GoerTek: Stand-Alone Q2 Grew 0.67%, FY2025's Profit Jump Was Largely an Accounting Gain, and CNY 23.51 Already Pays the Base Case

The most important near-term finding is the one the half-year headline conceals. Q1 2026 revenue was CNY 18.659 billion, up 14.44%, and attributable profit was CNY 0.503 billion, up 7.28%. H1 revenue was CNY 40.046 billion, up 6.65%, and attributable profit CNY 1.466 billion, up 3.46%. Subtracting Q1 from H1 gives stand-alone Q2 revenue of CNY 21.388 billion versus CNY 21.245 billion a year earlier: growth of only 0.67%. Q2 attributable profit was CNY 0.963 billion versus CNY 0.948 billion: growth of only 1.57%. On the cleaner adjusted-profit measure, Q2 growth was about 3.38%. These are direct arithmetic from the two exchange filings.

H1 2026 was a sharp Q1-to-Q2 deceleration, not a uniformly accelerating recovery. The filings identify pressures from component procurement and stocking, foreign-exchange movements and a still-soft smart-hardware business; the company does not disclose stand-alone Q2 segment revenue, so a precise allocation between product-cycle timing, customer mix and base effects cannot be proven. H1 smart acoustic revenue did rise 35.9%, while smart hardware fell 2.2% and precision components rose only 2.0%, making a mix/product-cycle explanation plausible but not fully quantifiable.

The second correction concerns the multi-year revenue narrative. The filed series is down-up-down: FY2023 revenue fell 6.03% to CNY 98.57 billion; FY2024 rose 2.41% to CNY 100.95 billion; FY2025 fell 4.36% to CNY 96.55 billion. Any description of H1 2026 as ending “two consecutive years of revenue decline” is inconsistent with the annual reports. The company has instead experienced a multi-year plateau around CNY 100 billion, with composition and profitability changing more than total sales.

The third correction is more important for valuation. FY2025 statutory attributable net profit rose 47.85% to roughly CNY 3.94 billion even as revenue fell. Gross-margin recovery was real: precision-component margin rose 2.01 percentage points, and smart-hardware margin rose 2.20 points. Yet smart acoustic margin collapsed from 9.47% to 3.85%, and the annual report recorded CNY 2.142 billion of non-recurring remeasurement gain when GoerTek Optical ceased being consolidated. Total non-recurring profit attributable to shareholders was about CNY 2.46 billion. Adjusted attributable profit was only about CNY 1.48 billion, down 38.17%, rather than up 47.85%.

The FY2025 “revenue down, profit up” scissors was partly operational improvement and predominantly a below-the-line accounting event. That does not erase the genuine margin improvement in precision components and smart hardware; it does mean the headline net-profit growth rate is a poor foundation for a forward P/E. The strongest evidence of operational repair appears in segment gross profit and H1 2026 adjusted earnings, not FY2025 statutory profit.

This distinction produces an unusually wide valuation range. At CNY 23.51 and roughly CNY 83.5 billion equity value, trailing statutory earnings imply about 20.9 times P/E. But rolling FY2025 adjusted profit forward with H1 2026 produces only about CNY 1.56 billion of trailing adjusted earnings, equivalent to roughly 53.5 times. Neither number is wholly satisfactory: 20.9 times is flattered by the optical remeasurement gain; 53.5 times capitalizes a trough year in the acoustic-assembly business. A normalized forward owner-earnings approach is therefore more informative.

Customer concentration is the other defining fact. The top five customers represented 88.75% of revenue in 2023, 88.57% in 2024 and 79.05% in 2025. The largest customer alone declined from 43.38% in 2023 to 31.96% in 2024 and 29.72% in 2025. Diversification is moving in the right direction, but 79% remains extraordinary concentration for a business with limited ability to set finished-product pricing. Overseas revenue represented 87.72% of FY2025 sales and 84.34% in H1 2026.

History shows why that risk belongs in valuation rather than merely in boilerplate. In November 2022 GoerTek disclosed that a major overseas customer had instructed it to stop production of one smart-acoustic finished product, with estimated FY2022 revenue impact of up to CNY 3.3 billion. The customer and product were not identified in the filing. Media and analysts linked them to Apple and AirPods Pro 2, but GoerTek itself did not confirm that attribution. The stock hit its daily down limit on November 9 and again the following day.

GoerTek Microelectronics adds a second valuation layer. The parent approved a Hong Kong spin-off plan, and the July 2025 filing described GoerTek as owning approximately 83.4% of the sensor subsidiary. Application materials characterized the subsidiary as the world's largest acoustic-sensor supplier by 2024 sales, with 43% share, and as a top-five intelligent sensing and interaction supplier overall. HKEX's application index now marks the relevant application as lapsed; the exchange's 2025 index records the July 21, 2025 submission, and no subsequent active application under the same applicant name was identified through the 2026-08-30 base date.

Two lapsed filings are meaningful because there is no active capital-markets event to value today. They do not reveal why the applications lapsed. A lapse is not evidence by itself of rejection, regulatory trouble or a failed valuation negotiation. A future IPO could make GoerTek's higher-quality sensing asset easier to value and raise external capital, while simultaneously diluting the parent's economic interest and potentially creating a holding-company discount. Until a new filing provides offering size and valuation, any quantified “IPO unlock” is speculation.

Vertically, GoerTek has proved that it can migrate from individual acoustic parts into increasingly integrated consumer-electronics subsystems and finished hardware. Horizontally, that breadth is both the moat and the discount: AAC Technologies more cleanly represents higher-margin component economics; Luxshare represents a much larger, more diversified integrated manufacturing platform; Sunny Optical concentrates on optics; BYD Electronic shows how thin large-scale assembly economics can become. GoerTek sits in the middle, with better proprietary component content than a generic EMS company but far less pricing power than a pure IP supplier.

The qualitative portrait is “company in transition.” It is transitioning from a company whose capital-markets identity depended heavily on a few blockbuster consumer-electronics assembly programmes toward one where precision components, sensors, optical know-how and AI/XR reference designs are supposed to carry a larger portion of value. That transition is real enough to reject a structural-decline label: precision-component margins are near 25%, customer concentration has fallen from its 2023 peak, and H1 adjusted profit grew. It is incomplete enough to reject a high-quality-compounder label: group economics remain tied to assembly, gross margin remains in the low teens, customer concentration remains extreme, and H1 operating cash flow turned negative.

Vertical history and financial anatomy

GoerTek began in Weifang in June 2001 as Weifang Yilida Electroacoustic Co. The institutional logic was straightforward: China was becoming an increasingly important electronics-manufacturing base, while miniature electroacoustic components were becoming more important inside mobile and portable devices. The predecessor was reorganized into the joint-stock company that became GoerTek Acoustics in 2007. Founder Jiang Bin brought an engineering background in electronics and decades of acoustic-industry experience; the business grew around component engineering and high-volume manufacturing rather than around a consumer brand.

The Shenzhen IPO on May 22, 2008 sold 30 million shares at CNY 18.78 each, implying gross proceeds of about CNY 563 million and 120 million shares outstanding after the offering. The prospectus positioned the company around miniature electroacoustic components, R&D, customer relationships and manufacturing expansion. That starting point matters because today's precision-component segment is closer to the original economic core than the later finished-device assembly businesses.

The company's post-listing history is best understood in four stages rather than as a chronology.

The first stage, from founding through the early post-IPO years, was acoustic-component specialization. GoerTek built the process engineering, tooling, miniature acoustics and customer qualification capabilities needed to become embedded in global consumer-electronics supply chains. The economic attraction of this stage was technical content per unit; the constraint was that component suppliers still negotiated against very large downstream brands. Over time, major-customer qualification became an asset in itself because it allowed the company to move from supplying a part to participating in more complex modules and complete products. The cost was growing customer concentration. Company responses to the exchange show the top five customers accounted for 56.28% of sales in 2016, 65.86% in 2017 and 66.45% in 2018.

The second stage was the painful 2018 reset followed by the TWS/wearables expansion. FY2018 revenue fell about 7%, attributable net profit dropped roughly 59% and ROE fell to 5.73% from 16.58%. FY2019 revenue then jumped 48% to CNY 35.15 billion, while operating cash flow more than doubled to CNY 5.45 billion. The business was beginning to capture the expansion of true-wireless audio and more integrated consumer-electronics programmes.

That rebound became the third stage, the 2020–21 capacity and category boom. Revenue reached CNY 57.74 billion in 2020 and CNY 78.22 billion in 2021; 2021 attributable profit was CNY 4.27 billion and ROE 17.61%. GoerTek issued CNY 4 billion of convertible bonds in 2020, adding capital during a period of rapid capacity and product expansion. By then, the company was no longer an acoustic-components story alone. TWS assembly, wearables and VR hardware had created a business able to scale revenue far faster than the original component franchise.

The fourth stage began in 2022 and still defines the company. FY2022 revenue reached CNY 104.89 billion, up 34.1%, but attributable profit fell 59.1% to CNY 1.75 billion and ROE fell to 6.17%. The revenue-profit divergence signalled what happens when manufacturing scale grows faster than economic quality. Then came the November 2022 major-customer production suspension. GoerTek said only that a major overseas customer had stopped one smart-acoustic finished-product programme and that the revenue impact could reach CNY 3.3 billion. The market treated it as evidence that a concentrated contract manufacturer could lose years of expected earnings very quickly: the shares suffered consecutive daily-limit declines.

In hindsight, that customer event genuinely changed GoerTek's capital-market identity. The effect outlasted the two-day stock shock. FY2023 revenue declined 6.03% to CNY 98.57 billion, gross margin sank to 8.95%, and attributable profit was only around CNY 1.09 billion. The company subsequently emphasized higher-value precision parts, margin restoration, broader smart hardware and the monetization or restructuring of specialized subsidiaries.

The financial recovery since then has been uneven but visible.

† FY2022 reported series differs by product classification and is shown approximately here. ‡ H1 2026 filing shows 14.72% gross margin for the main electronic-components business; consolidated revenue/cost arithmetic is approximately 14.5%. Sources: company annual and interim filings.

The revenue line tells a mature-cycle story more than a structural-growth story. GoerTek multiplied revenue severalfold during the TWS/VR expansion, but since FY2022 the group has remained clustered around CNY 100 billion. The task now is to increase profit per yuan of sales rather than simply restore the old revenue trajectory. That is possible because the mix still contains large volumes of low-margin finished-product work. It is difficult because those assembly businesses provide the scale, customer access and fixed-cost absorption that support the rest of the system.

Gross margin shows the turn more clearly than revenue. The group moved from 8.95% in 2023 to 11.09% in 2024 and about 11.85% in 2025. The first two steps came from both cost repair and mix: in 2024 precision-component gross margin reached 21.51%, acoustic assembly 9.47%, and smart hardware 9.17%. In 2025 precision rose further to 23.52% and hardware to 11.37%, while acoustic assembly moved violently in the opposite direction to 3.85%. Group improvement did not mean every business improved.

That 3.85% acoustic-assembly margin is a useful reminder of customer bargaining power. A finished-device assembler can carry large revenue and industrial importance while creating little gross profit. CNY 22.98 billion of smart-acoustic revenue in FY2025 generated only about CNY 0.88 billion of segment gross profit by simple revenue-times-margin arithmetic. Precision components generated about CNY 4.23 billion of gross profit on CNY 17.98 billion of revenue. Smart hardware generated roughly CNY 6.11 billion because of its scale. The profit pool is thus split: hardware supplies absolute gross profit; precision components supply much better economics per sales yuan.

The FY2025 optical restructuring complicates the net-profit series. The annual report booked about CNY 2.14 billion of remeasurement gain after GoerTek Optical ceased to be consolidated, part of total non-recurring attributable income of approximately CNY 2.46 billion. The counterparty was Sunny Optical, one of the peers compared later in this report. GoerTek Optical issued CNY 529.51 million of new registered capital to acquire two Ningbo Sunny micro-nano optics companies in Shanghai valued at about CNY 1.903 billion; GoerTek's holding fell from 56.66% to 37.77%, and to 38.57% after a further capital injection, while Sunny took roughly 31% and became the second-largest shareholder. The deal closed on December 8, 2025, which is what moved the optical business out of the consolidated group and into equity-method accounting. Government grants and financial/FX derivative gains contributed additional non-operating items, while asset impairment was roughly CNY 1.04 billion. The adjusted profit of CNY 1.48 billion is therefore a more conservative measure of continuing earnings than the CNY 3.94 billion headline.

The optical transaction also changes segment comparability. In 2026 the parent continued optical activities through the GoerTek Optical associate rather than consolidating all optical revenue into precision components. That means the 2026 precision-component segment is economically cleaner but not perfectly comparable with earlier years. Investors should avoid interpreting every percentage-point change in segment mix as purely organic.

Cash conversion needs equally careful treatment. Across FY2021–FY2025, operating cash flow was approximately CNY 38.1 billion against about CNY 13.7 billion of cumulative attributable profit, a ratio around 2.8 times. That looks excellent until capital intensity is considered. GoerTek carries a large manufacturing asset base, so depreciation is added back in operating cash flow and recurring equipment spending consumes much of it. FY2023–FY2025 capex was roughly CNY 6.95 billion, CNY 3.63 billion and CNY 7.13 billion respectively. Simple operating cash flow less capex came to only about CNY 1.2 billion in 2023, CNY 2.6 billion in 2024 and slightly negative in 2025.

H1 2026 made the capital requirement more visible. Operating cash flow swung from positive CNY 1.87 billion a year earlier to negative CNY 0.245 billion, while cash paid for property, equipment, intangibles and other long-term assets rose to CNY 4.49 billion from CNY 2.87 billion. The filing attributes the operating-cash-flow deterioration mainly to increased raw-material purchases, production stocking and employee-related cash payments. This may unwind during the seasonal second half, but H1 free cash flow before disposals was deeply negative.

The balance sheet can fund that cycle, but it is no longer obviously cash-rich. At end-2025 GoerTek held about CNY 18.89 billion of cash, with CNY 13.41 billion of short-term borrowings and CNY 3.55 billion of long-term loans. By March 2026 short-term borrowings had increased 49.7% to CNY 20.08 billion, which management attributed to operating and funding needs. At end-H1 the consolidated asset-liability ratio was 60.23%, up from 58.85% at end-2025.

R&D is a genuine recurring cost of staying qualified rather than an optional growth experiment. FY2025 R&D expense was CNY 5.30 billion and reported R&D investment was CNY 5.03 billion, about 5.2% of revenue; only about 7.4% of R&D investment was capitalized. H1 2026 R&D expense increased another 10.6% to CNY 2.36 billion. These levels support the case that GoerTek has real engineering depth, while also limiting how much operating leverage investors should expect from a simple recovery in revenue.

Management control remains concentrated. Jiang Bin is chairman and a founder; the H1 2026 shareholder table showed GoerTek Group at 15.63%, Jiang Bin at 8.09%, Jiang Long at 7.05% and Hu Shuangmei at 0.46%. The filing identifies Jiang Bin and Jiang Long as brothers, Jiang Bin and Hu Shuangmei as spouses, and GoerTek Group as controlled by the Jiang brothers. Their directly disclosed aligned holdings add up to roughly three-tenths of the company before employee plans.

Capital allocation has mixed signals. The company completed a CNY 1.20 billion repurchase programme by April 2026, acquiring about 44.2 million shares, or 1.25% of total share capital, at CNY 20.35–34.09. Those shares were intended for employee stock ownership or equity incentives rather than automatic cancellation, so the transaction is better viewed as compensation financing/alignment than as a permanent reduction in share count. The FY2025 proposed cash dividend of CNY 0.20 per share was modest relative to the capital intensity of the business.

The capability GoerTek has proved over twenty-five years is industrial migration: it repeatedly converts component know-how into access to larger device-level programmes. Its unresolved problem is converting that industrial scale into consistently high returns on shareholder capital.

Its share-price history reflects that tension. The market rewarded the TWS/VR expansion as revenue and ROE accelerated into 2021; the 2022 customer shock then exposed how much that growth depended on a small number of programmes. The subsequent re-rating has followed margin recovery and AI/XR enthusiasm more than a return to peak revenue growth. The current CNY 23.51 price is about 42% below the 52-week high of CNY 40.68 reported by Google Finance, illustrating how quickly the thematic premium has already contracted.

Historical P/E percentiles are less useful here than they appear. FY2022–23 earnings were depressed by the customer shock; FY2025 statutory earnings are elevated by the optical remeasurement gain. Comparing today's headline 20–21 times P/E with a decade-long mechanical distribution would mix radically different earnings bases. The more relevant history is the market's changing identity for the stock: acoustic component growth company, then TWS/VR manufacturing growth company, then customer-loss recovery, and now an AI-glasses/component-mix re-rating candidate.

Business model, industry structure, and horizontal comparison

The most useful way to read GoerTek's segments is to separate revenue scale from economic quality.

Source: GoerTek FY2024, FY2025 and H1 2026 filings.

Precision components are the closest thing GoerTek has to a classic moat business. At a 24.86% H1 2026 margin, an incremental CNY 1 billion of component revenue is economically much more valuable than the same amount of low-margin assembly revenue. The business benefits from miniature acoustic design, sensor packaging, process know-how, reliability qualification and the ability to build at very high volumes. GoerTek Microelectronics is central to this economic layer.

Smart acoustic assemblies are the strongest evidence against treating all customer wins as equal. The segment recovered rapidly in H1 2026 after a disastrous FY2025 margin, but its history shows that volume can be huge while gross profit collapses. At 3.85% gross margin in FY2025, price concessions, yields, utilisation, new-product ramp costs or customer mix can overwhelm a large revenue base. The H1 recovery to 9.70% is economically meaningful and is one of the clearest metrics to monitor over the next year.

Smart hardware is the paradox. It is GoerTek's largest business and in H1 2026 produced the largest absolute segment gross profit, about CNY 2.71 billion by simple arithmetic, versus about CNY 1.93 billion for precision components. Yet much of the value belongs downstream to brands, platforms, displays, processors and content ecosystems. GoerTek can improve its share of economics when it supplies optics, acoustics, sensors and system engineering in addition to assembly, but a reference-design win and an ODM contract still deserve a lower valuation multiple than proprietary component IP.

The cost structure reinforces that hierarchy. Materials and direct labour vary with volume; factories, automated lines, tooling, engineering teams and customer-specific qualification costs do not disappear when a programme is delayed. R&D alone exceeded CNY 5 billion in 2025. Asset intensity then creates depreciation and maintenance capex even in a flat-revenue year. This combination produces operating leverage in good product ramps and reverse operating leverage during customer transitions.

The company's strongest moat is manufacturing-plus-engineering integration rather than brand or network effect. A global device customer that has validated a supplier's microphone, speaker, sensor, optics and finished-device process has reasons not to switch casually: new suppliers must pass reliability, yield and capacity qualification. GoerTek's scale also spreads tooling and process-development costs across large programmes. The 2022 production suspension shows that these switching costs are asymmetric. They make entry difficult for a new supplier but do not give GoerTek equivalent bargaining power against a dominant customer.

A second real moat sits in acoustics and sensing. GoerTek's roots predate the TWS and XR waves, and GoerTek Microelectronics' market position in acoustic sensors is evidence that the company owns meaningful process and product expertise rather than merely labour capacity. Application materials for the Hong Kong listing described the subsidiary as having shipped more than five billion sensors cumulatively and holding 43% of the global acoustic-sensor market by 2024 sales. Those are stronger indicators of technical position than the parent company's overall low-teens gross margin would suggest.

A third moat is breadth across acoustics, sensing, optics and system integration. CES 2026 illustrates the point: the company showed AI-glasses designs, waveguide modules, MEMS acoustic sensors, microphones, ASR processing and other interaction technology in one portfolio. That allows GoerTek to participate in a product architecture at multiple content points. The economic test is whether that breadth raises blended gross margin, not how many technologies appear at an exhibition.

The moat is real at the component and co-design layer and much weaker at the pure assembly layer. The consolidated accounts are the weighted average of those two realities.

The industry cycle is similarly bifurcated. Mature smartphones, headphones and wearables are governed by replacement cycles, inventory digestion and annual product launches. VR/MR depends on a still-fragile consumer adoption curve and new device introductions. AI glasses are in an early penetration phase in which unit growth can exceed 100% because the starting base is small. GoerTek faces consumer, inventory and technology-iteration cycles at the same time.

IDC's 13.6 million-unit 2026 forecast for display-less smart glasses is a useful reality check. A simple 18.9% CAGR from that base would imply roughly 16 million units in 2027, 19 million in 2028 and 23 million in 2029 before reaching IDC's 27.3 million 2030 forecast; these intermediate numbers are a mathematical interpolation, not IDC forecasts for each individual year. Retail ASP is expected to decline materially as volumes rise. This combination is excellent for component unit volumes but less automatically attractive for assemblers because OEM procurement normally seeks to capture scale savings.

The current category is also concentrated at the brand level. Counterpoint put one vendor at about 82% of H2 2025 smart-glasses shipments, while IDC put that vendor at 69.2% in Q1 2026 under its category definitions. That means a supplier's AI-glasses success can initially depend on winning very few customer programmes. The technology may diversify faster than supplier economics do.

VR creates the opposite problem. Global shipments fell 28% in 2025, and Q1 2026 declined another 17%. The weakness stems from limited new-device stimulus and still-incomplete consumer use cases. A company with major VR/MR assembly exposure cannot treat AI glasses as pure incremental growth while the older XR category is contracting.

Geopolitics primarily enters GoerTek through customers, manufacturing geography, FX and trade rules rather than through direct regulation of a licensed business. Overseas sales were 87.72% of FY2025 revenue and 84.34% in H1 2026. Foreign-exchange movements were already large enough to drive Q1 finance expense to CNY 344 million from a small negative expense a year earlier, even though derivative investment gains partially offset the movement. The company bears both operational and accounting volatility from currency exposures.

The peer landscape is best described as Scenario C: plenty of adjacent competitors, but no single perfect comparable. AAC Technologies is the closest listed acoustic/precision-component analogue. Luxshare Precision is the closest large integrated-assembly and components analogue. Sunny Optical is a clean reference for what specialist optics economics look like. BYD Electronic is a useful reference for large-scale device manufacturing with thin margins. Lens Technology is relevant in precision structural parts. Cirrus Logic and Knowles provide higher-IP audio/sensing references, but their semiconductor/specialty-component models are too different to use directly for GoerTek's consolidated valuation.

† Main electronic-products margin in GoerTek's filing was 14.72%; consolidated arithmetic is slightly lower. ‡ Sunny's latest market data and interim materials put gross profitability around the high-teens/20% level. Sources are company results and current financial-result services.

AAC became the component-centric version of what GoerTek might look like with much less finished-product assembly. Its H1 2026 group gross margin of 22.4% was roughly eight percentage points above GoerTek's. Its acoustics business itself carried still higher margins. Investors pay for more component content and less dependence on huge low-margin hardware revenue. GoerTek's precision-component margin is already comparable with that economics; the group margin is not.

Luxshare became the scale-and-integration winner. Its H1 2026 revenue was more than four times GoerTek's and rose about 40%, with attributable profit growth around 18%. It can accept low gross margins because its breadth, scale, customer programmes and operating efficiency convert enormous revenue into much larger absolute earnings. GoerTek competes with Luxshare most directly where finished devices and integrated manufacturing overlap, while GoerTek remains differentiated by its older acoustic and sensor heritage.

Sunny Optical became a specialist in optical value pools rather than a broad electronics assembler. H1 2026 revenue rose about 11.5% to CNY 21.9 billion and profit about 9.9% to CNY 1.81 billion. Its profitability is supported by optics content in smartphones, vehicles and emerging devices. The relationship is closer than pure competition. Sunny was the counterparty to the December 2025 GoerTek Optical share swap and now holds roughly 31% of that company against GoerTek's 38.57%, so part of the group's optical and micro-nano capability sits in a jointly held associate that GoerTek no longer consolidates. Sunny remains the cleaner optical exposure and GoerTek the broader manufacturing system, but on XR optics the two are partners as much as rivals.

BYD Electronic illustrates the danger on the other side. H1 2026 revenue was CNY 82.23 billion, but gross margin fell to 4.91% and attributable profit to about CNY 0.43 billion. Large assembly revenue alone does not create an economic moat. GoerTek's hardware margin is currently much healthier, but the peer shows why investors should value finished-device manufacturing on sustainable margin and cash return rather than on revenue TAM.

Cirrus Logic provides an intentionally different comparison. On 2026-08-28 its US-listed shares traded at about 13.3 times trailing earnings, according to market data, despite its far higher semiconductor/IP content; Knowles traded around 45 times, reflecting a different specialty-components portfolio and earnings base. Neither multiple should be transplanted into an A-share ODM. They mainly show how meaningless a single “acoustics peer P/E” can become once business models diverge.

Cross-border comparisons also need FX discipline. The CFETS USD/CNY spot closing rate was 6.7204 on 2026-08-28. A-share valuation can differ from Hong Kong and US valuations because liquidity, retail participation, domestic theme premiums, capital controls and comparable-company sets differ. The analysis below therefore derives GoerTek's fair value in CNY from its own earnings power first and uses peers as a reasonableness check, not as the valuation engine.

The GoerTek Microelectronics spin-off sits exactly at this intersection. Shareholders approved the Hong Kong separation plan in September 2024. Earlier public material showed that the subsidiary had reached a domestic ChiNext listing-committee stage in 2022 before the capital-markets path changed. The July 2025 Hong Kong application subsequently lapsed, and the HKEX index did not show a newer active GoerTek Microelectronics application through the research date.

As of 2026-08-30, GoerTek Microelectronics has no verified active Hong Kong IPO application in the HKEX material reviewed. That is the relevant status for valuation. A hypothetical subsidiary equity value of CNY 20–30 billion would make the parent's pre-dilution 83.4% stake worth roughly CNY 16.7–25.0 billion, or 20–30% of GoerTek's current market capitalization. Those numbers are sensitivity mathematics, not an estimate of what an IPO would price at. Without a live prospectus, offering size and market valuation, using them as a target-price uplift would be circular.

A successful listing could still matter in three ways. It could reveal a standalone market value for the group's highest-quality sensor assets, provide external capital for MEMS expansion and establish a liquid benchmark against which the parent's conglomerate discount can be measured. The offsets are dilution and the possibility that investors apply a holding-company discount after the asset is separately listed. Two lapses increase event uncertainty, but they do not tell us whether valuation expectations, timing, regulation or simply transaction execution caused the delays.

GoerTek's ecological niche is unusually clear. It is a technology supplier and integrated manufacturer sitting between specialist components and global consumer-device brands. Its most defensible profit pool is miniature acoustics/sensing and the engineering required to integrate those components into compact devices. Its largest revenue pool is hardware assembly. Competitors can take the first pool through better component technology; customers and larger EMS firms can squeeze the second through procurement power. The company's strategic task is to make the former grow faster than the latter without losing the scale benefits the latter provides.

Current fundamentals and expectation gap

The cleanest current picture starts with the Q1/Q2 bridge rather than the H1 headline.

Q2 is derived by subtracting the Q1 filing from the H1 filing; comparison periods are derived the same way.

The quarter-by-quarter pattern weakens the “accelerating AI hardware” interpretation. Q1 produced double-digit revenue growth; Q2 was almost flat. Q2 adjusted profit did better than revenue, which suggests the margin-repair programme remained alive, but 3.4% adjusted-profit growth is not the earnings acceleration implied by a high-beta AI-glasses narrative.

Segment data explain part of the gap. H1 smart acoustic assembly revenue rose 35.9% and its gross margin recovered to 9.70%. This is a genuine turnaround from FY2025's 3.85%. Precision-component margin reached 24.86%, another record-like level within the reported recent series, but revenue rose only 2.0%. Smart hardware revenue, where the AI/XR story lives most visibly, fell 2.2%; its margin nonetheless improved to 13.61%.

This gives the current earnings story an unusual shape: the fastest-growing segment is the previously damaged acoustic-assembly business, while the highest-margin segment is barely growing and the AI/XR-heavy hardware segment is shrinking in revenue but becoming more profitable. That is a better mix story than a growth story. The bull case needs the margin improvement to persist until the next device cycle restores top-line growth.

Cash flow is currently worse than earnings. H1 adjusted profit was CNY 1.12 billion, yet operating cash flow was negative CNY 0.245 billion. Capex was about CNY 4.49 billion. Management's filing attributes the operating cash movement partly to raw-material purchases and production preparation, which creates a plausible second-half reversal. The burden of proof is now on H2 cash generation: if inventory and customer ramps do not turn the working-capital build back into cash, the explanation shifts from timing toward structural capital consumption.

Foreign exchange also distorts operating interpretation. Q1 finance expense rose to CNY 344 million largely because of FX losses, while investment income from derivatives increased. In H1 finance expense was roughly CNY 759 million, up more than sixfold, while investment income and fair-value gains from hedges and investment products partly offset it. Investors should examine the net economic currency exposure across finance expense, derivative investment income and fair-value movements instead of reading any one line in isolation.

The balance sheet shows that management is willing to pre-finance the next production cycle. Short-term debt reached CNY 20.08 billion at Q1 versus CNY 13.41 billion at end-2025; H1 leverage stood at 60.23% of assets. This is manageable for a CNY 100-billion-revenue manufacturer with established customer receivables, but it makes the quality of the next cash conversion cycle a first-order issue.

The market is currently trading four overlapping expectations. The first is continued margin repair after the 2022–23 shock. The second is a stronger TWS/acoustic cycle. The third is AI glasses becoming a new device category through which GoerTek sells assembly plus acoustic, optical and sensing content. The fourth is an eventual valuation event around GoerTek Microelectronics. Only the first two are already visible in reported earnings; the AI-glasses and spin-off stories remain partly prospective.

Smart glasses deserve optimism about units but restraint about GoerTek's earnings capture. IDC's 2026 forecast of 13.6 million display-less units at about US$376 retail ASP corresponds to roughly US$5.1 billion of end-market value. Retail value includes the brand's software, processor, cameras, channels, lenses, frames, warranty and margin. An assembler earns a fraction of that. A microphone or sensor vendor earns an even smaller dollar value per unit but potentially much higher gross margin. GoerTek's valuation depends on which role it wins in each programme.

This is why the component portfolio is more important than the “AI glasses” label. If the category grows and GoerTek mainly supplies finished assembly, revenue can rise while the valuation-worthy profit uplift remains modest. If GoerTek supplies acoustic sensors, microspeakers, optics, waveguides, structural parts and system design alongside manufacturing, content per device and blended margin can improve materially. The company has shown capability across those layers, but it has not disclosed enough customer-level content to choose the second outcome as the base case.

The bull/bear disagreement is quite specific.

Bulls can point to the precision-component margin moving from roughly 20.5% in 2023 to 21.5% in 2024, 23.5% in 2025 and 24.9% in H1 2026; the hardware margin moved from roughly 7.3% in 2023 to 9.2%, 11.4% and 13.6% over the same sequence. Those trends suggest that management is successfully moving away from the least profitable work or improving manufacturing efficiency.

Bears can answer that the smart-acoustic margin collapsed once already, group revenue remains below FY2022, H1 operating cash flow was negative, and Q2 2026 revenue growth was under 1%. In other words, the market may be capitalizing several years of better mix before GoerTek has shown that the improvement can coexist with renewed top-line growth.

Bulls can also point to customer diversification: top-five concentration fell roughly 9.5 percentage points from 2024 to 2025, and the largest customer dropped below 30% of revenue. Bears can answer that 79.05% top-five concentration is still extreme and that the 2022 order suspension has already proved the transmission mechanism from one programme to revenue, utilization, earnings and valuation. Both statements are true; the disagreement is whether the direction of travel or the absolute level matters more.

The Microelectronics spin-off creates a similar split. Bulls see a sensor subsidiary with a dominant acoustic-sensor position whose standalone listing could expose value hidden inside a lower-margin consolidated manufacturing group. Bears see two lapsed Hong Kong filings and an event whose timing, dilution and valuation remain unknown. The evidence currently supports neither zero value nor an automatic IPO premium.

The current fundamental evidence supports a margin-recovery thesis; it does not yet support a high-growth earnings thesis. Q3/Q4 2026 need to show that smart-hardware revenue can turn positive, acoustic margin can stay around the high single digits or better, and working capital can convert back to cash.

The latest official financial set is the H1 2026 report filed in August; the company IR page also lists the April Q1 report and the FY2025 report. I found no reliable public consensus series broad enough to state that analysts as a group have raised or cut 2027 estimates, so no such claim is made here.

Valuation, cash-flow passthrough, and margin of safety

The first valuation question is what earnings number to capitalize.

At CNY 23.51 and about CNY 83.5 billion market capitalization, the reported trailing earnings series gives roughly CNY 3.99 billion of TTM attributable profit: FY2025 CNY 3.94 billion plus H1 2026 CNY 1.466 billion less H1 2025 CNY 1.417 billion. That produces about 20.9 times P/E, consistent with contemporaneous market data.

Replacing reported FY2025/H1 earnings with adjusted profit gives a very different number. FY2025 adjusted profit was about CNY 1.48 billion; adding H1 2026 adjusted CNY 1.116 billion and subtracting H1 2025 adjusted CNY 1.033 billion gives approximately CNY 1.56 billion of trailing adjusted profit. The current market cap is about 53.5 times that amount.

The correct normalized multiple lies between those extremes. The 20.9 times headline ratio capitalizes a CNY 2.14 billion optical remeasurement gain as though it were recurring. The 53.5 times adjusted ratio capitalizes FY2025's unusually weak acoustic-assembly profit as though it were permanent, despite the H1 2026 margin rebound. I therefore value GoerTek on forward owner earnings under explicit margin assumptions rather than choosing one of the two mechanical trailing multiples.

Cash-flow passthrough initially looks strong. Cumulative operating cash flow during 2021–25 was about CNY 38.1 billion against roughly CNY 13.7 billion of attributable profit, or around 2.8 times. The ratios in individual years were often above two because depreciation and other non-cash charges are heavy in a manufacturing business.

But free cash flow is the more relevant denominator. FY2023–25 capex absorbed roughly CNY 6.95 billion, CNY 3.63 billion and CNY 7.13 billion respectively. On simple OCF-minus-capex arithmetic, the three-year cumulative free cash flow was only around CNY 3.5 billion, despite cumulative statutory profit above CNY 7.6 billion over those years. The economic reason is that much of GoerTek's depreciation reflects machinery and factories that eventually must be renewed.

GoerTek does not disclose a clean maintenance-versus-growth capex split. My valuation estimate treats roughly CNY 3.5–4.5 billion a year as a plausible maintenance/replacement range under normal production, using FY2024's CNY 3.63 billion capex as a rough low-cycle anchor and the much larger FY2023/FY2025 spending as evidence that growth/tooling cycles add several billion above that baseline. This is an analytical assumption, not company guidance.

Under that construction, depreciation and maintenance capex largely offset in owner earnings, making normalized adjusted profit plus sustainable working-capital conversion a reasonable shorthand. My central 2027 owner-earnings estimate is CNY 3.4 billion. It assumes smart-acoustic margin remains near normalized high-single digits rather than FY2025's 3.85%, hardware margin remains in low-teens, precision components retain a low-to-mid-20s margin, and the group does not repeat FY2025's optical gain.

Historical valuation percentile gets a qualified answer: a precise percentile would create false precision because the denominator was severely depressed in 2022–23 and boosted by a large one-off in 2025. At approximately 21 times reported TTM earnings, the stock appears ordinary rather than euphoric; at more than 50 times adjusted TTM earnings, it appears expensive. On my normalized forward owner-earnings range of roughly CNY 2.8–4.4 billion, the market is paying about 19–30 times depending on which recovery case occurs.

Peer valuation does not solve that ambiguity. AAC's much higher group gross margin deserves a business-quality premium; Luxshare's scale, growth and diversification deserve a different multiple; Sunny's optics specialization is a different profit pool; BYD Electronic's 4.91% H1 gross margin illustrates why assembly deserves a discount. US-listed Cirrus Logic was at about 13.3 times trailing earnings on 2026-08-28, while Knowles was near 45 times, a spread so wide that the label “acoustic peer” carries little valuation information by itself.

The valuation scenarios below use approximately 3.56 billion shares and capitalize 2027 owner earnings. They intentionally exclude any explicit Microelectronics IPO premium; a successful listing at an attractive valuation would be an upside catalyst rather than something already required for base-case fair value.

This is valuation-scenario analysis within a research framework, not investment advice.

The multiple range is intentionally below what I would apply to a pure high-margin component/IP company but above a generic EMS business. Twenty-two times conservative owner earnings recognizes GoerTek's technology franchise but refuses to pay for a new growth cycle. Twenty-six times in the base case assumes the market continues to give an A-share premium to AI/XR exposure and component quality. Twenty-eight times requires both stronger earnings quality and credible multi-year device growth. The range is not cheap enough to ignore customer concentration, and it is not high enough to assume GoerTek becomes a pure technology-licensing company.

The market expectation embedded around CNY 23.51 is close to the base case. A CNY 24.8 fair value on CNY 3.4 billion of owner earnings means investors are already assuming that normalized earnings move materially above FY2025 adjusted profit and that H1 2026 margin recovery persists. The expectation gap is therefore skewed toward execution: simply reporting higher statutory profit is insufficient if cash flow, adjusted profit or segment margin disappoint.

The next earnings release should be judged on four variables before revenue growth itself: stand-alone quarterly smart-hardware direction, acoustic-assembly gross margin, operating cash flow and working capital, and whether precision-component growth reaccelerates. Any fresh GoerTek Microelectronics filing would be a separate capital-markets catalyst rather than a substitute for those operating metrics.

The margin-of-safety test is harsher than the base-case valuation.

First, CNY 23.51 is about 36% above the CNY 17.3 value implied by the conservative scenario. On that test, the current price offers no discount to conservative intrinsic value.

Second, the most fragile base assumption is owner-earnings normalization. If the CNY 3.4 billion base estimate reaches only 70%, owner earnings would be CNY 2.38 billion. At the same 26 times multiple and current share count, the resulting value is only about CNY 17.4 per share. That single sensitivity removes roughly 30% of base fair value.

Third, suppose earnings remain flat for three years and the valuation multiple does not expand. FY2025's proposed CNY 0.20 dividend represents only about a 0.85% cash yield at CNY 23.51. China's official ChinaBond 10-year government yield was 1.6949% on 2026-08-28. On that flat-earnings thought experiment, the equity's cash return is below the government-bond yield. There is no margin of safety at this buy price.

This is also not quite a “good company but bad price” setup. GoerTek is a good collection of component technologies embedded in a mixed-quality consolidated business. At today's price, the market asks the investor to underwrite successful normalization of that mix. A deeper discount is appropriate because the conservative case is not a disaster case; it merely assumes weaker device demand and lower earnings normalization.

Margin-of-safety sufficiency verdict: none.

Risks, catalysts, tracking framework, and research uncertainties

Customer concentration is the first permanent-loss risk. I assign medium-to-high probability to periodic customer/programme disruption and high impact if it affects one of the top two customers. The observable indicators are the annual top-five and largest-customer shares, smart-acoustic revenue/margin and any company disclosure of a production suspension. The transmission path is already documented: programme loss cuts revenue, leaves fixed capacity underutilized, compresses assembly margin, weakens market confidence in supplier status and causes the P/E to contract simultaneously. Top-five concentration of 79.05% in FY2025 is lower than the prior two years but still makes this the dominant risk.

A second risk is that the current margin recovery reflects favourable project mix rather than structurally greater bargaining power. I assign medium probability and high impact. Precision-component margins appear steady, but smart acoustic moved from 9.47% in 2024 to 3.85% in 2025 and back to 9.70% in H1 2026. That volatility is too large to treat the current number as a permanent floor. A drop below roughly 7% in acoustic assembly or below 10% in smart hardware for two reporting periods would signal that customer pricing and ramp costs are again absorbing the economics.

A third risk is that AI glasses become a large revenue story but a thin-profit story. I assign high probability to rapid category unit growth but medium-to-high probability that supplier economics lag the end-market narrative. IDC sees 13.6 million display-less units in 2026 and a path to 27.3 million by 2030, while ASP declines materially. If GoerTek captures mostly assembly, declining end-device ASP and concentrated procurement can limit gross profit. The observable indicator is whether smart-hardware margin rises with AI-glasses volumes and whether precision-component revenue grows faster than hardware revenue.

A fourth risk is capital intensity. Probability is medium; impact becomes high if the H1 pattern persists. H1 operating cash flow was negative CNY 0.245 billion while capex was CNY 4.49 billion, and short-term debt had already risen sharply in Q1. If a second consecutive half produces weak OCF while capex remains above CNY 4 billion, the market will have to reassess whether margin improvement is generating owner earnings or merely financing the next equipment cycle.

A fifth risk is the combination of FX, trade/geopolitical exposure and overseas customer dependence. Probability is medium and impact medium-to-high. More than four-fifths of sales remain overseas, while 2026 finance expense has already been distorted by currency losses. A geopolitical or tariff shock can affect customer demand, production location, logistics and valuation even if GoerTek itself is not directly sanctioned. The observable indicators are overseas-revenue share, Vietnam/other overseas capacity disclosure, finance expense and derivative gains, and any named export-control measures affecting customer programmes.

Valuation is the final amplifier. At approximately 21 times statutory TTM profit the stock does not look obviously expensive, but those earnings include the optical gain. If investors begin valuing the company on adjusted owner earnings before those earnings have normalized, the multiple is much higher. A narrative unwind can produce sharp price downside even without an accounting loss.

Positive catalysts are most powerful when they change one of those economics rather than merely add a product announcement. A Q3/Q4 print showing positive smart-hardware revenue growth together with acoustic gross margin around 10% and positive operating cash flow would establish that revenue and margin recovery can coexist. Precision-component growth returning to double digits would matter because that segment earns more than twice the gross margin of FY2025 acoustic assembly. A new active HKEX filing for GoerTek Microelectronics with clear offering economics could turn an uncertain SOTP story into a measurable event.

The largest negative catalysts are symmetrical: smart-hardware revenue remaining negative into a new device cycle; acoustic margin falling back toward FY2025 levels; another major-customer programme loss; H2 operating cash flow failing to reverse the first-half working-capital build; or a downward industry revision to smart-glasses shipments at the same time VR remains weak.

Sources: GoerTek filings, IDC and calculations described earlier.

The dashboard should be used causally. Precision margin tells whether the component moat is intact. Acoustic and hardware margins tell whether customer pricing is allowing GoerTek to keep any of the value created by volume. Customer concentration measures whether the 2022 failure mode is becoming less severe. OCF and capex determine how much accounting profit belongs economically to shareholders. Smart-glasses shipments test the external narrative, while owner-earnings valuation tests how much of that narrative is already prepaid.

For the next earnings date, the company had not published a clearly indexed 2026 Q3 appointment date in the IR material reviewed as of the research base date. Its IR page shows the 2025 third-quarter report was published on October 27, 2025. I use late October 2026 as the expected reporting window, not as management guidance; investors should replace this estimate with the official appointment once CNInfo/GoerTek IR posts it.

Research uncertainty remains in four places. First, GoerTek does not identify its major customers by name, so programme-level supplier rumours cannot be converted into forecasts. Second, the company does not disclose maintenance capex; owner-earnings calculations necessarily estimate it. Third, there is no live GoerTek Microelectronics prospectus as of the base date from which to derive a defensible IPO valuation or dilution. Fourth, the private Research Hub reports referenced in the brief for Luxshare and Sunny Optical were not accessible in this research session, so the competitive work above was rebuilt independently from public filings and current company/result sources rather than represented as a reading of those private reports.

The source hierarchy used throughout was: Shenzhen Stock Exchange/CNInfo and GoerTek IR for company financials; HKEX's application index for listing status; company filings/results for competitors; IDC and Counterpoint for device-market data; ChinaBond/CFETS for bond and FX data; and established financial-market services only for current quotations where an exchange closing-price feed was not directly available.

Cross-synthesis and final research conclusion

Looking vertically, GoerTek's history contains a consistent pattern. The company repeatedly found a new layer of integration to add on top of its original acoustic competence. It started with miniature components. It moved into modules, TWS finished products and then VR/MR and other smart hardware. It built sensors, optics and system-design capability around that manufacturing base. Revenue rose from CNY 35 billion in 2019 to almost CNY 105 billion in 2022 because that migration let GoerTek capture far more dollars per end device.

The same migration weakened the purity of the economics. A microphone business can earn attractive gross margins because a tiny component carries meaningful technical content. A complete headset or XR device contains much more revenue but also far more pass-through material, customer-directed design, labour and procurement pressure. GoerTek consequently became strategically more important to its customers while its consolidated margin remained low. That is why scale and quality cannot be treated as synonyms here.

The 2022 order suspension exposed the cost of that model. GoerTek had spent years becoming deeply embedded in a small number of global device customers. Deep integration created revenue growth and manufacturing scale; it also created the possibility that one product decision could hit utilization, earnings and valuation simultaneously. The top-five revenue share above 88% in 2023–24 showed that the issue persisted even after the original shock.

Management's answer since then has been economically sensible: restore margins, grow precision components, broaden smart-hardware categories and create standalone visibility for specialized assets such as GoerTek Microelectronics. The filings show evidence that the first two pieces are working. Precision-component margins have approached 25%, hardware margins have improved several percentage points, and top-five customer concentration fell to 79%. These are operational changes, not slogans.

The statutory FY2025 net-profit number nevertheless overstated the recovery. More than CNY 2.1 billion came from the optical deconsolidation remeasurement, and adjusted profit fell 38%. Investors who look only at CNY 3.94 billion of FY2025 attributable profit will conclude that GoerTek has already restored its old earnings power. The adjusted accounts say the restoration remained unfinished at year-end. H1 2026 is more encouraging because adjusted profit rose 7.96% without another transaction of that magnitude.

The Q2 arithmetic prevents us from taking that recovery too far. A 14.4% Q1 revenue increase became only 0.7% in Q2. Attributable-profit growth slowed from 7.3% to 1.6%. H1 segment margins were better, but the largest hardware business still declined in revenue. Thus the company enters H2 with healthier unit economics and less convincing top-line momentum.

Horizontally, GoerTek's best advantage is that no single adjacent peer reproduces its combination of acoustics, sensors, optics, XR engineering and large-scale finished-device manufacturing. AAC is more component-rich. Luxshare is larger and more diversified in integrated manufacturing. Sunny is more optically concentrated. BYD Electronic is more obviously an assembly economics reference. GoerTek's customer pitch is the ability to co-design several interaction technologies and then industrialize the whole device.

That breadth is valuable in AI glasses because weight, microphones, speakers, cameras, sensing, power consumption, optics and manufacturability interact. The CES reference designs make sense strategically. A 35-gram glasses design or compact acoustic sensor is exactly where an acoustics-plus-miniaturization company can add value.

Yet the size of the AI-glasses category does not presently justify treating GoerTek as a new secular-growth stock by itself. IDC's 13.6 million 2026 unit forecast is fast growth from a low base. VR remains in decline. One brand still dominates smart-glasses shipments. And GoerTek does not disclose the contract economics that would tell us whether it captures the category through low-margin assembly, higher-margin components or both.

The market's most likely misjudgment is not that AI glasses are fictitious. The category is growing. The misjudgment risk is equating category growth with GoerTek profit growth at the same rate. A CNY 1 billion increase in 25%-margin component revenue and a CNY 1 billion increase in 5–10%-margin assembly revenue have very different values. The share price ultimately needs the former type of growth, or a sustained improvement in the latter's margin.

The second likely misjudgment is on FY2025 earnings quality. Headline 20.9 times trailing P/E looks far more comfortable than the adjusted accounts. The appropriate response is not to insist the shares are “really” at 53 times either, because the acoustic business has already recovered from its 2025 trough. The correct answer is to normalize forward owner earnings and demand a discount for the uncertainty around that normalization.

Over the next twelve months, the critical variable is whether the Q2 slowdown was product-cycle timing or the beginning of another weak demand period. A second-half revenue acceleration with acoustic margin around 10%, hardware margin above 12% and positive operating cash flow would make the base case considerably firmer. A second half with flat hardware revenue, renewed acoustic margin pressure and weak OCF would move the conservative case toward the centre of the distribution.

Over three years, the key variable is mix. GoerTek needs precision components and other proprietary content to grow faster than pure assembly, or it needs device-level engineering to turn hardware assembly into a structurally higher-margin business. A group still dominated by roughly 10%-margin manufacturing deserves a manufacturing multiple even if the products contain AI. A group where a rising share of gross profit comes from 20–25%-margin sensing/acoustic/optical content can re-rate.

Over five years, the deepest question is whether GoerTek can reduce dependence on a handful of global consumer brands while retaining their scale. Customer concentration has improved but remains enormous. The sensor subsidiary, new device categories, automotive interaction and broader component platforms could diversify the company. The wrong path would be replacing one concentrated blockbuster programme with another and calling it diversification because the end-device category changed.

The Microelectronics listing could accelerate that five-year transition but should not drive today's base valuation. Its sensing position is strategically attractive. A live IPO would put a market price on an asset that probably deserves a different multiple from assembly. The two lapsed filings mean timing cannot be relied upon. A third filing with transparent finances and offering terms would be genuinely new information; continued silence is neutral-to-negative for near-term SOTP hopes but does not change the operating value of the subsidiary.

The balance sheet does not currently threaten the company, but cash economics place a ceiling on valuation. GoerTek can report operating cash flow comfortably above earnings because depreciation is large, then spend most of that cash on equipment. H1 2026 shows the opposite extreme: negative OCF combined with almost CNY 4.5 billion of capex. A company that needs recurring multibillion-renminbi manufacturing investment cannot be valued as though accounting profit were distributable cash.

This also explains why the current 11–15% consolidated gross-margin range is structurally important. A bull case can plausibly lift group margin through mix, but a move to 20% would require a profound change in what the company sells, not just better utilization. AAC's 22.4% group margin gives a real-world reference for a more component-heavy model. GoerTek would need precision/sensing/optical content to become much more economically dominant before that comparison became appropriate.

At CNY 23.51, the stock is not priced for failure. My base fair value of CNY 24.8 essentially says the market already discounts a sizeable owner-earnings normalization toward CNY 3.4 billion. The optimistic CNY 34.6 case requires a stronger device cycle, sustained margin gains and CNY 4.4 billion of owner earnings. The conservative CNY 17.3 case does not require another 2022-style disaster; it merely assumes weak hardware growth, partial acoustic normalization and a lower multiple. That asymmetry is why the margin-of-safety verdict is “none.”

The stock becomes much more interesting when price no longer requires the base case. Around CNY 13.0–13.8, an investor would be paying at least 20% below the conservative CNY 17.3 valuation, provided customer relationships and segment margins remain intact. A price collapse caused by a new major-customer loss would not automatically be a buying opportunity because the conservative earnings assumption itself would need to be rewritten.

Conversely, the research should be overturned to a more positive view if GoerTek produces several reporting periods in which precision-component revenue grows at double digits, hardware grows without giving back its 12–14% margin, acoustic margin remains around 9–10%, OCF normalizes above earnings, and customer concentration continues to fall. That combination would show that the transition is no longer merely a mix repair but a better business model.

The bearish reassessment is equally concrete. If smart-hardware revenue remains negative through a major new-product cycle, acoustic margin falls below 7%, top-five customer concentration moves back above 85%, and operating cash flow remains weak despite rising debt, the evidence would say GoerTek is still a volatile programme manufacturer with a few high-quality component assets rather than a transforming technology platform.

Bull reasons:

• Precision-component gross margin increased to 24.86% in H1 2026 from 21.51% in FY2024, evidence that the highest-quality consolidated business is improving economically.
• Smart-acoustic margin rebounded from 3.85% in FY2025 to 9.70% in H1 2026 while segment revenue rose 35.9%, showing meaningful recovery from the prior trough.
• Top-five customer concentration fell from 88.57% in 2024 to 79.05% in 2025, reducing, though not removing, the largest structural risk.
• GoerTek possesses credible AI-glasses content across acoustics, sensing, optics and system design just as IDC forecasts 13.6 million display-less smart-glasses shipments in 2026.
• GoerTek Microelectronics gives the parent exposure to a sensor asset described in listing materials as holding 43% of the global acoustic-sensor market, creating optional SOTP value.

Bear reasons:

• Stand-alone Q2 2026 revenue grew only 0.67% and attributable profit 1.57%, a sharp deceleration from Q1 despite the AI/XR narrative.
• FY2025's 47.85% statutory profit growth was dominated by a CNY 2.14 billion optical remeasurement gain; adjusted profit actually fell about 38%.
• Top-five customers still generated 79.05% of FY2025 sales, and the 2022 order suspension proved that customer concentration can cause simultaneous earnings and valuation damage.
• H1 2026 operating cash flow was negative while capex reached CNY 4.49 billion and leverage rose, weakening the passthrough from accounting profit to owner cash.
• Smart-hardware revenue fell 2.2% in H1 while worldwide VR shipments remained weak; fast-growing AI glasses are not yet large enough to erase the legacy XR cycle automatically.

Pre-mortem, first script: in 2027, a major overseas customer reallocates another high-volume acoustic or wearable programme after a quality, yield or strategic sourcing change. GoerTek's smart-acoustic revenue falls 20–25%, capacity utilization drops and segment gross margin moves from around 10% back toward 4–6%. Precision components cannot absorb the lost factory overhead quickly enough. Normalized group owner earnings fall toward CNY 2 billion, and the A-share market removes the AI/XR premium, compressing the multiple from roughly 25–30 times normalized earnings to 15–18 times. An equity value around CNY 30–36 billion would correspond to roughly CNY 8–10 per share, a decline of 55–65% from the research-date price. The 2022 production suspension provides the historical transmission mechanism, although this specific future script is hypothetical.

Second script: smart glasses grow in units through 2027–28, but brands force ASP reductions down the supply chain while GoerTek wins mostly finished-device assembly rather than component-rich content. At the same time VR stays weak. Smart-hardware revenue rises but its gross margin falls from 13.6% toward 9–10%; precision components grow only mid-single digits. Investors eventually realize that “AI glasses” added revenue rather than returns. Owner earnings plateau near CNY 2.5 billion, and a 18-times multiple implies about CNY 45 billion of equity value, roughly CNY 12–13 per share. That would be close to a 45–50% loss without any existential company failure. The industry's declining ASP trajectory and current concentration make this a credible stress test.

GoerTek is a more capable industrial company than its low consolidated margin suggests, because the group contains a genuine precision acoustic/sensing franchise. It is also a lower-quality consolidated business than the AI-glasses narrative suggests, because assembly remains the majority of revenue and the strongest customers retain substantial bargaining power. The post-2022 repair is real: component margins are higher, hardware margin is healthier and customer concentration has eased. The evidence does not yet show that GoerTek has converted those repairs into a durable new growth curve.

At CNY 23.51, I would not pay in advance for that conversion. The base valuation is close to the market price, the conservative value is materially below it, and the flat-earnings cash return does not beat the China 10-year government yield. The appropriate posture is ownership only for investors already comfortable with the transition and willing to monitor margins and customer concentration closely; new capital has a better risk/reward at a much larger discount.

What would change that judgment is evidence rather than another product theme: sustained positive smart-hardware growth, precision-component growth above group growth, acoustic margin remaining around 9–10%, full-year OCF comfortably above profit, lower customer concentration, or a new Microelectronics filing whose valuation makes the SOTP discount quantifiable. Without several of those, GoerTek remains a transition story whose share price already values a meaningful portion of the transition as successful.

【Company-profile scores】

• Fundamental quality: medium
• Growth: medium
• Moat: medium
• Financial soundness: medium
• Management credibility: medium
• Valuation attractiveness: medium
• Risk level: high
• Suitable investor type: cyclical / event-driven / long-term growth investors able to tolerate concentrated customer and product-cycle risk

【Investment rating】

• Rating: Hold
• One-line thesis: Component margins are recovering, but Q2 growth nearly stalled and the current price already discounts meaningful owner-earnings normalization.
• Current-price classification: acceptable hold
• Whether to wait for a better price: yes. For new money, I would require CNY 13.0–13.8 with intact customer relationships and no collapse in precision/acoustic margins. The opportunity cost is missing a faster AI-glasses ramp or a favourable Microelectronics IPO event.
• Target holding horizon: 3–5 years
• Expected annualized return: conservative approximately -7% to -8%; base about 9%; optimistic about 24%, based on illustrative three-year values near CNY 18, CNY 30 and CNY 44 respectively plus modest dividends.
• Max-loss risk: roughly 45–65% in a combined major-customer loss, margin compression and multiple-reset scenario; the stress zone is approximately CNY 8–13 per share, spanning the two pre-mortem scripts above.
• Reassessment triggers: smart-acoustic gross margin below 7% for two reporting periods; smart-hardware margin below 10%; top-five customer concentration returning above 85%; full-year OCF/net profit below 1.0x while capex remains elevated; or a new GoerTek Microelectronics filing that materially changes parent-level SOTP economics.

【Ideal Buy Price】13.0–13.8 CNY

Basis: this is at least 20% below the conservative scenario's CNY 17.3 estimated fair value and therefore supplies a valuation buffer before requiring the base-case recovery.

Acceptable hold price: CNY 21.1–28.5, approximately ±15% around the CNY 24.8 base-case fair value.

Clearly overvalued price: CNY 38.1–42.0, beginning at roughly 10% above the CNY 34.6 optimistic fair value. At that level the share price would require earnings outcomes beyond this report's already demanding optimistic assumptions.

【Valuation Range】

• current: 23.51 (close as of 2026-08-28)
• bear (conservative · ideal buy zone): [13.0, 13.8]
• base (fair · acceptable hold zone): [21.1, 28.5]
• bull (optimistic · above the clearly-overvalued line): [38.1, 42.0]

Other tickers mentioned

• 2018.HK: AAC Technologies is the closest listed reference for higher-margin acoustic and precision-component economics.
• 002475.SHE: Luxshare Precision is the large-scale integrated manufacturing and components comparison.
• 2382.HK: Sunny Optical is the specialist optics comparison and an important reference for XR optical value pools.
• 0285.HK: BYD Electronic illustrates the economics and margin risk of large-scale finished-device manufacturing.
• 300433.SHE: Lens Technology is an adjacent precision structural-component and consumer-electronics manufacturing peer.
• CRUS.US: Cirrus Logic provides a higher-IP audio-semiconductor valuation reference rather than a direct manufacturing comparable.
• KN.US: Knowles is an acoustic/sensing technology reference whose portfolio economics differ materially from GoerTek's consolidated model.
• META.US: Meta is an important end-device demand driver in smart glasses and VR; industry trackers show unusually high category share, increasing programme concentration for suppliers.

This report is based on public information and does not constitute investment advice. Markets carry risk; invest with caution.