China optical communications infrastructure and potential FCC restrictions on optical transceivers Report Interpretation

Overview

This Citi note examines whether potential FCC action could restrict China-made optical transceivers and how different regulatory designs would affect Chinese optics companies. Citi distinguishes enacted FCC rules from a separate reported draft proposal and concludes that near-term enforcement is unlikely, although the downside would be material for directly exporting module makers.

Core views

Citi separates the enacted FCC framework from reports of a possible optical-module import ban. FCC 26-50, adopted on 22 July 2026, formalizes two Covered List mechanisms: a producer/provider-based approach that blocks new FCC equipment authorizations for named entities and their affiliates, and a production-location-based approach that restricts product categories made outside the United States. The latter currently applies to unmanned aircraft systems, routers, power inverters and advanced robotic devices. Optical transceivers are not an enacted restricted category; their appearance in FCC 26-50 is only an illustrative reference in proposed hardware/software bill-of-materials disclosure discussion. Citi says the Reuters report dated 4 August concerns a separate proposal-stage possibility rather than a formal FCC rule. Citi argues that a producer/provider-based restriction targeting named Chinese optics companies would be difficult to implement near term. Under the FCC’s broad proposed “produced by” definition, offshore subsidiaries controlled through design, development, manufacturing or assembly could still be captured, so shifting only final assembly to Thailand, Vietnam or Taiwan would not provide meaningful protection. Eoptolink and Dongshan Precision (DSBJ) would face high exposure because of their direct US optical-transceiver exports, while non-Chinese competitors in Japan, Korea and Europe would benefit. Citi nevertheless considers this route unlikely because Chinese suppliers provide an estimated 60–70% of high-speed modules to major US cloud service providers and domestic alternatives cannot fill the supply gap promptly. A production-location-based restriction is, in Citi’s view, the more plausible template if action materializes, because it follows existing treatment of UAS, inverters and robots. An all-foreign-production rule would restrict modules made in China, Thailand and Vietnam alike unless they met the US domestic-content standard of at least 65% US-origin content, rising to 75% from 2029. This would leave Eoptolink’s and DSBJ’s offshore capacity without regulatory protection, severely constrain US AI-infrastructure buildout, and likely create pressure for broad exemptions. It would also require more automated domestic production lines, benefiting equipment and automation suppliers. A China-only location restriction would provide more potential relief through offshore production, but Citi stresses that the definition of foreign versus China production remains unresolved. FCC 26-50 explicitly does not use a simple last-substantial-transformation test; design, firmware and supply-chain control are all relevant. Under this scenario, Eoptolink and DSBJ’s exposure would moderate as overseas capacity ramps, whereas TFC would remain the most insulated because it does not directly manufacture optical transceivers and would be affected mainly through weaker downstream customer orders. Citi’s scenario table classifies Eoptolink and DSBJ as high-impact under both named-entity and all-foreign-location scenarios, falling to moderate under a China-only restriction. Eoptolink’s overseas capacity is 88.4%, concentrated in Thailand; DSBJ’s overseas capacity is 30.3% in Taiwan and Thailand and is being expanded in Thailand during 2026. TFC has 0.3% overseas designed production capacity, but its impact is moderate at worst under the first two scenarios and low under a China-only scenario because of its indirect exposure. Citi assigns low probability to near-term enforcement because a genuine ban would impair the US AI buildout that the administration publicly supports, the FCC retains discretion to modify or shelve restrictions, and supply alternatives require time to ramp. The report identifies September and November 2026 US-China diplomatic events as potential points at which optical-module restrictions could be raised in negotiations involving rare earths, agricultural purchases or other bilateral issues. The report also restates valuation bases for selected covered companies. Citi values DSBJ at Rmb350 through SOTP, using 15x 2026E P/E for the original business, 20x 2027E for optical transceivers, 50x 2027E for optical chips and 25x 2027E for AI-PCB. Eoptolink’s Rmb701 target is based on 20.0x FY2027E P/E, set at 0.5 standard deviations below its five-year historical mean, reflecting both the 800G/1.6T, ASIC and scale-up opportunity and CPO cannibalization risk at Nvidia. TFC’s Rmb419 target uses 34.3x 2027E P/E, its five-year mean, balancing a new CPO re-rating against potential 1.6T/800G cannibalization in 2027. ZTE’s HK$25.40 target uses 15.0x 2026E P/E, 1.5 standard deviations above its five-year average, with AI servers, computing products and consumer businesses expected to offset carrier-network weakness.

Analysis framework

Citi first distinguishes enacted FCC rules from a reported draft proposal, then applies the FCC’s two existing Covered List mechanisms to three hypothetical optical-module scenarios. It compares each scenario with suppliers’ overseas capacity and role in the optical supply chain, and separately presents company valuation frameworks and stated downside risks.

Methodology notes

• Event-Driven and Behavioral FinanceEvent-driven analysis

  Regulatory scenario analysis based on FCC Covered List mechanisms

  Citi assesses how different possible FCC actions could affect companies and the US optical supply chain, while emphasizing that no formal optical-transceiver rule exists.

• Industry AnalysisUpstream-Midstream-Downstream Transmission

  Optical-module supply-chain transmission

  The note links restrictions on module suppliers to US AI data-center deployment, downstream customer demand and indirect effects on component suppliers such as TFC.

• Valuation methodsSOTP (Sum-of-the-Parts) Valuation

  DSBJ sum-of-the-parts valuation

  Citi applies separate P/E multiples to DSBJ’s original business, optical transceivers, optical chips and AI-PCB operations to derive its target price.

• Valuation methodsP/E and PEG Valuation

  Forward P/E valuation

  Citi bases target prices for Eoptolink, TFC and ZTE on stated forward P/E multiples relative to historical or peer valuation references.

Asset mapping & comparison

Structured mapping from thesis to named assets (strengths, weaknesses, peers, risks).

• Eoptolink Technology (300502.SZ)

  Direct US optical-transceiver exporter with high regulatory exposure under named-entity and all-foreign-production restrictions.

  Strengths

  88.4% overseas capacity in Thailand; exposure moderates under a China-only restriction.

  Weaknesses

  Offshore capacity would not protect it under an all-foreign-production rule.

  Comparison

  High exposure, alongside DSBJ; materially greater direct exposure than TFC.

  Risks

  Slower data-center investment, weaker optical-network capex, price competition, slower customer expansion, China-US tech disputes and faster CPO deployment.

• Dongshan Precision (002384.SZ)

  Direct US optical-transceiver exporter with high exposure under named-entity and all-foreign-production restrictions.

  Strengths

  Thailand capacity is being expanded in 2026 and may moderate impact under a China-only restriction.

  Weaknesses

  Only 30.3% overseas optical-transceiver capacity; offshore production would not protect it in an all-foreign scenario.

  Comparison

  High exposure, alongside Eoptolink; more directly exposed than TFC.

  Risks

  Slower module FPC progress and overseas share gains, weaker Tesla business growth, optoelectronics losses, higher material costs and US-China geopolitical risk.

• Suzhou TFC Optical Communication (300394.SZ)

  Passive-component supplier with mainly indirect exposure through downstream optical-module customer demand.

  Strengths

  Most insulated across Citi’s scenarios; low impact under a China-only restriction.

  Weaknesses

  Does not directly benefit from overseas module-production capacity because its exposure is indirect.

  Comparison

  Moderate impact at worst versus high exposure for Eoptolink and DSBJ in the strictest scenarios.

  Risks

  Slower product development, weaker global AI and data-center investment, China-US disputes, upstream component constraints, competition and delayed CPO deployment.

• ZTE (0763.HK)

  Covered communications-infrastructure company included with a separate valuation update.

  Strengths

  Citi expects AI servers, computing-related products and consumer businesses to offset carrier-network weakness.

  Weaknesses

  Carrier-network segment weakness remains a headwind.

  Risks

  Lower computing-infrastructure investment, faster 5G base-station price declines and global market-share pressure related to US-China disputes.

Key data

• Chinese supplier share of high-speed modules to major US CSPs60–70%Citi estimate supporting its view that a near-term restriction would disrupt US AI infrastructure.
• US domestic-content threshold≥65% US-origin content; 75% from 2029Threshold cited for the production-location-based framework.
• Eoptolink overseas capacity88.4%Thailand; high impact under named-entity and all-foreign-location scenarios, moderate under China-only restriction.
• DSBJ overseas capacity30.3%Taiwan/Thailand; Thailand capacity is expanding during 2026.
• TFC overseas designed production capacity0.3%TFC is indirectly exposed as a passive-component supplier.
• DSBJ target priceRmb350Derived using SOTP valuation.
• Eoptolink target priceRmb701Based on 20.0x FY2027E P/E.
• Suzhou TFC target priceRmb419/shareBased on 34.3x 2027E P/E.
• ZTE target priceHK$25.40Based on 15.0x 2026E P/E.

Impact & implications

Citi believes a broad restriction would create a conflict between regulatory objectives and the near-term needs of US AI infrastructure. Direct module exporters Eoptolink and DSBJ would be most affected in stringent scenarios, while TFC’s exposure would mainly flow through downstream order weakness; offshore manufacturing only clearly helps in a China-only formulation.

Risks

• A formal FCC optical-transceiver restriction could materially disrupt direct exporters Eoptolink and DSBJ, particularly under named-entity or all-foreign-production scenarios.
• For Eoptolink, Citi cites weaker data-center investment, reduced network capex, price competition, customer-expansion delays, China-US tech disputes and faster CPO deployment.
• For DSBJ, Citi cites slower module FPC progress, weaker Tesla growth, optoelectronics losses, higher material costs and geopolitical risk.
• For TFC, stated risks include weaker AI and data-center investment, component constraints, rising competition and delayed CPO deployment.
• For ZTE, stated risks include lower computing-infrastructure investment, faster 5G base-station price declines and US-China-related market-share pressure.

What to watch

• Whether a formal optical-transceiver rule is published; as of 9 August 2026, Citi states that none exists.
• How the FCC defines “produced in a foreign country” versus “produced in China,” including the role of design, firmware and supply-chain control.
• September and November 2026 US-China diplomatic events, which Citi identifies as possible negotiation inflection points.
• The ramp-up of US automated optical-module production and whether domestic alternatives can meet AI-related demand.
• Eoptolink’s and DSBJ’s overseas-capacity expansion and the resulting protection under any China-only restriction.