Three Major Investment Banks Interpret Optical Module Ban Rumors: China's Capacity Irreplaceable in Short Term, U.S. Suppliers May Seize Multi-Billion Dollar Opportunity
0xBroomberg
Citi, Morgan Stanley, and JPMorgan have all weighed in on the reported FCC plan to ban Chinese optical transceivers from U.S. data centers. The consensus: China's supply cannot be replaced overnight and a clean break is impractical, but non-Chinese suppliers stand to capture nearly $10 billion in redirected demand.
Why is this ban so hard to enforce overnight?
Citi notes that seven of the world's top ten optical-transceiver makers are Chinese, supplying more than 50% of high-speed modules to major U.S. cloud providers. This means → if the ban takes effect, over half the supply chain goes dark with no ready substitute.
Morgan Stanley flags a deeper chokepoint: InP substrates — indium phosphide, the semiconductor material used to make high-speed lasers inside modules — still come from China. Lumentum just signed a new supply deal last week; Coherent's CEO visited China with a Trump delegation months ago. Both have active ties to Chinese InP suppliers.
In plain terms = the dependency isn't just finished modules. The raw material needed to *make* modules also runs through China — turning enforcement into a "cut both ends at once" problem.
If the ban lands, who gets the orders first?
Morgan Stanley estimates Zhongjiguangchuang (中际旭创) and Eoptolink (新易盛) together hold roughly 50% of the transceiver market. Among non-Chinese vendors, Coherent — vertically integrated from chips to modules — is the clearest scaled beneficiary. AAOI and Fabrinet could also pick up incremental share.
JPMorgan puts a number on it: third-party forecasts project the data-center optical market at over $50 billion in coming years. Zhongjiguangchuang and Eoptolink hold 20%+ and ~15% respectively, with more than half tied to U.S. revenue. That implies a potential addressable opportunity for U.S. suppliers approaching $10 billion.
JPMorgan also ranks beneficiaries: Coherent, Lumentum, Fabrinet, and Credo are more directly exposed than Ciena or Corning. This reflects a simple pricing logic — the closer a company sits to finished optical modules, the more certain the upside.
How long can Chinese suppliers play the "Thailand card"?
Citi ranks the impact on Chinese names: Tianfu Communication (indirect) < Dongshan Precision (lasers + modules) < Eoptolink (modules). Tianfu mainly makes passive components; its direct customer is a Thailand-headquartered overseas module company, and its Thai factory is still ramping — short-term impact is manageable.
Eoptolink derives roughly 88% of revenue from its Thailand base, where tariffs and restrictions are significantly lighter; the company has pledged to keep expanding there. Dongshan Precision ships all U.S.-facing modules from Thailand and Taiwan today, targeting 35 million 800G/1.6T units by 2027 — about two-thirds from Thailand, one-third from Taiwan.
But Citi flags a key risk: if Washington extends restrictions to Chinese-affiliated capacity in third countries, the Thailand buffer weakens. In plain terms = the Thailand card works today, but the rules of the game can change at any time.
Is "supply-chain disruption" actually the biggest risk?
JPMorgan argues the greatest risk to U.S. suppliers' upside is not the supply chain itself — it is that the rule gets shelved or watered down. Policies like this often acquire exemptions and workarounds between headline and final text.
Morgan Stanley adds that most cloud operators have already anticipated this scenario and qualified backup suppliers, though those backups still cannot cover full demand.
This means → the policy-negotiation window between "rumor" and "final rule" is the real pricing event. The ultimate scope of exemptions matters more than the binary question of ban-or-no-ban.
Content is for reference only, not financial advice.