Rating: Neutral (Cautiously Constructive) | Target Price: HKD 45–55 | Current Price: HKD 44.5 (close on 2026-08-31) | Margin of Safety: approx. +1% (vs. base-case fair value floor of HKD 45) | Time Horizon: 12–18 months
AAC Technologies is a global platform leader in precision components for smartphones, with 2026H1 revenue of RMB 14.51 billion, a record high for the period (+8.9%), outperforming the industry against the headwind of declining global smartphone shipments (Q1/Q2 YoY -4.1%/-6.7%). Recurring net profit excluding fair value gains/losses rose +38.9% YoY, with gross margin rising to 22.4%. The thermal management business (+~400%) and new directions such as AI glasses and robotics form a second growth curve; management has guided that AI-related new businesses could "exceed RMB 8 billion" in 2027.
The current price of HKD 44.5 corresponds to approximately 13.3x consensus sell-side 2027E PE, with PE(TTM) at roughly the 23rd percentile of the past 5 years, implying normalized earnings of about RMB 3.05 billion—below the company's recurring earnings trajectory (two-year CAGR 30%+), suggesting valuation is "fairly low"; however, PB sits at the 81st percentile of the past 5 years, indicating re-rating has partially occurred, and the base-case fair value floor of HKD 45 is only +1.1% above the current price—a thin margin of safety. The key constraints are the expected record decline in global smartphone shipments in 2026 (IDC: -16.7%) amid the memory chip crisis, and extremely high customer concentration (top five customers account for 70.9%). The Q3 earnings release in early November 2026 is the nearest and most information-rich checkpoint to validate the "AI thermal management hedging handset downturn" thesis; we recommend using it as the decision point for adding or trimming positions.
Future Changes: Over the next 12–24 months there are three trackable trajectories of change—① Thermal business ramping from RMB 1.67 billion in 2025 (+410.9%) and RMB 1.10 billion in 2026H1 toward RMB 2.5–3.0 billion for full-year 2026 (iPhone 17 Pro adopts VC vapor chamber for the first time, with AAC as a core supplier); ② AI new businesses (thermal/liquid cooling/AI glasses/robotics) climbing from about 5–8% of revenue toward the 2027 guidance of RMB 8–9 billion (liquid cooling from ~RMB 200 million to nearly RMB 1 billion; robotics-related RMB 400–700 million); ③ the evolution of the memory crisis determining whether smartphone shipments can recover in 2027 (Xiaomi has raised its 2026 shipment target to 110 million units, betting on a memory reversal).
Primary Stock Price Drivers: ① 2026H2 elasticity from Apple's new-device stocking (Goldman Sachs expects H2 revenue up +53% QoQ, implying full-year revenue of ~RMB 36.7 billion/+15.3%); ② the impact of falling total handset volumes amid the memory crisis on the core business that accounts for 60–70% of revenue; ③ the pace of conversion of AI new businesses from design wins to revenue, which determines whether the valuation framework can shift from "consumer electronics components" to "AI hardware platform."
Demonstrable Expectation Gap: The current price implies normalized earnings of about RMB 3.05 billion (back-solved under a 15x assumption), between 2026E (RMB 2.68 billion) and 2027E (RMB 3.43 billion) sell-side estimates; only 13.3x assigned to 2027E, with PEG at ~0.79—the market is pricing "steady-state traditional business + partial delivery of AI new businesses" and has not paid a sufficient premium for the thermal business's five-year, multi-billion path. However, this expectation gap is "fairly low" rather than deeply mispriced: PB at the 81st percentile of 5 years shows re-rating has partially occurred, and hitting 2026E consensus requires H2 net profit attributable of ~RMB 1.78 billion (report basis +8.7% YoY); memory price increases are the main downward-revision risk. The divergence between the two ends of the sell-side anchors is real: CICC 50 / Huatai 60 / Goldman Sachs 63.6 HKD vs. Jefferies 43 HKD (Hold, "management execution to be validated").
Validation Catalysts and Falsification Conditions: Upside validation: September iPhone 17 Pro teardown (VC supplier and usage volume), November Q3 results (revenue QoQ, thermal segment, gross margin holding above 22%), year-end robotics mass production. Falsification: Q3 revenue weakening YoY or gross margin falling below 21%, memory prices not peaking in Q4, or loss of flagship model design wins. Evidence grade B: core changes (thermal +400%, CDU mass production, IDC -16.7%) all verified against interim report originals and industry sources; the sell-side anchor divergence at both ends is real. The system's probability-weighted expected value across three scenarios at the current price is ~HKD 46.7 (+4.9%), with symmetric odds slightly skewed upward.
Key Evidence
It must be emphasized that the Group's +8.9% is the result of stacked business mix: all growth came from non-handset or low-base businesses, while optics and mid-frame segments directly tied to handset volumes are actually contracting—"outperforming the industry" holds at the Group level but not for handset-exposed segments. The H2 test has not yet been passed: IDC's full-year -16.7% forecast implies H2 shipments down roughly -27% YoY, and the interim report's risk section explicitly notes "rising memory costs due to supply chain constraints, posing resistance to the smartphone market."
Key Evidence
Full-year thermal of RMB 2.5–3.0 billion implies H2 of RMB 1.4–1.9 billion, up 27%–73% QoQ from H1's RMB 1.1 billion—almost entirely betting on the ramp of the single iPhone 17 Pro product. Revenue scale of AI glasses components (Quark S1 five-microphone array + bone conduction, microphone design win with a leading overseas customer) has not been quantitatively disclosed; confidence weighting for including it in the RMB 8–9 billion guidance must be conservative. The market bear side (Jefferies) has directly questioned "management execution to be validated."
Key Evidence
FCF basis note: based on the cash flow statement's "payments for purchase of fixed assets" of RMB 1,599 million, 2026H1 FCF = 19.14 − 15.99 = RMB 315 million; on the company's capex basis of RMB 1.231 billion, it would be RMB 683 million (2025H1 same basis ~RMB 1.45 billion, ~-53% YoY). The rise in payables days to 135 is not inconsistent with the interim report's "cash conversion cycle improved by 30 days to within 20 days" (receivables 76 days + inventory 76 days − payables 135 days ≈ 17 days), but its mechanical effect is OCF supported by occupying supplier cash—operating cash flow deterioration excluding this effect is worse than apparent. The one-off lower tax rate contributed to H1 reported net profit; a reversion to 15–16% in H2 will suppress reported growth by about 6–7%.
Key Evidence
Three basis notes: ① "Consensus" is actually 4 verifiable samples (CICC/Huatai/Goldman Sachs/Jefferies) and is sensitive to sample composition; ② the PE(TTM) denominator includes 2026H1's one-off 9.4% low tax rate and 2025 fair value gains, compressing the multiple—percentile is for reference only; ③ the divergence between PB (81st percentile of 5 years) and PE (23rd percentile of 5 years) is attributed to earnings growth outpacing the stock price—the market has only partially re-rated, and this is not a direct contradiction between "undervalued" and "overvalued."
Key Evidence
Short-selling data note: the 31.88% short-selling ratio on 28 August is a single-day short-selling-to-turnover ratio (previous two days: 23.5% and 15.6%); it is a single-day sentiment indicator, not outstanding short positions, and cannot serve as evidence that "risk is priced" or of structural bearishness. Quite the opposite: the most bearish named sell-side's target price is below the current price, indicating some downside risk has not yet been absorbed by the market price.
| Metric (RMB) | FY2023 | FY2024 | FY2025 | 2026H1 |
|---|---|---|---|---|
| Revenue (RMB 100M) | 204.1 | 273.28 (+33.8%) | 318.17 (+16.4%) | 145.06 (+8.9%) |
| Net profit attributable (RMB 100M) | 7.40 | 17.97 | 25.12 (+39.8%) | 9.01 (+2.9%) |
| Recurring net profit (excl. one-off items, estimated) | — | 18.34 | 22.52 (+22.8%) | 8.71 (same basis +38.9%) |
| Gross margin | — | 22.1% | 22.1% | 22.4% (+1.7pp) |
| Net margin | 3.6% | 6.6% | 7.9% | 6.2% |
| Operating cash flow (RMB 100M) | 46.33 | 52.03 | 71.83 | 19.14 (-33.8%) |
| Free cash flow (RMB 100M, OCF − payments for fixed assets) | 32.69 | 31.32 | 46.02 | 3.15 |
| Cash + short-term time deposits (RMB 100M) | — | — | 86.12 (2025-12-31) | 70.72 (as of 2026-06-30) |
| Interest-bearing debt (RMB 100M, excl. lease liabilities) | — | — | 96.17 | 93.14 |
| Gearing ratio (loans + bonds/total assets, company basis) | — | — | 19.6% | 19.7% |
| Net debt (RMB 100M) | — | — | 10.05 | 22.42 (net gearing 4.7%) |
| EBITDA (RMB 100M) | — | 55.5 | 63.24 (+13.9%) | TTM 62.80 (+6.3%) |
Reasons for Metric Changes (items ≥±20% YoY, per official company explanations)
2026H1 revenue of RMB 14.506 billion (+8.9%) hit a record high for the period, but net profit attributable of RMB 901 million rose only +2.9%—reported growth clearly diverged from revenue for three reasons: ① 2025H1 included a RMB 320 million fair value gain on PSS contingent consideration that did not recur this period (other gains and losses -87.8% YoY); ② the effective tax rate fell to 9.4%, partially offsetting this; ③ recurring net profit excluding fair value gains/losses was RMB 871 million, +38.9% on the same basis—core earnings actually accelerated. Gross margin of 22.4% (+1.7pp) benefited from a higher mix of high-margin acoustics/electromagnetic actuation and thermal businesses. Market reaction confirmed the divergence: the day after results (8/20), the stock closed down 6.87% at HKD 38.8 (profit-taking triggered by net profit missing expectations on a reported basis), then recovered on buybacks and bullish sell-side notes, closing at HKD 44.5 on 31 August (+8.54% on the day, the same day Huatai set a HKD 60 target price). Relative to sell-side expectations: CICC said "net profit was slightly below expectations but new business profitability beat expectations" and raised its 2027 net profit forecast by 7%; hitting CICC's 2026E net profit attributable of RMB 2.68 billion requires H2 attributable profit of ~RMB 1.78 billion (report basis +8.7% YoY; recurring basis ~+7.6%), with memory price increases squeezing BOM costs as the main variable.
Business model snapshot: An asset-heavy precision manufacturing platform—vertical integration across multiple categories: acoustic devices, electromagnetic drives (X-axis linear motors), optics (WLG glass-plastic hybrid lenses/modules), precision structural components and thermal management (VC vapor chambers/liquid-cooling CDUs), and sensors (MEMS). It delivers packaged "acoustics + motors + thermal + structural" solutions to Apple (core supplier for iPhone/iPad/Mac/AirPods) and leading Android brands, raising per-device content and customer switching costs. Revenue has no subscription/recurring attributes and fluctuates with downstream device cycles; bargaining power is weak versus downstream (annual price-reduction mechanism with major customers) but strong versus upstream (concentrated scale purchasing). A gross margin of around 22% is maintained through technology exclusivity and yield barriers.
Cash conversion test: OCF/attributable net profit: FY2023 6.26 → FY2024 2.90 → FY2025 2.86 → 2026H1 2.12; FCF/attributable net profit: FY2024 1.74 → FY2025 1.83—well above 1 for multiple years, indicating excellent cash backing of book profits, with no signs of receivables build-up or capex eroding profits. The main marginal blemishes in 2026H1: OCF -33.8% and FCF (cash flow statement basis) of only RMB 315 million. Accounts payable turnover days rose to 135 days (+20 days each over two years), meaning part of the cash conversion is supported by stretching payables; inventory turnover days jumped from 62 to 76 days (unexplained by the company). If order cuts occur in H2, there is dual pressure from inventory writedowns and payables contraction. Recurring earnings test: FY2025 attributable profit RMB 2.512 billion vs. recurring RMB 2.252 billion, a +10.4% difference (mainly RMB 320 million fair value gain on the PSS contingent consideration)—below the 15% warning threshold, but the direction is "one-off gains inflating attributable profit"; the core earnings trend (recurring CAGR of 30%+ over two years) is genuine and healthy. The company does not disclose a self-reported Non-GAAP measure; recurring net profit in this report is self-calculated (excluding fair value gains/losses, FX effects, and disposal gains/losses).
Return on capital: FY2025 ROIC estimated at ~7.5% (NOPAT ≈ pre-tax profit RMB 2.887 billion × 88% ÷ invested capital RMB 33.97 billion)—the large book net assets (including RMB 1.82 billion goodwill from the PSS acquisition) depress the ratio, while marginal returns on capital (new thermal revenue/new capital) are significantly above the stock average. The sub-8% headline ROIC should not be directly equated with a "fake growth" warning, but it is not a >15% wide-moat signal either.
Maintenance capex: CapEx/depreciation FY2023 0.49 → FY2024 0.69 → FY2025 0.88, consistently below 1—a cash-cow characteristic. Expansion has been driven mainly by M&A (PSS RMB 3.45 billion, Dispelix RMB 620 million, Yuandi RMB 174 million, etc.) rather than asset-heavy organic build-out, at the cost of goodwill accumulation (see E4).
Moat and red flags: The moat comes from precision manufacturing know-how and patents (WLG wafer-level glass, ultra-thin VC, X-axis motors), Apple's multi-category certification barriers, and multi-component packaged solution capabilities. Red flags: ① High and rising customer concentration—top five customers at 70.9% (2026H1), two customers at 47.1% (2025: three customers at 59.2%; 2024: three customers at 55.5%); ② Optics segment gross margin of only 10.2%, the weakest segment; ③ Automotive acoustics growing revenue without profit (revenue +23.1%, gross margin -3.2pp).
Words-deeds consistency: pragmatic, good track record of delivery. ① The 2025 annual report promised robotic finger actuators would "ship in mass production on schedule, with revenue exceeding RMB 100 million within the year" → confirmed in the 2026 interim report: robotic motors and modules won design wins from multiple leading customers, with partial mass production; complete units expected in mass production by end-2026 (delivered). ② Thermal business +410.9% to RMB 1.67 billion in 2025 → 2026H1 continued at ~+400% (delivered). ③ April 2025 announcement promised buybacks of up to HKD 1.2 billion within 12–18 months (up to HKD 2.0 billion including automatic buybacks) → FY2025 buybacks of HKD 1.113 billion + 2026H1 HKD 332 million + continued buybacks in July–August (delivered). Management gives no quantitative revenue/profit guidance; outlooks are mostly directional, with no high-profile missed commitments. However, "AI new business reaching RMB 8–9 billion by 2027" is a forward-looking vision, and Jefferies' caveat that "execution remains to be proven" deserves respect.
Shareholder friendliness: friendly. Payout policy of 15%; actual FY2024 14.4%, FY2025 14.1%; final dividend per share HKD 0.10 (FY2023) → 0.24 (FY2024) → 0.35 (FY2025), rising with profits. No share issuance/rights issues in the past three years; the 2016 share award scheme expired in March 2026 and never issued new shares to the trustee—award shares were paid from repurchased treasury shares, so no equity dilution. Cumulative buybacks from FY2024–2026H1 of ~HKD 1.44 billion equivalent; buybacks continued in August 2026 at HKD 40.0–40.5.
Risk signals: Fast M&A pace (cumulative consideration exceeding RMB 4.5 billion over the past two years), goodwill rising to RMB 2.72 billion; Dispelix booked a loss of RMB 10.23 million upon consolidation with zero revenue. Step-by-step buyouts of minority interests in Chenrui Optics (2024–2026 cumulative ~6.4% equity for ~RMB 1.19 billion) with the difference credited to attributable equity—neutral-to-positive internal consolidation. No public anomalies found regarding concentrated selling by major shareholders or directors/supervisors/executives.
| Segment | Revenue share | Gross margin | YoY | Business logic in one sentence |
|---|---|---|---|---|
| Acoustics & Electromagnetic Drives | 41.4% | 28.6% | +14.2% | Vertically integrated premium acoustics + motors core base; robotic motors/cooling fans as new curve |
| Precision Structural Parts & Thermal | 22.4% | 23.1% | +12.5% | Driven by thermal +400%; metal frames -13.5%; AI server liquid cooling as new market |
| Automotive Acoustics | 15.4% | 20.1% | +23.1% | Global delivery (PSS sites), system-level solutions; competition intensifying, margins under pressure |
| Optics | 14.1% | 10.2% | -22.7% | Weakest segment: hit by smartphone volume decline; hedged via ASP and premiumization (7P/1G6P) |
| Sensors & Semiconductors | 5.9% | 14.6% | +41.0% | High-SNR microphones + MEMS, entry point for AI voice interaction; rapid growth |
(Segments sum to 99.2%; remainder is other/eliminations)
Primary profit segment: Acoustics & Electromagnetic Drives contributes ~52.9% of segment gross profit (RMB 1.72 billion ÷ 3.25 billion) on 41.4% of revenue—the undisputed profit engine; the thermal segment contributes ~23.1% of gross profit, automotive ~13.8%. Gross margin structure spread: the gap between the highest (Acoustics & Electromagnetic Drives 28.6%) and the lowest (Optics 10.2%) is 18.4pp—the former is a technology-exclusive component in a duopoly, while the latter is a scale-competition market dominated by Sunny Optical and Largan, where AAC chases differentiation via WLG; the difference in bargaining power is the root cause. Under the old FY2025 classification, optics gross margin was 11.5% (+5.0pp), showing an improving trend, but the volume decline in 2026H1 interrupted it again.
Accounting red flags: ① Continuous lengthening of accounts payable days (severity: medium)—115 days (+20 days vs. end-2024) → 135 days (a further +20 days in 2026H1); operating cash flow inflated by stretching payables for two consecutive periods, raising sustainability doubts (FY2025 results / 2026 interim). ② Large swings in effective tax rate (severity: low)—15.6%→9.4%; lower tax burden boosted current profit; the Group still has unrecognized tax losses of RMB 6.87 billion (2026 interim · taxation). ③ Rapid goodwill and intangible asset accumulation driven by M&A (severity: low)—goodwill RMB 2.72 billion (+19.0%), intangibles +44.7% to RMB 2.37 billion; Dispelix loss-making upon consolidation (2026 interim · Note 19).
Cross-period consistency: ① Inventory turnover days 60 (end-2024) → 62 (end-2025) → 76 (2026H1), rising continuously with a 14-day jump in six months, unexplained by the company—track in Q3/annual results whether this relates to stockpiling; ② Other gains and losses swing widely with the fair value of the PSS contingent consideration (-RMB 83 million → +RMB 251 million → +RMB 27 million), disturbing net profit comparability; management's explanation is consistent with disclosures; ③ Decline in OCF/attributable net profit ratio (6.26→2.90→2.86) reflects recovery from a low profit base; absolute OCF hit new highs each year, consistent with management's statement that "operating cash flow reached a record high." Apart from unexplained inventory days, no obvious accounting manipulation was found; lengthening payables days is the only item requiring continued vigilance.
Current market data (close 2026-08-31): Share price HKD 44.5 (+8.54% on the day), market cap ~HKD 53.3 billion; PE(TTM) 17.6x (percentile: 35% over 1 year / 12% over 3 years / 23% over 5 years / 24% over 10 years); PB 1.85x (81st percentile over 5 years); dividend yield 0.79%; EV/EBITDA (2025A) ~7.6x. Note on PE percentile: TTM earnings include one-off low tax rate in 2026H1 and 2025 fair value gains, so the percentile is for reference only; the PB percentile is significantly higher than the PE percentile, indicating earnings are growing faster than the share price and the market has only partially re-rated. Liquidity is normal (Stock Connect eligible; 8/31 single-day turnover of HKD 689 million, ~1.3% of market cap), no low-liquidity warning.
Peer comparison:
| Company | PE(TTM) | PB | Gross margin | Recent growth | Core difference |
|---|---|---|---|---|---|
| AAC Technologies | 17.6x | 1.85x | 22.1% | 2026H1 revenue +8.9% | Multi-category component platform; leader in acoustics/motors |
| Goertek | 21.6x | 2.27x | 11.8% (FY2025) | 2026H1 revenue +6.6% | No.2 in acoustics + VR/AR ODM; heavy ODM characteristics |
| Sunny Optical | 13.9x | 2.31x | 19.7% (FY2025) | 2026H1 revenue +11.5% | Optics leader (lenses + automotive + XR); most direct valuation anchor |
| Largan Precision | 40.7x | 5.22x | 49.4% | 2026Q2 net profit +352% | Niche monopoly in premium lenses; low comparability |
| Q-Tech | 4.3x* | 0.99x | 6.6% (2026H1) | 2026H1 attributable profit -10.5% | Price war in mid/low-end modules; PE distorted by low base |
*Q-Tech's PE is distorted by 2025 disposal gains from an associate plus a low base; excluded from the median. Peer PE median ~17.7x; AAC is in line with the median and at ~30% premium to Sunny—the premium must be supported by AI thermal/multi-category differentiation.
Market-implied expectations: At the current price of HKD 44.5, applying a fair multiple of 15x (a conservative midpoint between the peer median of 17.7x and the company's own 5-year median of 22x) implies normalized earnings of ~RMB 3.05 billion—between sell-side estimates for 2026E (RMB 2.68 billion) and 2027E (RMB 3.43 billion), corresponding to an implied 2025–27E attributable profit CAGR of ~10%; equivalently, the current price implies only 13.3x on 2027E consensus. Against reality: recurring net profit TTM ~RMB 2.496 billion with a two-year CAGR of 30%+, recurring net profit +38.9% in 2026H1, and management guidance of AI new business revenue exceeding RMB 8 billion by 2027. In one sentence: the current price requires the company to deliver only the lower bound of a ~10–17% earnings CAGR range, while both the demonstrated earnings trend and the AI new business guidance exceed that level—market expectations are reasonably low, but the 81st-percentile 5-year PB shows the AI option is already partially priced in.
Three-layer value (EPV): Asset value floor ~HKD 23.5/share (BVPS RMB 20.11 × 1.1664; the company is in a net debt position of RMB 2.24 billion, so the floor has limited meaning); EPV with zero growth ~HKD 24.6/share (normalized EPS HKD 2.34 = recurring net profit ~RMB 2.4 billion ÷ 1.1985 billion shares, WACC 9.5%); of the current HKD 44.5, ~55.4% is supported by zero-growth EPV (HKD 24.63/share) and ~44.6% (HKD 19.9) is a growth option—corresponding to AI thermal on a multi-billion-RMB five-year path, automotive acoustics, AI glasses/robotics, and other new businesses. The growth option share is not small; whether it is worth paying for is decided by scenario distribution and implied-expectation math: the implied growth at the current price is below the company's demonstrated earnings trend, so the option is priced conservatively rather than excessively, but the EPV layer means that if the AI narrative is falsified, downside is significant.
Three scenarios and odds (probabilities are this report's judgment):
| Scenario | Fair range (HKD) | Probability | Key swing factors |
|---|---|---|---|
| Bear | 23–29 (center ~26) | 25% | H2 Android order cuts + memory squeezing gross margin below 21% + AI new business misses: 2026 attributable profit ~RMB 2.0 billion, zero growth in 2027, at 12–15x (anchors: Sunny at current 13.9x; AAC's own low 5-year PE percentile) |
| Base | 45–55 (center ~50) | 55% | H2 revenue surges sequentially as expected, AI thermal/AI glasses ramp per guidance, memory headwinds contained: 2027E attributable profit RMB 3.1–3.3 billion, at 15–17x (anchors: midpoint between peer median 17.7x and own 5-year median 22x, independent of the company's current 18x) |
| Bull | 60–67 (center ~63) | 20% | AI new business fully delivers RMB 8–9 billion in 2027 + memory crisis eases: 2027E attributable profit ~RMB 3.6 billion, at 17–19x (anchors: lower end of own 3-year PE median 26x and 5-year median 22x + scarcity of AI thermal; cf. Goldman Sachs TP 63.6) |
Bear-case anchor check: the most pessimistic credible published forecast is Jefferies at HKD 43 (Hold, 2026-08-25)—this bear-case range of HKD 23–29 is significantly below that anchor, covering the valuation implications of the market's most pessimistic scenario; the base-case center of HKD 50 falls between Jefferies' 43 and Goldman Sachs' 63.6, not exceeding either end of the market distribution.
The current price of HKD 44.5 sits at roughly the 30th percentile of the three-scenario distribution: upside of +1.1% to +23.6% versus the base range of HKD 45–55; downside of -48.3% to -34.8% versus the bear range of HKD 23–29; upside of +34.8% to +50.6% versus the bull range of HKD 60–67—roughly symmetric, slightly skewed upward. Probability-weighted expected value ~HKD 46.7 (+4.9%).
In-house earnings forecast (results reconciliation anchor):
Conclusion: Assess quality (improving, direction clear) separately from price (reasonably low, thin margin of safety)—judgment: reasonably low (modestly undervalued). Target price range HKD 45–55 (= base-case fair value), +1.1% to +23.6% versus the current price. This is not a deeply undervalued stock, but an observe-and-allocate opportunity with "symmetric, slightly upside-skewed odds and dense catalysts"; Q3 results are the directional decision point.
Industry Size: AAC truly competes in four tracks within "smartphone/AI terminal precision components." The demand base is global smartphone shipments: 1.26 billion units in 2025 (+2%, Counterpoint), with 2026 facing the steepest decline on record amid the memory chip crisis—IDC's latest August 2026 forecast of -16.7% to about 1.04 billion units (ASP +27.6% to US$581), while Counterpoint forecasts -13.9% to 1.08 billion units (lowest since 2013); IDC further expects China's decline to widen to around 20% in 2026H2, with ASP unlikely to return to 2025 levels before 2027. However, the structure shows a divergence of "declining volume + rising mix": GenAI phone penetration rising from 36% in 2025 to 45% in 2026 (Counterpoint; cumulative shipments surpassing 1 billion units in 2026Q3); foldables up +20~21% in 2026 (Counterpoint; Apple's iPhone Fold potentially accounting for 22%+ of 2026 foldable shipments, IDC); AI glasses with global shipments of 8.7 million units in 2025, roughly doubling YoY (China Academy of Information and Communications Technology), +130.1% globally in 2026Q1 (IDC), and Chinese-made AI glasses projected at 22.67 million units globally in 2026 (+56.3%). By track size: global smartphone lens market (including modules) at about US$21.08 billion in 2025 (Fortune Business Insights; aggregator-sourced, pending verification); no first-tier institution has published market sizes for acoustic devices and haptic motors (aggregator figures carry low reliability, used only for cross-reference); no authoritative size for vapor chambers (VC), but AAC's standalone thermal revenue reached RMB 1.67 billion in 2025, up +410.9%, indirectly confirming the track's expansion speed. Data center liquid cooling is a larger option: market estimates put liquid cooling procurement at RMB 74.1/176.1 billion in 2026/2027 respectively (+138% YoY; third-party estimate, pending verification).
Supply Chain and Value Distribution: Upstream comprises MEMS wafers, optical glass/plastic pellets, NdFeB magnetic materials, and metals/copper—mostly fully competitive commodity components where AAC's scale-driven purchasing gives it strong bargaining power. However, in 2025–2026 memory price increases are transmitted through device BOMs, compressing brand budgets and indirectly squeezing component price cuts (magnet steel and other rare-earth-containing materials are also pushed up by export controls, already reflected in the -3.2pp automotive gross margin). The midstream—AAC's segment—is component design and precision manufacturing, with gross margins showing a clear gradient: high-end optical lenses highest (Largan's 2025 gross margin of 50.4%), acoustics/motors next (AAC at 22.1%), and modules/structural parts thinnest (Q-Tech at just 6.6% in 2026H1); value accrues mainly to "component design with the highest technical barriers + precision processes." Downstream consists of Apple, leading Android brands, and automakers (automotive acoustics revenue of RMB 4.12 billion in 2025, or 12.9% of total)—customers are concentrated with long certification cycles and strong bargaining power over component suppliers. AAC's bargaining power over downstream is medium-to-weak, offset by sole-design wins and yield barriers.
Supply-Demand and Competitive Landscape: Acoustic devices are a duopoly of AAC and GoerTek (dual leaders on Android; AAC/GoerTek/Luxshare splitting iOS—broker report basis, dated data pending verification), with concentrated capacity and no large-scale new entrants. In optical lenses, Sunny Optical ranks first (about 594 million units in 2025, >28% market share—Chaodian Intelligence basis, pending verification), Largan second (about 546 million units), while AAC differentiates with WLG glass-plastic hybrid lenses as a challenger in scale. In motors, AAC leads with Japanese players (TDK/Alps) coexisting. Structural parts/thermal management has many players (Lingyi, Lens, Feirongda, etc.); AAC added 150 million units of ultra-thin VC capacity in early 2026, with total capacity to exceed 300 million units, and ultra-thin VC market share in domestic high-end models has surpassed 50% (Hong Kong Commercial Daily, July 2026 basis). Entry barriers: precision manufacturing know-how and patents, Apple's certification cycles and switching costs, billion-unit shipment scale diluting yields, and multi-component vertical integration. Competitive intensity: modules > lenses > acoustics/motors—price wars are fierce in mid/low-end lenses/modules (Q-Tech gross margin at 6.6% and falling further YoY), with profit concessions emerging even in high-end lenses (Largan 2025 net profit -18%); industry profits are concentrating in technology-monopoly segments.
Cycle and Regulation: The industry is in a divergent phase of "declining volume cycle + structural upside from AI-driven premiumization." Leading indicators: ① memory chip prices and BOM costs (the core driver of this downturn); ② monthly/quarterly shipments (Counterpoint 2026Q2 -11%); ③ Largan's monthly revenue (its chairman noted momentum declining month by month in Jan–Feb 2026); ④ component suppliers' gross margins (Q-Tech at 6.6% in 2026H1); ⑤ quarterly AI glasses shipments. On regulation: US-China tariffs are driving "China+1" restructuring of the Apple supply chain (US-bound iPhones shifted to India; iPad/Mac/Watch to Vietnam, under a 20% Vietnam-US tariff agreement), forcing AAC to build overseas capacity—raising costs but deepening customer ties. Memory price hikes are a quasi-exogenous shock; China has no major restrictive policies targeting component suppliers. The direction is neutral-to-negative, but the policy environment for premiumization and new AI categories is friendly.
Peer Benchmarking (see table in Section IX): GoerTek posted 2025 revenue of RMB 96.55 billion (-4.4%; net profit attributable of RMB 3.94 billion, +47.9%) but gross margin of only 11.8%, reflecting heavy EMS characteristics. Sunny Optical's 2025 revenue was RMB 43.23 billion (+12.9%; net profit RMB 4.64 billion, +71.9%) with 19.7% gross margin—the most direct valuation anchor in optics (PE 13.9x, at the 7th percentile of 5 years). Largan's 2025 revenue was NT$61.15 billion (about RMB 19 billion), gross margin 50.4% but net profit -18%. Q-Tech's 2025 revenue was RMB 20.88 billion (+29.3%) but 2026H1 gross margin only 6.6%. AAC's 2025 revenue of RMB 31.82 billion (+16.4%) sits between Sunny and Q-Tech, with a 22.1% gross margin—the highest of the four (excluding Largan)—validating its positioning as a "high-margin component platform."
Company Positioning in the Industry: A platform leader (niche leadership + diversified catching-up). Global No. 1 in acoustic devices (duopoly with GoerTek; media reports claim coverage of 90%+ of global flagship acoustic solutions—unattributed basis, pending verification); first tier in X-axis linear motors; a challenger in optics (2025 optics revenue of RMB 5.73 billion, a fraction of Sunny's optics-related revenue, entering Apple/Huawei supply chains via WLG differentiation); thermal management (VC) and automotive acoustics (a top-ten global automotive audio supplier) are new growth poles. In 2026H1, revenue still grew +8.9% against an industry headwind of -11% shipments, with both share and per-unit content rising. Moat sources: precision manufacturing know-how and patents, Apple's multi-category certification barriers, and bundled multi-component solution capabilities. Share trends: traditional acoustics/motors stable; thermal/sensors/automotive share rising rapidly; key risks are volume declines under the memory crisis and annual price-down pressure from the Apple chain.
Overall View: stance = neutral (cautiously constructive), confidence = 0.60, time horizon 12–18 months. AAC is a name with "improving quality + an emerging AI second curve + a reasonably valued but thin margin of safety": recurring profit CAGR of 30%+ over two years, excellent cash conversion, and a solid buyback and dividend record. However, the current price offers only +1.1% safety margin versus the lower bound of base-case fair value, and four constraints—memory crisis, customer concentration, tax rate normalization, and marginally weakening cash flow—cap near-term upside. Probability-weighted expected value is about +4.9%, with roughly symmetric, slightly upside-skewed odds—not suitable for chasing; better suited to an observe-then-allocate strategy with Q3 results as the decision point: if Q3 revenue growth lands at the upper end of Citi's +9%~11% range, the thermal segment jumps sequentially, and gross margin holds above 22%, the probability of converging toward the base/bull scenarios rises; if revenue weakens or gross margin falls below 21%, reassess at a retest of the HK$38–40 buyback-heavy zone.
Risk Disclosures: ① Memory crisis deterioration leading to H2 order cuts (IDC implies -27% H2 shipments); ② annual price-downs from the top two customers (47.1% of revenue) exceeding expectations or share transfers; ③ continued automotive acoustics margin decline; ④ impairment of RMB 2.72 billion goodwill from the Dispelix and other acquisitions; ⑤ H2 tax rate normalization reducing reported net profit by 6–7%; ⑥ the 135-day payables model amplifying cash flow pressure in an order-cut scenario; ⑦ elevated single-day short-selling ratio (31.88% on 8/28) amplifying volatility.
Monitoring Points: September iPhone 17 Pro teardown and supply chain production schedules; early-November Q3 results (revenue YoY/QoQ, thermal segment, gross margin, tax rate); memory contract price sequential inflection; buyback frequency and upward shift in price range; major broker rating changes (CICC 50 / Huatai 60 / Goldman Sachs 63.6 vs Jefferies 43).
Primary data sources: AAC Technologies 2026 interim results announcement (2026-08-20), FY2025 annual results announcement (2026-03-19), 2025 interim results announcement (2025-08-21); IDC/Counterpoint industry forecasts; CICC/Jefferies/Goldman Sachs/Huatai research reports (as cited in news); earnings call transcripts (ifeng.com/Zhitong Caijing). All multiple percentiles and scenario probabilities are this report's own estimates/judgments; sell-side target prices not independently verified against original texts are noted as such.