Beijing Jingyi: A Genuine Semiconductor Subsystem Business at an Increasingly Demanding Price

Meta

• Ticker: 688652.SHG
• Company / project: Beijing Jingyi Automation Equipment Co., Ltd.
• Price & market cap: CNY196.01 per share; CNY32.93bn market cap as of 2026-06-29, calculated from the 2026-06-29 close and 168.0m shares outstanding. At USD/CNY 6.7940 on 2026-06-29, that is about USD4.85bn.
• Currency: CNY
• Report date: 2026-06-29
• Industry: Semiconductors
• One-line positioning: Domestic semiconductor subsystem maker focused on chillers, local scrubbers and wafer sorters; 2025 revenue reached CNY1.43bn.

Bottom Line Up Front

This report covers Beijing Jingyi Automation Equipment under a SUN-R lens as of 2026-06-29, with the horizon extended to 2028. The business is real semiconductor infrastructure, not a shell built on social heat. But the stock already trades like a scarce-policy beneficiary with several years of flawless execution pre-paid. That split runs through the whole case: a genuine business with an increasingly demanding share price.

On the “old-world-underpriced, new-world-migrating” question, the answer splits in two. The operating business still sits inside a structural domestic-substitution trend that many generalist investors underestimate: China remains the world’s largest semiconductor-equipment market, 300mm fab equipment spending is still rising into 2026 and 2027, and Jingyi has already cleared the hardest hurdle for a subsystem vendor by entering leading customers’ volume fabs. The equity, though, is not underpriced on today’s numbers. At the 2026-06-29 close the stock sat near 208x trailing twelve-month earnings and around 23x 2025 sales, and 2025 profit actually fell year on year despite 39% revenue growth, because R&D rose sharply and tax-rebate support fell. Usage-first company, price-first stock.

Any upside from here has to come mainly from real order conversion and continued domestic share gains in temperature control and process exhaust treatment, above all at memory and foundry customers. What can still move earnings is straightforward: more Chinese fab capacity, deeper local subsystem penetration, more aftermarket service, and better operating leverage once the current R&D build-out turns into qualified products. The biggest single opportunity isn’t wafer-sorter optionality. It is the core mix itself: chillers already account for roughly two-thirds of revenue, scrubbers roughly one-quarter, and both sit in an industry where customer validation takes time and then tends to persist.

The main rerating risk is just as plain. Contract liabilities climbed from about CNY406.9m at 2023 year-end to CNY736.2m in 2024, CNY1.35bn in 2025, and CNY1.49bn by 2026 Q1, which tells you demand is real. Inventory, though, reached CNY2.39bn in 2025 and CNY2.51bn in 2026 Q1, and the company itself warns that goods shipped but not yet accepted make up a very large share of it. Let customer capex pause, acceptance slip, or a new product cycle take longer to qualify, and the market will stop paying a scarcity multiple for revenue that has not yet fully become earnings or cash.

My settled rating is Thematic opportunity: the business sits in a durable structural shift with real deployment, but the share price already reflects a large part of it. In the project’s standard vocabulary, that maps to Watch.

S = Structural Shift

Jingyi represents a narrow but important shift inside Chinese semiconductor localization. The company is not chasing the whole wafer-fab-equipment stack; it goes after imported subsystems that sit close enough to process yield, uptime, and safety to make customers cautious, yet are not so far out of reach that a domestic entrant cannot qualify. Subsystem replacement tends to run ahead of full tool replacement, because fabs can localize chillers, process exhaust treatment, handling modules, and support equipment earlier than they can localize lithography or some leading-edge process chambers. Jingyi’s own disclosures list its core products as chillers, local scrubbers, and wafer sorters, and show it has already entered major customers including YMTC, SMIC, Hua Hong, Intel Dalian, CanSemi, and CXMT.

The old world has historically underpriced this kind of company. Part of it is reflex: many investors still rank subsystem vendors below the more visible front-end champions, even though fabs care intensely about temperature stability, exhaust safety, and tool uptime. The rest is that the business model looks messier than software or fabless semis. Revenue recognition runs through shipment, installation, debugging, and acceptance; inventory can swell before earnings do; customer concentration is baked into the market structure. So the reported numbers look most awkward exactly when underlying demand is accelerating. Jingyi’s 2025 annual report shows it plainly: revenue rose 38.95% to CNY1.426bn, yet reported net profit fell 3.26%, because R&D expense rose nearly 49% and other income from VAT rebates fell. Read only the near-term earnings and you miss that the company was also building product depth and backlog.

The migration is real, not theoretical. In the prospectus the company called itself the only domestic maker with scaled deployment of semiconductor-specific chillers and the only domestic scaled-installation vendor among scrubber suppliers, while showing that its key metrics in both categories were broadly comparable to the mainstream foreign products it benchmarked against. By 2022 its domestic scrubber share had risen to 15.57%, fourth in the market, and its chillers were already described as domestically leading. Foreign peers did not disappear. What changed is that fabs had stopped asking “can a domestic supplier work?” and started asking “where can a domestic supplier be qualified at scale?”

The migration also sits inside a capex regime that stays supportive through the current horizon. SEMI expects global 300mm fab equipment spending to rise 18% to USD133bn in 2026 and another 14% to USD151bn in 2027, with China, Taiwan, and Korea remaining the top destinations through 2026. SEAJ separately put China’s 2025 semiconductor-equipment billings in the top regional spot at USD49.3bn. Jingyi does not need the whole equipment market to keep compounding. It needs Chinese wafer-fab expansion and further localization of the subsystem layers where it already has proof points, and both the macro and the micro line up for that today.

Over a three- to five-year window, the plausible industry change is not “foreign out, domestic in” across the board. It is more selective. Imported leaders will probably keep the highest-value positions in many categories, while qualified local specialists should keep taking share where reliability, service responsiveness, integration, and political resilience count for more than absolute bleeding-edge process leadership. Jingyi’s product descriptions fit that pattern: its chillers are already qualified in logic and storage applications, its scrubbers adapt to process tools from Lam, TEL, Applied Materials, Kokusai, AMEC, NAURA and others, and its on-site service model is built around domestic fabs’ demand for fast response. That is classic structural migration in semiconductor equipment: not a change in physics, but a change in approved vendor lists.

U = Usage & Unit Economics

The usage case is tangible and measurable; this is not a pre-revenue concept stock. Revenue rose from CNY742.3m in 2023 to CNY1.026bn in 2024 and CNY1.426bn in 2025, and 2026 Q1 revenue reached CNY392.3m, up 16.13% year on year. Net profit climbed to CNY152.9m in 2024, dipped to CNY147.9m in 2025, then recovered to CNY46.4m in 2026 Q1, up 29.45% year on year. So the business clears the first usage test: customers are buying, installing, and paying for the equipment.

A compact view of the recent operating record is below.

Source: company annual reports and 2026 Q1 report; author formatting.

The numbers say demand is real, and also that the income statement alone won’t capture it. Contract liabilities almost tripled between 2023 and 2025, then kept climbing in 2026 Q1; with no disclosed backlog, that is the best public read on order growth. Inventory rose faster still, to CNY2.39bn at 2025 year-end and CNY2.51bn by 2026 Q1. Management warned that 2025 inventory was 55.98% of current assets and that shipped-but-not-yet-accepted goods were 64.31% of inventory book value. This is not fake prosperity of the wash-trading kind; it is real industrial demand running through a long acceptance cycle. That cuts both ways: it strengthens the case for underlying usage and weakens the case for near-term earnings visibility.

Revenue quality is mixed, and investors need to respect that. The good news: 2025 operating cash flow turned strongly positive at CNY381.4m after running roughly flat in 2024, which is hard to fake in a capital-equipment company. The weaker news: 2025 recurring earnings lagged the topline. Attributable net profit fell 3.26%, deducted net profit fell 0.47%, and non-recurring items still mattered, including CNY20.75m of recognized government subsidies and CNY18.14m of fair-value and disposal gains on financial assets. None of that breaks the thesis, but the clean operating gear has not yet shown up as strongly as the sales growth.

The product mix shows what the usage engine actually is. In 2025, chillers generated CNY937.0m of revenue at 36.37% gross margin; scrubbers CNY348.1m at 26.73%; wafer sorters only CNY35.6m at 5.00%; spare parts and support equipment CNY83.6m; service CNY21.4m. This is not “three equal legs.” It is one large leg, one meaningful second leg, and one small option: chillers are the thesis, scrubbers are the second engine, and sorters are still in incubation, with technical value but too little current revenue to carry valuation.

For unit economics, that mix has real consequences. Chillers and scrubbers carry industrial logic that can improve with scale: a bigger installed base supports better procurement, more field data, faster debugging, spare-parts pull-through, and eventually service revenue. This is not a software model where gross margins drift up on their own. Jingyi’s 2025 company-level gross margin was 32.61%, down 0.18 percentage points from 2024, as cost growth broadly matched sales growth. Chillers held up, scrubber margins slipped, and the sorter business stayed low-margin. Scale helps here, but it does not help automatically; the business still needs mix improvement and cleaner acceptance conversion before meaningful operating leverage shows.

Customer stickiness looks real. Jingyi’s 2025 report describes local service teams stationed near major customers and products that can respond quickly to customized needs, and it stresses that fabs run strict supplier-screening because equipment reliability, safety, and process fit feed straight into production. That is exactly where installed vendors gain persistence: once a subsystem is validated into a critical production flow, buyers do not switch casually to save a little money. The direct-sales model and long on-site presence only deepen the lock-in.

A commercially useful technology breakthrough by 2028 is a medium probability, not a high one. The R&D engine is real: 2025 R&D expense reached CNY140.0m, 9.81% of revenue, invention patents reached 130, and the lead in-process projects span higher-end temperature-control iterations, hydrogen and other harsh-gas treatment platforms, and new wafer-transfer products. One sorter project had already produced a sample machine that stores 800 wafers and passed end-customer validation; a high-flow combustible-gas treatment project had formed partial prototypes. Enough to support iterative product upgrades and broader process coverage, yes. Not enough to pencil in a category-defining step change on the order of a new process-module monopoly.

CXMT is the read-through investors ask about most. The record: Jingyi had a framework agreement with CXMT in place by June 2022, and prospectus disclosures showed CXMT as a significant customer in 2023 H1. The public expansion path is large. Reuters cited about 200,000 12-inch wafers per month in mid-2025, and later Reuters reporting described a path toward around 600,000 wafers per month as the Shanghai, Hefei, and Beijing expansions roll forward; SemiAnalysis coverage summarized by Barron’s has CXMT continuing to gain global share into 2028. What public materials won’t confirm is Jingyi’s exact present revenue share from CXMT after 2023 H1, or the exact subsystem content per 100k wafers of future CXMT capacity.

So my estimate has to stay an inference. If CXMT’s effective capacity path from the present cycle to 2028 adds roughly 250k to 300k wafers per month, and if Jingyi holds a meaningful but not dominant position in chillers and scrubbers there, CXMT expansion could plausibly add about CNY120m to CNY300m of annual revenue by 2028. At a 12% to 15% incremental net margin, that is about CNY15m to CNY45m of incremental annual net profit. The base case I use later in valuation sits near the middle, roughly CNY30m of annual net profit by 2028. Meaningful, but it is not large enough by itself to justify a CNY32.9bn market cap. To grow into today’s price Jingyi needs broad domestic-fab demand, not one memory champion. This estimate rests on the customer-history disclosure, the public CXMT capacity path, and Jingyi’s own margin profile.

N = Narrative Liquidity

Jingyi’s one-line story travels fast: China’s scarce domestic chiller-and-scrubber proxy for memory and foundry capex. It is a clean narrative because it bundles three things investors like, namely import substitution, visible downstream customers, and a product set that sounds closer to “must-have infrastructure” than to discretionary electronics. The company’s disclosures back the core of it. The risk is that the market compresses a nuanced operating case into something cruder: “every new Chinese fab benefits Jingyi.” That is directionally true, but not proportionally true.

Narrative liquidity is strong right now. On 2026-06-29 the stock closed at CNY196.01, up 15.06% on the day, on 11.87m shares against a three-month average of 6.36m. The 52-week range had stretched from CNY53.00 to CNY197.77, a one-year move of about 252%. These are not quiet-compounding numbers. The market has decided Jingyi is one of the handier listed expressions of the domestic-memory and semiconductor-localization trade.

The narrative rests on real usage, so it is not a bubble in the pure sense. Revenue grew sharply in 2024 and 2025, contract liabilities kept climbing, and the company already sells to major fabs. The trouble is not that the story is fake; it is that price has run far ahead of demonstrated earnings power. On trailing twelve-month profit including 2026 Q1, the stock was at roughly 208x earnings at the 2026-06-29 close. At that multiple even small disappointments in acceptance timing, customer capex, or margin conversion hit hard.

Reflexivity is central here. On the way up, a higher stock price lifts brand visibility, helps recruiting, reassures some customers about supplier durability, and nudges the market to judge Jingyi less by present profits and more by a future localization roadmap. On the way down the same mechanism reverses: a derating instantly changes what investors look at. They drop contract liabilities and fix on inventory, fourth-quarter profit softness, customer concentration, and the missing disclosed backlog. The narrative turned dangerous once price outran the conversion of orders into earnings, and I think Jingyi is now much closer to that dangerous zone than to a neglected-asset zone.

That is why I split the business narrative from the stock narrative. The business side is healthy: real installed products, hard customer validation, actual capex exposure. The stock side is stretched, with the market already discounting a long runway of customer-order growth and technology progress, even though the reported 2025 earnings record came in weaker than the share-price trend implies. The business can keep winning while the stock still derates.

Bridge: Bridging the Traditional World

Jingyi bridges into the traditional world better than many high-growth A-share equipment stories, because it lives inside existing fab workflows instead of asking customers to adopt a new industrial architecture. A chiller that holds process temperature steady, or a local scrubber that safely treats exhaust, is not a speculative line in an experimental budget; it is part of the operating stack of a working plant. The customer list runs to leading Chinese fabs and Intel Dalian, and the scrubbers are built to pair with major domestic and foreign process tools. That makes for a stronger bridge than a pure domestic-policy story: Jingyi is not merely approved by policy, it is usable inside existing global fab ecosystems.

The bridge is institutional too, in capital-market terms. Jingyi is STAR-listed, with audited reports, state-linked industrial lineage, a disclosed dividend policy, and a customer base of large fabs rather than speculative channels. Within the SUN-R framework that widens the buyer universe, and a domestic-substitution story institutions can hold is more durable than a theme confined to retail chatrooms. The bridge still cuts both ways: institutional ownership and benchmark visibility add liquidity on the way up, but they also make crowded positioning quicker to unwind once the multiple becomes the whole story.

Commercially, the strongest bridge is service and integration. Jingyi’s reports point to local service teams stationed near major customers and quick turnarounds on customized requirements. Unglamorous features, but in semiconductor equipment they count almost as much as headline spec sheets. A domestic vendor that can show up faster, debug faster, and tweak a subsystem faster holds a real edge when a fab is trying to cut its dependence on imported maintenance chains, all the more so when export controls and geopolitical friction turn “local support certainty” into a product feature in its own right.

The move is commercial, not just public relations. Customers had signed framework agreements with YMTC and CXMT before listing, and the prospectus listed large historical contracts and framework arrangements with major semiconductor players. None of that proves future share gains line by line, but it does prove the company crossed from “pilot discussion” into “purchasing system.” In capital equipment, that is the bridge that counts.

Anti-cycle: Counter-cyclical and Distressed Opportunities

The mid-2026 industry backdrop is not a trough. SEMI’s latest outlook still points to another step up in 300mm fab equipment spending in 2026 and 2027, China remains a major destination for equipment demand, and memory, foundry localization, and AI-related infrastructure are all pulling the same way. So I am not analyzing Jingyi from the bottom of a cycle where everything is hated and cheap; this is an expansion phase, and the scarcity premia are already back.

That weakens the anti-cycle argument. Jingyi is not a natural distressed-asset buyer. At 2025 year-end it held about CNY554.3m of cash and equivalents, CNY942.9m of trading financial assets, just CNY50m of short-term debt, and no long-term borrowings; by 2026 Q1 short-term borrowings had fallen to zero. A healthy balance sheet, with room to keep building and funding R&D. But set against the global equipment leaders it benchmarks against, Jingyi is far too small to drive consolidation in a downturn. Its anti-cycle edge is survival and share-taking, not acquisition.

There is still a useful counter-cyclical angle. If the fab-capex cycle cools in 2027 or 2028, large foreign vendors may defend share through price and service bundling, while cash-poor domestic challengers retreat. Jingyi’s net cash, growing installed base, and the ongoing Anhui production-and-R&D build-out improve its odds of staying on customer roadmaps through a softer stretch. A downturn could even widen the moat at the margin, since qualification persistence counts for more when customers turn selective. That is a second-order benefit, though, not a reason to pay any price today.

So the right anti-cycle read is modest. If the sector falls 50%, Jingyi is more likely to be a relative winner than a system-level consolidator. The balance sheet should let it keep investing, yet the stock would almost certainly trade as a high-beta victim first, because the current multiple leaves very little protection.

R = Regulatory, Reflexive and Ruin Risk

The most important risks are summarized below.

Sources: company risk disclosures, balance sheet and operating data, current trading statistics, and public reporting on Chinese fab expansion.

Customer concentration is the first risk to read as structural, not incidental. Jingyi itself says concentration is high and that a drop in major-customer capex would hit order volume and earnings stability. That is normal for semiconductor equipment, but it bites harder here because the market prices Jingyi like a broad-based platform rather than a concentrated capital-goods supplier. If memory or foundry expansion pauses, this stock gets no benefit of the doubt.

Inventory and acceptance are the second major risk. The company is explicit that large amounts of shipped goods sit unaccepted for a while, and its 2025 figures show it clearly. That does not imply channel stuffing. It does imply that investors have to watch the pace of acceptance, not just the pace of shipments and prepayments. A long equipment lead time is manageable in a steady capex cycle; it turns dangerous when the market prices the stock as if every shipment converts smoothly and on time.

Technology-step execution is more nuanced. The R&D breadth is genuine: Jingyi is putting money into next-generation temperature-control systems, harsh-gas exhaust handling, and more advanced wafer-transfer equipment. But these are customer-qualification businesses, not app releases. A product can work in the lab and still take far longer than investors expect to reach meaningful revenue, which is why I marked the technology-breakthrough likelihood as medium in the opportunity section and the execution risk as medium here. The payoff is real; so is the timing risk.

Policy and sanctions risk cuts both ways. The broad localization push helps Jingyi, yet sanctions and export-control regimes can also delay customer build-outs, change tool configurations, or squeeze some upstream components and service chains. Jingyi’s earlier reporting warned that tighter overseas restrictions could weigh on domestic semiconductor development. The more likely outcome for the company is a stop-start ordering cycle that worsens quarterly visibility, not an outright existential ban.

Governance risk is not the headline issue, but it is not zero either. The 2024 annual report disclosed that one independent director could not be contacted; the 2025 report shows that director left and a replacement was elected. I read that as a resolved blemish, not a live thesis-breaker. Still, when a stock is priced this aggressively, even small governance oddities register, because investors are paying for trust as much as for growth. Some senior executives also trimmed holdings in 2025; the amounts were not thesis-defining, but they are worth watching if the market keeps treating the stock as a pure scarcity asset.

The reflexive risk is the one I would weight most heavily in an actual portfolio decision. At the 2026-06-29 close, on trailing twelve-month profit, the stock was around 208x earnings, which means the market is capitalizing several years of future success in advance. A plain rerating to still-generous growth-equipment multiples can gut the stock even if Jingyi keeps growing revenue and winning customers. That is why the downside here is more likely to start as valuation compression than as business failure.

Ruin risk in the literal “go to zero” sense is lower than the quote’s behavior might suggest. The company has real products, real customers, net cash, and audited profits, so a total business-model collapse is not my base case. A 50% to 70% equity drawdown, though, is easy to picture if three things land together: customer capex slows, inventory conversion disappoints, and the market stops paying scarcity multiples for domestic-equipment proxies. For the shareholder, that is the risk that counts. The stock does not need the business to break to inflict permanent capital damage.

SUN-R score

I score Jingyi as follows.

• Structural shift: 13/15. Domestic substitution in subsystems is real, customer migration has already happened, and the broader fab-capex backdrop remains supportive into the stated horizon.
• Real usage: 18/20. Revenue, cash receipts, contract liabilities, and customer references all show the company is already used in production environments.
• Unit economics: 10/15. Gross margins are respectable and some scale benefits exist, but 2025 showed that revenue growth does not yet translate cleanly into profit growth, and non-recurring items still matter.
• Network effects: 7/15. This is not a digital network, but qualification lock-in, service density, and customer process knowledge do create switching costs and local data advantages.
• Narrative liquidity: 11/15. The story is simple and marketable, and recent trading shows clear thematic sponsorship, but the narrative is no longer under-owned.
• Bridge to the traditional world: 8/10. The company serves mainstream fabs, integrates with global and domestic process gear, and is investable for institutional capital.
• Counter-cyclical opportunity: 5/10. The balance sheet is good enough to endure a downturn and perhaps take share, but the company is too small to become a genuine distressed-asset buyer.
• Regulatory & ruin risk: -16/40. The business is unlikely to disappear, but customer concentration, inventory-heavy conversion, and extreme valuation make the equity unusually fragile.

That sums to a total of 56, which places Jingyi in Thematic opportunity. The score reflects the split already laid out: a real infrastructure winner inside a structural shift, paired with a stock price that already discounts a large amount of future success and leaves little room for timing misses.

【Project Standard Rating】Watch

Position Sizing and Tracking

For a balanced-risk investor, I would classify Jingyi as watch-only at the current price. Anyone who insists on exposure to the theme should hold it as a small research position, not a core holding. This is not a “buy and forget” setup; it is “track the operating conversion carefully, and wait for either earnings catch-up or a materially lower entry.” The aim is not to deny the business. It is to refuse to pay for an immaculate future while the company is still converting backlog through a long acceptance cycle.

What would push me to do more work on the bull case over the next three months is not another sharp price move; it is evidence that the order story is turning into cleaner earnings quality. The signals I care about: contract liabilities staying firm without inventory quality slipping, chiller margins stabilizing or improving, scrubber mix improving, sorter optionality starting to matter, and 2026 interim results showing that the fourth-quarter 2025 profit slump was timing rather than deterioration. If revenue grows while deducted profit lifts too, the case changes.

What would overturn the thesis is just as clear. If customer prepayments flatten, inventory keeps outrunning sales, or new products eat R&D without widening the revenue base, the stock loses both halves of the argument at once: the growth story slows and the premium multiple becomes indefensible. The company itself flags customer concentration, receivables risk, inventory write-down risk, and competition risk. Those are not boilerplate here; they map exactly onto the public balance-sheet stresses the market should be watching.

Over the next three months, I would track interim order proxies and sentiment: contract liabilities, inventory composition, receivables movement, any read on CXMT, YMTC, SMIC, and Hua Hong capex cadence, and whether daily turnover stays far above the three-month average. Over the next twelve months, whether 2026 full-year profit regains a clear upward slope and whether R&D productivity shows up as broader process coverage or new qualified products. Over three years, whether Jingyi becomes a two-engine company with chillers and scrubbers both compounding at scale, or stays a single-engine chiller story with expensive optionality bolted on top.

Share-price elasticity is unusually high because the multiple is unusually high. Hold the market multiple constant, and every additional CNY10m of annual net profit is worth about CNY0.0595 of EPS, or roughly CNY12.4 per share at the current trailing multiple near 208x. Switch to a more reasonable fair-value frame using a 2028 growth multiple in the 50x to 60x range and discounting back to today, and that same CNY10m of annual net profit is worth only about CNY2.5 to CNY3.1 per share. That gap is the whole risk: the present quote responds to short-term earnings revisions far more violently than a sane long-term valuation would. Run my CXMT estimate through the same logic and a CNY30m annual net-profit contribution by 2028 is worth about CNY8 to CNY9 per share in a fair-value frame, not enough on its own to justify today’s premium. This valuation logic is my calculation based on the current quote, share count, and the scenario multiples used below.

I set the ideal buy range at CNY80 to CNY100. That roughly matches the discounted present value of a credible 2028 base case in which Jingyi keeps gaining share but the market stops paying a triple-digit P/E for that growth. At those levels you would be paying for the structural shift and still getting paid for the execution risk. Above CNY140, most of the reasonable 2028 upside is already in the price.

【Valuation Range】

• current: 196.01 (as of 2026-06-29)
• bear (conservative · ideal buy zone): [80, 100]
• base (fair · acceptable hold zone): [105, 135]
• bull (optimistic · fully-priced zone): [145, 170]
• mode: price