UBS Initiates Coverage on China Fiber Optics: 2027 Overcapacity Risk Overestimated, YOFC H-Share Target Price Set at HK$330
nashnova research
UBS initiates coverage on China's fiber-optic sector, calling market fears of 2027 oversupply significantly overstated — demand has shifted from telcos to global data centers, and YOFC's H-share target is raised to HK$330, implying 73% upside.
The demand engine has changed — changed to what?
China's fiber sector used to run on capex budgets from three state telcos. UBS now sees the engine shifting to global data-center buildout.
UBS forecasts global fiber demand growing at a ~11% CAGR from 2026 to 2030, with data-center demand alone compounding at 37%.
This means → data centers will go from just over 10% of global fiber demand in 2025 to more than 45% by 2030. Fiber is no longer just "telecom infrastructure" — it is becoming the circulatory system of AI compute.
Why did the market panic, and why does UBS say it's wrong?
China's fiber sector has fallen 37% from its June high, triggered by a wave of capacity-expansion announcements from second-tier players and new entrants in Q2 2026.
The market drew a parallel to the glut at the end of the 5G cycle, pricing in a "2027 earnings peak."
In plain terms = the market copy-pasted last cycle's script onto this one. UBS argues the demand structure and supply constraints are fundamentally different — the analogy itself is flawed.
Can new capacity really arrive that fast?
UBS breaks the supply ramp into three waves: leading players complete brownfield expansion — building out on existing sites — in 1.5 to 2 years by late 2026 to early 2027; second-tier greenfield projects announced in Q2 2026 take 2+ years and may stretch to three, given longer equipment lead times and talent competition; overseas capacity lands after late 2028 at the earliest.
The core bottleneck is the preform — the glass rod from which fiber is drawn. Preform equipment must be imported from South Korea and the US, with lead times now exceeding one year. Even after delivery, new entrants face installation, process tuning, and customer qualification.
This means → UBS projects industry utilization stays above 90%, with tight supply lasting at least through 2027–2028. The oversupply timeline is far later than the market assumes.
Why is high-end fiber so hard to replicate?
AI network architecture is pushing demand away from standard G.652.D fiber toward higher-spec products: G.654.E ultra-low-loss fiber for long-haul data-center interconnects, G.657.A1/A2 bend-insensitive fiber for rack-to-rack links, and polarization-maintaining and hollow-core fibers for more advanced use cases.
These products require years of R&D, manufacturing know-how, and customer certification — barriers far above commodity telecom fiber.
In plain terms = new entrants will almost certainly start with standard fiber, but demand is moving in the opposite direction. High-end data-center fiber supply will remain structurally tight, and leading players with preform capacity and specialty portfolios keep widening their moat.
What does the YOFC H-share valuation gap reveal?
UBS raises its YOFC (长飞光纤, 6869.HK) H-share target to HK$330, versus a current price of HK$190.5 — 73% upside. That implies 14.8x / 8.6x 2026/2027 P/E, with a 45% EPS CAGR over 2026–2028. Rating: Buy.
YOFC's H-share trades at a 57% discount to its A-share. By comparison, dual-listed AI PCB peers sit at only a 23% discount, and optical/semiconductor interconnect names actually trade at an H-share premium.
This reflects offshore investors over-pricing fiber oversupply risk while undervaluing YOFC's strategic position in the global AI supply chain — and that mismatch is precisely the opportunity UBS sees.
When will this call be tested?
UBS flags Q4 2026 as the key verification window — the next round of telco fiber procurement results and the company's Q3 earnings will be disclosed around the same time.
This means → whether the supply-demand tightness holds and whether the leader premium materializes will face their most direct test in Q4.
YOFC's A-share keeps a Neutral rating, with the target raised from RMB 310 to RMB 510 — UBS's bullish thesis centers on H-share discount repair, not on absolute A-share valuation.
市场有风险，内容仅供研究参考，不构成投资建议。