, P/E 37.55 as of 2026-08-14 — the same P/E arithmetic that looked stretched at $110 now sits on a raised revenue floor.

Bear Case

• A beat-and-raise that could not hold a bid. The stock is -36.6% below its $126.89 52-week high with the print already banked. When the best quarter on record fails to produce a higher high, the marginal buyer is being supplied by someone.
• Margin direction is the wrong way. Non-GAAP gross margin 46.0% in Q2 2026 versus 48.7% a year earlier; non-GAAP operating margin 19.2% versus 19.3%.
• The raise came off the floor. FY2026 guidance moved to $1.125B–$1.175B — the low end lifted, the $1.175B ceiling untouched. That is a narrowing, and it caps the upside surprise the guide can carry into Q3.
• Estimates have not been marked to price. An average target of $115.14 against a $80.42 close (2026-08-14) leaves roughly 43% of implied upside unreconciled; a heavy tape into September invites target cuts rather than a re-rating.
• Dilution is approved and undated. ACM Shanghai's board approved a Hong Kong H-share listing on 2026-05-26 — up to 7% of post-offering capital plus a 15% over-allotment, priceable at board discretion inside a 24-month window. Asked on 2026-08-07, Wang said the company "really cannot comment too much."
• The export-control binary was not addressed. No substantive discussion of export controls on the Q2 call, while revenue remains concentrated in China fabs and ACM subsidiaries already sit inside the US control net from December 2024 actions.
• Insider supply into the June strength. An ACM Shanghai senior VP sold 18,750 shares (~$1.59M) on 2026-06-04; Director David H.
• Retrospective-return coverage arrived at the top. Benzinga's "Here's How Much You Would Have Made Owning ACM Research Stock In The Last 5 Years" ran 2026-08-13, six days after the print and with the stock roughly a third off its high.

Setup & Price Structure

• Reference levels. 2026-08-14 close $80.42. 52-week high $126.89 (set between 2026-07-26 and 2026-08-14). Prior record close $109.87 (2026-06-18). Old 52-week high $94.21, the ceiling of the $90–94 shelf that broke. June base low close $76.29 (2026-06-05).
• Momentum. RSI(14) 50.5 and a three-month return of +26.3% as of 2026-08-14 — the drawdown has already been partly absorbed, so this is not an oversold washout print, and no higher low has been confirmed since the high.
• The narrative is saturated. Dating it: the 2026-08-07 beat-and-raise produced no sustained new high; the 2026-08-13 five-year-returns retail piece is retrospective mainstream coverage; published targets ($100 Craig-Hallum 2026-08-10, $115.14 average 2026-08-14) cluster far above a price that has already given back the move. Well-known story, still fundamentally working, thin new bid. A weekly close that loses the June base would move the label to dead; sustained weekly closes back above the $94 shelf would argue the structure is repairing rather than distributing.
• Crowding observables, stated plainly. Sell-side targets 40%+ above spot and un-cut; retail-facing performance content clustering days after the high; an approved but unpriced equity issuance at the operating subsidiary; insider sales dated into the June advance; and no earnings inside the next 30 days, so nothing scheduled forces a repricing either way.
• Sector context. Third-party coverage in August 2026 attributed part of the decline to sector-wide selling across semicap and materials peers rather than a company-specific headline (AlphaStreet, August 2026) — relevant because it means the drawdown is not yet evidence that the China-localization thesis itself has been repriced.

Catalyst Calendar (next 30 days)

• ~2026-08-31 (est.) — ACM Shanghai (STAR 688082) H1 2026 interim report. Exact date unconfirmed.
• ~2026-11-05 (est.) — Q3 2026 print. Outside the 30-day window. This is the point: the 30 days ahead contain no company-scheduled event, so price action through September is flow and sector beta rather than news resolution.

Elapsed catalysts

• No fixed date — HKEX filing for the H-share offer. Approved 2026-05-26 with a 24-month window; a listing application or pricing announcement is the dilution headline, and it can land in any week. (passed 92d ago)
• No fixed date — US export-control action touching ACM Shanghai. Structural, unscheduled, and not addressed on the 2026-08-07 call. (passed 19d ago)

What Would Change Our Mind

The structure that carried this name out of June is already gone. The $90–94 shelf named on 2026-07-26 broke, the $126.89 high round-tripped inside weeks, and the prior published break level was taken out — so the question is no longer whether the breakout holds but whether a base forms above the June low. The gradeable break from here: a weekly close below $76 loses the 2026-06-05 base (low close $76.29) and the last higher low, at which point the fundamental acceleration is being ignored outright by the tape and the label goes to dead. Secondary conditions that would break the thesis independently: the FY2026 floor guided back under $1.125B at the Q3 print (~2026-11-05, est.); non-GAAP gross margin printing below the 42% bottom of management's stated band; or an HKEX pricing that lands into a weak tape. On the other side, the read repairs if weekly closes reclaim the $94 shelf with the order curve confirmed at Q3 — that would restore an accelerating structure rather than a distribution pattern.

Correlation Notes

• Semicap beta. Moves with AMAT, LRCX, KLAC, ONTO, MKSI and ENTG on WFE-cycle and capex headlines; August 2026 declines were peer-wide per third-party coverage, so single-name attribution of the drawdown is unsupported.
• China localization complex. Correlated to STAR-listed peers (Naura, AMEC) and to SMIC / Hua Hong capex commentary;
• Memory capex. The second demand leg runs through Micron, SK Hynix and Samsung DRAM/NAND spending — the same leg Roth cited in its 2026-06-17 step to $125. A memory capex deferral hits ACMR's 2027 order conversion, not its 2026 guide.
• Policy factor. Export-control and Entity List news is a shared shock across the China-exposed WFE group; it does not respect chart levels and has no calendar.