Recently, the domestic optical communications sector has experienced volatile trading, with the core disturbance consistently revolving around expectations for US trade policy. From reports in early August that the FCC might impose a full ban on Chinese optical modules, to the current market chatter about a new "65% US-origin content exemption," market sentiment has been oscillating between pessimism and recovery. Is this 65% content rumor actually credible? For domestic optical module manufacturers, is this threshold high or low? If the policy materializes, what kind of transformation will the entire industry chain undergo? Nomura Securities recently released a research report analyzing these questions. The report argues that the broader logic of surging AI computing power driving optical module demand remains unchanged, and trade policy has always been merely a short-term sentiment disturbance. For the industry and investors alike, there is no need to be overly pessimistic about trade rumors, nor should there be complacency merely because of exemption expectations.
Where the 65% rule comes from
Many people's first reaction upon hearing the number 65% is that it is arbitrary speculation, but in reality, this standard has precedents in the US industrial regulatory system. Previously, the US introduced a federal procurement "Made in America" standard for the drone industry, explicitly requiring that the proportion of US domestic content in products must exceed 65%. The similar requirement now rumored for the optical module industry is essentially an extension of existing regulatory logic into the high-tech domain. Tracing the entire chain of events, the most pessimistic expectation previously held by the market was that the US would fully ban next-generation optical modules produced by Chinese manufacturers from entering the US market. For domestic optical module leaders deeply bound to North American AI data center customers, this would be equivalent to directly losing their largest core market, which triggered significant sector volatility. If the 65% US-origin content exemption clause comes to pass, it would be equivalent to opening a door in the wall of a comprehensive ban: as long as US-produced components in an optical module account for more than 65% of the total bill of materials (BOM) cost, Chinese manufacturers can still supply next-generation 3.2T optical modules to US customers. From an industry fundamentals perspective, Chinese optical module manufacturers have already established a highly solid global leadership position, with overall global market share exceeding 50% in high-end products such as 800G and 1.6T. However, it is unavoidable that in core segments such as high-end chips, the domestic supply chain still objectively depends on US suppliers, which also forms the realistic basis for the 65% rule's feasibility.
Is the 65% threshold actually high?
To judge whether the 65% requirement can be met, the key is to calculate one thing clearly: how much of the cost in a high-end optical module comes from US-origin components. The Nomura report uses current mainstream 800G and 1.6T optical modules as examples, where material costs are concentrated in four types of chips: digital signal processors (DSP), optical chips/lasers, transimpedance amplifiers (TIA)/driver chips, and photodiodes (PD). According to estimates, these four core chip categories alone account for 66% to 72% of the total BOM cost for 800G EML-solution and silicon photonics-solution optical modules, and for the two technology routes in 1.6T products, the proportion is basically in the same range. Among these, DSP chips account for the highest share, generally exceeding 30% of BOM costs, making them the most critical "brain" of optical modules. This segment is currently basically monopolized by US companies such as Broadcom and Marvell, which are inherently pure US-origin components. Beyond that, in segments such as high-end optical chips and TIA driver chips, US suppliers also occupy a dominant position. In other words, even excluding other scattered US-origin components, the natural US-origin share from core chip segments alone is already close to or even above the 65% threshold. For Chinese optical module manufacturers, as long as they maintain their existing core chip procurement structure, meeting the 65% US-origin content requirement is not an unrealistic goal. Among these, optical chips/lasers, which account for 14% to 23% of BOM costs, are the key variable determining whether the 65% threshold can be stably met.
Who wins and who faces pressure?
If the 65% US-origin content rule is formally implemented, the profit landscape across the optical module industry chain will clearly diverge, with impacts varying dramatically across different segments. The first beneficiaries are US high-end optical chip manufacturers. After the rule takes effect, downstream optical module manufacturers will prioritize purchasing optical chips from US suppliers to meet the US-origin content requirement, which will further increase US companies' market share in high-end optical chips and strengthen their pricing power. Correspondingly, Chinese optical chip manufacturers such as Yuanjie Semiconductor, which have made rapid breakthroughs in the global market over the past one to two years, will face greater competitive pressure, and the pace of their global expansion may be constrained. Next are the DSP, electrical chip, and PD chip segments. This area has always been the home turf of US companies, with players such as Broadcom, Marvell, and MACOM already occupying absolutely leading market positions. After the rule takes effect, the existing supply chain structure will not change much and will basically remain as is. Third are supporting segments such as passive optical components and PCBs. These products themselves account for a very low proportion of total optical module BOM costs, and their supply chains are highly globalized. They are neither areas monopolized by US companies nor core targets of this policy regulation, so they will hardly face any noticeable impact. Finally, turning to the optical module manufacturing segment, namely the domestic leading manufacturers: in the short term, business continuity will not be substantially disrupted. After all, the 65% threshold can be met, and the channel for shipping to the US has not been completely blocked, which is far better than the previous pessimistic expectation of a comprehensive embargo. But the potential risk lies on the cost side: if, in order to meet the content requirement, manufacturers are forced to purchase more expensive US-origin components, or if increased concentration of core chip supply leads suppliers to raise prices, this will push up overall manufacturing costs for optical modules and squeeze manufacturers' profit margins.
Chinese optical modules: the long-term test has only just begun
For domestic leading optical module companies such as Innolight and Eoptolink, if the 65% rule is implemented, the overall impact is neutral to slightly positive, significantly better than the previous expectation of a full embargo. Such industry leaders are themselves deeply bound to leading North American cloud providers and AI customers, have mature supply chain management capabilities, and are in the global first tier for R&D iteration speed. As long as the channel for shipping to the US remains open, their global leadership position will be difficult to shake in the short term, given their years of accumulated manufacturing expertise, customer resources, and economies of scale. But it must also be clearly recognized that this rule once again highlights the core weakness of China's optical module industry chain. External dependence on core components has always been a sword hanging over the industry. The 65% threshold may appear easy to meet, but in essence it trades supply chain autonomy for market access. In the future, if the US further raises domestic content requirements or tightens core chip supply, the industry will still find itself in a passive position. From a long-term perspective, domestic substitution in core segments such as optical chips will only become increasingly urgent. The short-term trade exemption is more like giving the domestic industry chain a valuable buffer period. What can truly enable China's optical module industry to travel steadily and far is never an access qualification granted by others, but core technologies held in its own hands. Overall, the 65% US-origin content rule currently remains at the stage of market rumor, and whether the FCC will ultimately introduce a formal policy and what the specific implementation details will be still carry considerable uncertainty. But this rumor itself has already brought room for sentiment recovery to the optical module sector that has been under persistent pressure.
Comments