Luxshare Precision prévoit une hausse de son bénéfice net de 15 à 25 % au cours des neuf premiers mois de 2026

Luxshare Precision prévoit pour les neuf mois clos le 30 septembre 2026 un résultat net attribuable aux actionnaires entre 13,25 Md et 14,40 Md RMB, soit +15% à +25% en glissement annuel, et un résultat net ajusté de 10,37 Md à 11,72 Md RMB (+8,63% à +22,79%). Le BPA de base devrait atteindre 1,79–1,95 RMB (contre 1,59 RMB l’année précédente), soutenu par l’accélération des segments électronique grand public (+19,3% à 122,48 Md RMB) et communications/centres de données (+49,7% à 16,61 Md RMB), ainsi que par la forte hausse de l’électronique automobile (+274,1% à 32,39 Md RMB). La marge brute s’améliore de +17 pb à 11,78%, malgré des capex en hausse à 10,27 Md RMB et des R&D en hausse de +43,1% à 6,57 Md RMB, signalant un cycle d’investissement orienté commercialisation.

Analysis

The market implication is not the headline growth rate; it is the evidence that the company is successfully moving mix toward higher-value systems work where pricing power is better and customer switching costs are higher. That should widen the gap versus consumer-electronics assemblers that remain stuck in lower-margin, design-led exposure, and it also raises competitive pressure on smaller China-based peers that lack integrated thermal, optical, and power capabilities. Over 6-18 months, the key re-rating case is less about unit growth and more about whether the business can sustain a structurally higher margin profile as AI infrastructure and vehicle electronics become a larger share of revenue.

Near term, the bigger issue is quality of earnings and cash conversion. Rising capex, elevated R&D, and inventory buildup are consistent with an inflection phase, but they also mean free cash flow can lag reported profit for several quarters if customer pull-through slows or program timing slips. That creates a useful falsifier: if the next report shows gross margin flattening or working capital absorbing more cash than revenue growth justifies, the market should take away the "platform upgrade" narrative and re-price this as cyclical execution rather than durable structural improvement.

The contrarian miss is that consensus may be extrapolating every growth bucket as equally sticky. AI/datacenter content is likely the highest-quality leg, while consumer hardware can still revert quickly and auto ramps can be lumpy by SOP timing. The strongest secondary beneficiaries are upstream thermal, optical, and connector suppliers tied to high-speed computing; the main losers are commoditized EMS names that do not have a route to system-level ownership. The immediate move may be modest if this is already partially in the tape, but the 1-3 month catalyst path is the third-quarter print and any evidence that margins and cash flow are compounding faster than revenue.

Key Decisions for Investors

• Long 002475.SZ / 02475.HK on post-announcement weakness, targeting a 6-12 month mix-driven re-rating; exit if Q3 gross margin slips back below roughly 11.5% or working capital absorbs disproportionate cash.
• Pair trade: long 002475.SZ, short a more consumer-handset-exposed EMS peer such as 002241.SZ (Goertek) over the next 1-3 months; the thesis is mix-quality divergence, not absolute sector beta.
• Use any sharp rally to sell strength rather than chase; the setup is better on confirmation in the Q3 filing than on the pre-announcement alone, because capex and inventory make free-cash-flow risk the key watch item.
• Watch the AI infrastructure cadence as the main catalyst: if high-speed interconnect, liquid cooling, and power modules keep compounding in the next update, the equity can sustain a higher multiple for 6-18 months; if not, expect multiple compression back toward a traditional hardware assembler.
• For credit-focused accounts, prefer the bonds over equity only if leverage remains controlled and operating cash conversion holds; if not, treat the name as an equity story and avoid reaching for spread compression too early.