JCET Group
Back-end IC manufacturing services -- contract assembly, advanced packaging (flip-chip, fan-out WLP, 2.5D/3D, SiP, chiplet integration) and final test billed per-unit/per-wafer to fabless designers, IDMs and foundries. Capital-intensive, capacity-utilization-driven margins.
Earnings, margins, COGS & capex
FY2025 delivered record revenue of RMB 38.87B (+8.1%) and net profit attributable to shareholders of RMB 1.57B (net margin ~4.0%), with pre-tax profit RMB 1.74B (+5.4%). The story is mix-shift: advanced-packaging revenue hit RMB 27B (~70% of sales), and computing, industrial/medical and automotive electronics segments grew 42.6%, 40.6% and 31.7% YoY respectively -- offsetting softer legacy consumer/communications packaging. Margins remain structurally thin (OSAT is a low-margin, high-capex business) but are being defended through advanced-node mix while R&D rose 21.4% to RMB 2.09B.
Income statement — where each revenue dollar goes
% of revenueOf every $1 of revenue, ~86¢ is cost of goods and ~9¢ operating expense, leaving ~5¢ of operating profit (~4¢ net).
Revenue trend
Margins
roughly stable; structurally low for OSAT
modest
up on cost control + mix
+26.6% QoQ
COGS structure
Dominated by direct materials (substrates, leadframes, bonding wire, wafers received from customers on consignment vs turnkey), depreciation of a very large assembly/test equipment base, labor, and utilities. Depreciation is the key operating-leverage lever -- margins swing sharply with utilization on the fixed-asset base.
Capex
Elevated and sustained -- funding high-end advanced-packaging lines (fan-out, 2.5D/3D, chiplet, automotive-grade), including a dedicated automotive-chip advanced-packaging facility. Capex has exceeded operating cash flow in build-out quarters, pressuring free cash flow.
Latest earnings
Record full-year revenue and higher pre-tax profit; framed as in-line-to-positive by management. No consensus-vs-actual beat/miss cleanly disclosed for an A-share OSAT.
No formal numeric guidance typical of A-share OSATs; management guides qualitatively toward continued advanced-packaging and automotive capacity ramp
- FY2025 revenue
- RMB 38.87B (+8.1%)
- Advanced-packaging revenue
- RMB 27B
- FY2025 R&D
- RMB 2.09B (+21.4%)
- Net profit attributable
- RMB 1.57B
Growth drivers
- Advanced packaging (~70% of revenue, RMB 27B) — fan-out, 2.5D/3D, chiplet integration riding the AI/HPC packaging cycle
- Computing/AI electronics segment (+42.6% YoY) as chiplet and heterogeneous-integration demand scales
- Automotive electronics (+31.7% YoY) -- new automotive-grade packaging capacity
- China domestic-substitution: local fabless designers, SMIC and HiSilicon localizing back-end away from Taiwan
- Industrial/medical electronics (+40.6% YoY)
Bull & bear
The indispensable advanced-packaging arm of China's semiconductor build-out, riding both the global AI/chiplet packaging cycle and a state-mandated onshoring wave, with real top-3 scale and a full advanced-packaging process menu.
- Advanced packaging is now ~70% of revenue (RMB 27B) and among the fastest-growing parts of the semiconductor value chain as node-scaling economics fade -- JCET is a top-3 global player in exactly the right pocket
- Computing/AI (+42.6%), industrial/medical (+40.6%) and automotive (+31.7%) segments are compounding at 30-40%+ YoY, offsetting a maturing consumer base and pointing to durable mix-driven growth
- Structural tailwind of China domestic substitution: SMIC, HiSilicon and local fabless firms have every incentive (and policy backing) to localize back-end onto JCET
- State-aligned capital access (National IC Fund history, SMIC alignment) de-risks funding for the capex a capital-hungry OSAT needs
- Record FY2025 revenue plus +26.6% QoQ Q4 profit growth suggest operating leverage improving as advanced-node utilization rises
A thin-margin, hyper-capital-intensive assembler whose ~4% net margin and cash-consuming capex leave little cushion, structurally behind TSMC/ASE at the true leading edge and squarely in the geopolitical crosshairs -- at a rich ~100x earnings multiple.
- Net margin ~4% and gross ~14% mean the business earns modest returns on an enormous, depreciating asset base -- one downturn in utilization can erase profitability
- Capex persistently exceeds free cash flow; the growth requires continuous reinvestment that may be slow to convert to owner cash returns
- The most valuable AI packaging (CoWoS-class 2.5D for GPUs) is being insourced by TSMC and led by ASE/Amkor -- JCET risks winning the lower-value tiers
- Export-control exposure cuts both ways: restricted access to Western advanced-packaging tools/materials could cap its leading-edge roadmap
- As a China national champion, it is exposed to further US/allied restrictions and to some Western customers de-risking supply chains away from it
- Valuation: at ~$23.6B market cap on ~$218M FY2025 net income, the stock trades near ~103x trailing earnings (P/E ~103 per stockanalysis.com) -- the AI-packaging and onshoring narrative is largely priced in
What it is worth
Market-cap cross-check vs OSAT peers on P/E and EV/sales; the business is capital-intensive and cyclical, so peer multiples plus mix-shift optionality frame the range.
A semiconductor downturn or export-control shock cuts utilization; ~4% margins compress and the ~100x multiple de-rates sharply toward peer levels.
Mid-to-high single-digit revenue growth with slowly improving margins; the rich multiple normalizes over time toward a still-premium level vs global OSAT peers.
Advanced-packaging mix drives margin expansion and 20%+ earnings growth; onshoring keeps utilization high -- premium multiple sustained or extended.
At ~$23.6B market cap (RMB ~170.2B) on ~RMB 1.57B (~$218M) FY2025 net income, JCET trades near ~103x trailing earnings (P/E ~103 per stockanalysis.com, 2026-07-07) -- a large premium to Western OSAT peers (Amkor and ASE trade at far lower multiples), reflecting an AI-packaging + China-onshoring + state-strategic-asset premium. On market-cap/sales (~4.4x on ~$5.4B revenue) it is also rich for a ~14% gross-margin assembler. The multiple embeds a bet that advanced-packaging mix lifts margins structurally; if utilization or the AI cycle disappoints, it can compress hard. Context only, not financial advice; a mainland-China A-share listing is not a buy/own recommendation for a US-first mandate.
SWOT
Strengths
- World's #3 OSAT (~12% share, ~$5.0B FY2024 revenue per TrendForce) and China's #1 -- scale, breadth and a global footprint (China, South Korea/Singapore via ex-STATS ChipPAC, US)
- Full advanced-packaging portfolio: flip-chip, fan-out WLP, 2.5D/3D, SiP, chiplet integration
- Strategic anchor of China's domestic semiconductor supply chain, historically backed by the National IC Fund ('Big Fund') and aligned with SMIC
- STATS ChipPAC acquisition (2015) added global blue-chip customer relationships and Singapore/Korea capacity
Weaknesses
- Structurally thin margins (net ~4%, gross ~14%) -- a commodity-leaning, capacity-utilization-driven model
- Very capital-intensive; free cash flow repeatedly consumed by capex during build-out
- Lags ASE/Amkor and especially TSMC in the highest-end 2.5D/CoWoS-class packaging for leading-edge AI accelerators
- Exposure to Chinese-designer demand ties fortunes to China's constrained access to leading-edge foundry capacity
Opportunities
- AI/HPC packaging cycle — chiplets and heterogeneous integration become the performance lever as node scaling slows
- China domestic-substitution mandate: onshoring back-end from Taiwanese/Korean OSATs
- Automotive-grade and industrial packaging as semiconductor content-per-vehicle rises
- Advanced packaging as the 'more than Moore' workaround where China can compete despite leading-edge equipment restrictions
Threats
- US/allied export controls could restrict advanced-packaging equipment and materials JCET depends on
- ASE and Amkor out-invest at the leading edge; TSMC insources the most valuable AI packaging (CoWoS)
- OSAT is cyclical -- a semiconductor downturn compresses utilization and the thin margins fast
- Geopolitics: as a China national-champion supplier, some US/Western customers may de-risk supply away from it
Moats, dependencies & bottlenecks
Moats
Top-3 global OSAT; assembly/test capacity and a global footprint are hard and costly to replicate
Full fan-out/2.5D/3D/chiplet menu; leads China but trails TSMC/ASE at the very top of leading-edge 2.5D/3D
Package qualification is sticky and long, especially automotive-grade; but multi-sourcing across OSATs is normal
Strong (in China) High while policy holds National IC Fund history / SMIC alignment; a designated advanced-packaging pillar of China's chip stack
China cost base helps, but OSAT is fundamentally commoditized -- price competition is real
Dependencies
Upstream wafer supply / demand pull Growth is coupled to Chinese fabless/foundry output; SMIC is both an aligned partner and a demand driver
Besi, Disco, Ibiden/Shinko-type substrates) Capital equipment / materials Leading-edge packaging tools and advanced substrates are import-dependent and exposed to export controls
Qualcomm, STMicroelectronics, memory makers) Customer concentration Specific concentration not disclosed; blue-chip base inherited via STATS ChipPAC plus growing Chinese designers
Capital & strategic Cheaper capital and demand mandates are a tailwind that could shift with policy or fiscal constraints
Utilization-driven margins amplify the cycle
Advantages
- China's only top-tier, globally-scaled OSAT -- default onshoring beneficiary
- Full advanced-packaging portfolio spanning fan-out, 2.5D/3D, SiP and chiplet
- Global footprint and blue-chip relationships from the STATS ChipPAC acquisition
- State-aligned capital access for a capital-hungry business
- Diversified end-market mix now tilting toward AI/compute, automotive and industrial
Weaknesses
- Structurally thin margins and heavy capital intensity
- Free cash flow repeatedly consumed by build-out
- Behind the leading edge for the most valuable AI packaging
- Rich ~100x earnings multiple leaves little room for cyclical disappointment
- Geopolitical/export-control overhang as a China national champion
Bottlenecks
- Access to leading-edge advanced-packaging equipment and high-end substrates under tightening export controls
- Capex funding cadence -- keeping advanced lines full to earn on the depreciation
- Talent/process yield at the 2.5D/3D and chiplet frontier vs TSMC/ASE
- Dependence on Chinese leading-edge wafer availability (SMIC node constraints) to feed the highest-value packaging
Top signals & trends
Top signals
Mix shift into the highest-growth part of the value chain
Direct read on chiplet/heterogeneous-integration demand
Investing to defend advanced-packaging position
Cash returns deferred; execution/utilization risk
AI-packaging + onshoring narrative largely priced in
Sell-side enthusiasm; sentiment risk if the cycle turns
Trends
Advanced packaging is the new performance lever as node scaling slows -- structural OSAT tailwind
State-mandated localization of back-end favors JCET
Threatens tool/material access and some Western customer relationships
Caps JCET's access to the most valuable AI packaging tier
New automotive-grade advanced-packaging capacity aimed here
Ecosystem & competitor graph
Suppliers feed the company; customers pull from it. Line thickness shows the strength of each tie (supply-chain dependency, customer earnings contribution). Hover to isolate a tie.
Assembly/packaging equipment (bonders, etc.)
Advanced-packaging die-attach/hybrid-bonding tools
Packaging deposition/etch and advanced-packaging equipment
Dicing/grinding tools for wafer-level packaging
High-end IC substrates (import-dependent bottleneck)
High-end IC substrates / package substrates
Chinese equipment supplier reducing foreign-tool dependence
Anchor Chinese fabless demand for chiplet/advanced packaging
Foundry partner and demand driver; strategically aligned within China's IC ecosystem
Blue-chip customer relationship (partly via ex-STATS ChipPAC)
IDM customer for packaging/test
Memory packaging demand served via Korea/Singapore lines
Taiwan; world's #1 OSAT (~$18.5B FY2024 rev, ~44.6% top-10 share per TrendForce). Scale and advanced-packaging leader.
US-headquartered #2 OSAT (~$6.3B FY2024 rev, ~15.2% share); the Western advanced-packaging alternative for US/EU customers.
Not an OSAT but insources the most valuable leading-edge AI packaging (CoWoS/SoIC), removing the top tier from the merchant market.
China #2 OSAT; AMD's key China packaging partner -- domestic rival for onshoring share.
Memory-packaging specialist; a top-5 global OSAT.
Test-focused OSAT; competes on the final-test side.
Taiwan; display-driver and memory test/assembly.
China OSAT growing double-digits; domestic competitor for substitution share.