Ligent Hong Kong IPO Targets $723 Million as Zhongji Sets a Much Bigger Benchmark

Ligent Technologies launched its Hong Kong IPO seeking HK$5.67 billion, or about $723 million, while investor confidence in AI hardware faces a fresh test.

The optical communications supplier is offering approximately 172 million shares before any overallotment exercise. Trading is scheduled to begin September 22 under stock code 9856. The deal would value Ligent at roughly HK$32.4 billion.

The offering follows Zhongji Innolight’s HK$53.4 billion listing in July. That deal was almost ten times larger, yet Zhongji’s shares still closed below their offer price on the first trading day.

That contrast defines the Ligent Hong Kong IPO. Demand for AI data-center networking is lifting manufacturers’ revenue, but investors are becoming less willing to reward growth without examining margins, customer exposure, and capital requirements.

Ligent now has to show that its smaller scale represents room for expansion, not a lasting disadvantage against Zhongji.

The Ligent Hong Kong IPO Turns an AI Supplier Into a Public-Market Test

Ligent is selling investors exposure to the physical connections inside AI infrastructure, not another application-layer software story.

The company develops optical transceivers, optical chips, and optical network terminals. An optical transceiver converts electrical and optical signals, allowing servers and switches to exchange data through fiber connections.

These components matter because large AI clusters must move information rapidly among accelerators, storage systems, and networking equipment. Processing capacity becomes less useful when data cannot travel through the cluster with comparable speed.

According to the company’s Hong Kong filing, Ligent operates across product development, manufacturing, and optical-chip production. That vertical reach supports the company’s claim that it can address several layers of an optical connection.

The final offering includes about 172 million shares, with roughly 17.2 million allocated initially to Hong Kong investors. The remaining 154.8 million shares are assigned to the international offering.

An overallotment option allows underwriters to sell approximately 25.8 million additional shares. If exercised fully, it would increase both the proceeds and the number of shares available to investors.

Ligent expects about HK$5.45 billion in net proceeds before that option. The difference from gross proceeds reflects underwriting fees and other listing expenses.

The company has assigned 52.9% of net proceeds to research and development. Another 25.1% is earmarked for production expansion and increased automation across optical-transceiver and optical-chip lines.

Ligent plans to direct 8% toward investments and acquisitions. Overseas promotion receives 4%, while the remaining 10% supports working capital and general operations.

That allocation makes the offering more than a liquidity event for existing owners. Ligent is asking public investors to finance its attempt to move further into faster, higher-value optical products.

The company’s offering timetable opened public applications on September 14. Applications close September 17, followed by an expected allocation announcement on September 21.

Cornerstone investors have committed $340 million, equal to about 47% of the base offering. Cornerstone investors agree to buy a fixed allocation before public trading begins, usually with a lockup period.

The group includes Primavera Investment Fund, GigaDevice, Amlogic Hong Kong, Mirae Asset Securities Hong Kong, PAG, Barings, and funds linked to ORIX.

That commitment provides a substantial demand floor. It also leaves fewer freely allocated shares, making the eventual public and international subscription levels important indicators of broader interest.

Ligent’s parentage adds another layer. Hisense Group controlled 48.6% before the offering and is expected to retain about 40.1% afterward, assuming no overallotment exercise.

Ligent maintains operations in China, Thailand, and the United States, including a Silicon Valley office in San Jose. Its international footprint supports overseas sales while exposing the business to several regulatory systems.

The transaction therefore combines three investor themes: AI infrastructure demand, Chinese technology manufacturing, and Hong Kong’s renewed role as a capital-raising center.

Those themes can attract buyers. They also bring different valuation risks into the same security.

Why Ligent Is Raising Capital Now

The timing reflects a rapid increase in optical demand, but it also shows how much capital suppliers need to keep pace with AI networks.

Ligent reported revenue of RMB8.35 billion for 2025, up 64.2% from RMB5.09 billion in 2024. Profit reached RMB872.6 million.

For the first six months of 2026, revenue increased 27.9% from a year earlier to RMB5.39 billion. Profit rose 29.7% to RMB661 million during the same period.

Those figures establish that Ligent enters the market with expanding operations. This is not a pre-revenue supplier attempting to fund an untested manufacturing plan.

However, the growth rates do not eliminate questions about earnings quality. Ligent’s 2025 profit included a RMB353 million gain from disposing of a business interest, according to prospectus-based financial analysis.

That one-time contribution means headline profit does not fully describe recurring performance. Investors need to separate operating improvement from gains that will not repeat.

Product mix is another important variable. Demand has been shifting from slower connections toward 800-gigabit and 1.6-terabit products, which handle more data during each second of operation.

The transition creates opportunities for suppliers that qualify products early. It also compresses development schedules and increases the cost of manufacturing equipment, testing systems, and customer certification.

A data-center operator cannot replace a networking component based only on a laboratory specification. New modules must satisfy performance, reliability, power, thermal, and interoperability requirements within an operating network.

That qualification process can strengthen incumbent suppliers. Once approved, a manufacturer gains a pathway to large orders, but a delayed product can miss an entire purchasing cycle.

Ligent’s planned spending addresses that race directly. Research funding supports new products, while manufacturing investment aims to convert approved designs into reliable volume.

The company had 1,581 granted patents and 726 pending applications as of September 5. It had also participated in 61 industry standards through 18 standards organizations.

Patent counts do not automatically establish commercial leadership. They do show that Ligent has accumulated intellectual property across a technically demanding product category.

Ligent ranked fifth among specialized optical-transceiver manufacturers by 2025 global revenue, according to market data cited in its prospectus. Its reported global share was 4%.

Within China, the company ranked third with a 10.1% share. That stronger domestic position gives Ligent a base for expansion, although it also highlights the gap between its Chinese and global standings.

Industry demand provides the favorable part of the timing. A competitor’s annual market review cited global optical-module and related-product sales of $23.8 billion in 2025.

That report also placed Ethernet optical-module sales near $18 billion after rapid annual growth. Ethernet products connect equipment across modern data-center networks.

The numbers explain why optical suppliers want capital now. Cloud operators are building larger clusters, and each cluster requires more high-bandwidth links between computing nodes.

Yet demand alone does not guarantee attractive returns for every manufacturer. Large customers can demand lower prices while requiring suppliers to finance capacity before firm orders arrive.

Production expansion therefore contains a built-in tradeoff. Too little capacity can forfeit orders, while too much capacity can leave expensive equipment underused after demand changes.

Ligent is raising money while revenue momentum remains visible. Public investors must decide whether that momentum justifies financing the next manufacturing cycle.

Zhongji Innolight Is the Benchmark Ligent Cannot Avoid

Ligent’s main challenge is not proving that optical networking matters, but proving it can narrow a large competitive gap without sacrificing returns.

Zhongji Innolight provides the clearest comparison because it sells high-speed optical products into the same AI infrastructure cycle. It also completed Hong Kong’s largest share sale of 2026.

Zhongji raised HK$53.41 billion by selling 54.5 million shares. The transaction ranked as Asia’s second-largest offering of the year, behind CXMT’s Shanghai listing.

The scale difference is striking. Ligent seeks HK$5.67 billion in gross proceeds, placing its offering at roughly one-tenth of Zhongji’s fundraising total.

Zhongji reported a 21.2% share of global optical-interconnect revenue for 2025, citing data from China Insights Consultancy. Ligent’s reported optical-transceiver share was 4%.

The categories are not perfectly identical, so the percentages should not be treated as a direct scorecard. They still demonstrate Zhongji’s much larger presence in high-speed optical connectivity.

Zhongji also entered its offering with exceptional financial momentum. First-quarter 2026 revenue nearly tripled to RMB19.5 billion, while profit increased to RMB6.32 billion.

The company attributed that expansion to major customers investing in AI infrastructure. The United States generated 61.7% of its first-quarter revenue.

Despite those results, the market delivered a warning. Zhongji closed 2% below its offer price on July 30 after falling as much as 10.2% during the session.

Its mixed trading debut occurred as investors reconsidered AI valuations and the growing cost of data-center construction.

The offering itself attracted demand. Hong Kong subscriptions reached 16.84 times the available allocation, while the international tranche was subscribed 9.73 times.

Strong orders before trading did not prevent first-day losses. That separation matters for Ligent because subscription demand and sustainable public-market support measure different things.

Cornerstone commitments can help complete an offering. They cannot guarantee how unrestricted investors will price the company after trading begins.

Zhongji’s experience reverses the easiest bullish narrative around the Ligent Hong Kong IPO. A company can combine AI exposure, strong growth, and a major fundraising event without receiving an immediate market premium.

Ligent must therefore compete on a different proposition. It cannot credibly present itself as another Zhongji at the same scale.

Instead, its case depends on moving from a 4% global share toward a larger position while protecting margins. The new capital must produce qualifying products, usable capacity, and customer wins.

Other manufacturers complicate that path. Coherent, Lumentum, and Eoptolink also compete across optical chips, modules, or related communications products.

These companies differ in product mix and geographic exposure. However, all face the same underlying pressure to deliver faster connections while lowering energy use and unit costs.

For Ligent, vertical integration could reduce dependence on outside chip suppliers and shorten development cycles. It could also increase capital needs and create execution risk across more manufacturing stages.

Zhongji remains the primary benchmark because investors have already priced its scale and growth in Hong Kong. Ligent’s debut will reveal whether the market also wants a smaller challenger.

Faster AI Networks Do Not Guarantee Better Supplier Economics

The mechanism behind Ligent’s growth is clear, but the value captured by component makers remains less certain than the demand for their products.

AI accelerators work in parallel, which creates heavy communication among thousands of computing devices. Optical connections move that traffic over fiber with lower signal loss across distance than conventional electrical links.

As clusters grow, networking can become a bottleneck. A slow connection leaves expensive processors waiting for data, reducing the useful output of the entire installation.

Faster transceivers address this constraint. An 800-gigabit module can carry twice the nominal data rate of a 400-gigabit module, while 1.6-terabit products double it again.

Real network performance depends on more than nominal speed. Power consumption, heat, error rates, reach, software compatibility, and switch architecture also shape deployment decisions.

These requirements favor manufacturers with established engineering teams and customer relationships. Ligent’s combination of optical chips and completed modules is meant to support that position.

The company operates production facilities in Qingdao, Jiangmen, Thailand, and the United States. Multiple locations provide manufacturing options, but they also add supply-chain and coordination demands.

Ligent says the planned proceeds will expand production capacity and improve automation. Automation can increase consistency when thousands of optical assemblies must meet tight tolerances.

It can also reduce the labor content of each unit. However, investors should not assume that higher automation automatically creates stronger margins.

Optical products often become less profitable as standards mature and more suppliers enter. Manufacturers must recover development costs before a module becomes widely available at lower prices.

Lumentum describes that pressure in its regulatory disclosures. The company warns that optical communications products face commoditization, pricing pressure, and concentrated customer demand.

Those are industry-wide risks, not evidence of a specific problem at Ligent. They explain why market growth can coexist with difficult economics for individual vendors.

Large cloud customers possess significant purchasing power. They can divide orders among suppliers, demand customized products, or shift architectures before a capacity investment earns its expected return.

Product generations also move quickly. A supplier that expands an older line too aggressively can face lower utilization as customers transition toward faster hardware.

Ligent’s R&D allocation therefore matters as much as its capacity budget. The company needs products that match where customers are going, not merely additional output from its present catalog.

Its optical-chip operations could become an advantage if they improve cost control and product coordination. Chip integration can also help a module supplier respond to performance and power requirements.

Still, ownership of more production steps does not remove market dependence. Customer certification, manufacturing yields, component availability, and delivery schedules remain decisive.

The practical use case sits inside each AI cluster. Servers equipped with accelerators must exchange model parameters, training data, and intermediate results across a network.

A delayed or unreliable optical module can impair that network even when processors perform correctly. Buyers therefore care about reliability and service alongside headline bandwidth.

That operating reality gives qualified suppliers a defensible role. It does not guarantee that every supplier earns premium returns throughout the product cycle.

The Ligent Hong Kong IPO asks investors to fund both sides of this mechanism. More AI computing creates more optical demand, but capturing that demand requires continuing investment before outcomes become certain.

Cornerstone Demand Does Not Remove the Offering’s Risks

The strongest skepticism concerns whether Ligent can convert rapid growth and new capacity into durable, recurring profit.

The first risk involves the composition of earnings. Ligent’s 2025 profit benefited from a one-time disposal gain, so comparisons based only on reported net income can overstate operating progress.

Its 2026 first-half figures offer a cleaner signal because revenue and profit both increased from the prior-year period. Future disclosures still need to show whether operating margins improve alongside sales.

The second risk is customer concentration. Optical suppliers often depend on a small group of cloud providers and networking vendors, even when prospectuses avoid naming customers for commercial reasons.

A large customer can change suppliers, reduce orders, delay a deployment, or demand a lower price. Any of those decisions can affect factory utilization quickly.

Purchase forecasts do not always become binding commitments. Suppliers sometimes acquire equipment and inventory before receiving final volumes, increasing exposure when plans change.

Ligent’s production expansion magnifies this issue. New factories and equipment can support revenue growth, but underused capacity increases depreciation and other fixed costs per unit.

The third risk is technological timing. The transition toward 800-gigabit and 1.6-terabit links creates demand, yet it also shortens the useful commercial window for earlier products.

Customer qualification can take months. A delayed certification may allow Zhongji or another supplier to secure the available order before Ligent enters production.

The fourth risk comes from trade policy. Ligent operates across China, Thailand, and the United States while serving customers in several markets.

Optical chips and advanced communications equipment sit near areas affected by export controls, tariffs, investment restrictions, and supply-chain security reviews.

The precise impact depends on individual products, customers, and rules. It would be premature to claim that current restrictions will block Ligent’s offering or operating plan.

However, Zhongji’s experience shows that regulatory exposure can enter the valuation quickly. The United States added Zhongji to a list of companies suspected of Chinese military ties in June.

Zhongji disputed that designation and said it had not caused material order cancellations, suspensions, or delays. Investors still had to price the possibility of future consequences.

Ligent’s controlling-shareholder structure presents another consideration. Hisense is expected to remain the largest owner after the IPO, giving it substantial influence over corporate decisions.

A committed parent can support long-term planning and commercial coordination. Minority investors must still consider whether future decisions consistently align with their interests.

The cornerstone allocation also deserves measured interpretation. A $340 million commitment demonstrates institutional willingness to participate at the offer valuation.

Because those investors receive predetermined allocations, their involvement is not equivalent to unrestricted buying after listing. Their shares are also generally subject to lockup arrangements.

A large cornerstone percentage can reduce the immediately available float. Lower float sometimes amplifies price movements when public trading begins.

Zhongji disclosed concentrated ownership among its Hong Kong shareholders and warned that concentration could produce sharp price changes. Its volatile first session illustrated that risk.

Ligent’s smaller deal may behave differently, but smaller size does not automatically mean lower volatility. Subscription levels, allocation concentration, and broader technology sentiment will all matter.

The company’s market-share claims also require context. The cited rankings rely on third-party market definitions that can differ across optical modules, interconnect solutions, chips, and terminals.

A 4% share indicates a meaningful position. It does not prove leadership in every high-speed product needed by the largest AI customers.

None of these risks invalidates Ligent’s growth. They set the conditions that investors should use to judge whether the offering’s proceeds create lasting competitive gains.

Three Signals Will Decide What the Ligent Hong Kong IPO Means

The listing result, operating margins, and high-speed product adoption will show whether Ligent is closing the gap with Zhongji or merely financing a costly expansion.

The first signal is demand through September 22. Investors should watch subscription levels, the final allocation, and trading during Ligent’s first session.

Heavy subscriptions would confirm interest in another optical-infrastructure company. A stable debut would carry more weight than subscription multiples because it reflects unrestricted market pricing.

A weak debut would not disprove the AI networking thesis. It would show that public investors want more compensation for Ligent’s smaller scale and execution risks.

The second signal is the next set of financial results. Revenue growth already shows demand, but operating profit and cash generation will reveal whether Ligent captures enough value.

Investors should separate recurring earnings from disposals and other one-time items. They should also compare capital expenditures with the additional revenue produced by new capacity.

Improving operating margins would strengthen the claim that product mix and scale are working together. Falling margins would suggest pricing pressure or expensive capacity expansion.

The third signal is commercial progress in faster optical products. Ligent needs visible adoption of its newer transceivers and optical chips among major cloud or equipment customers.

Product announcements alone will not settle that question. Customer qualifications, volume shipments, and repeat orders provide stronger evidence than laboratory performance.

Progress outside China will be especially informative. Ligent’s factories and San Jose presence give it an international platform, but revenue quality depends on sustained customer relationships.

These three indicators should be read together. A strong debut without operating progress would reflect sentiment, while better margins without product adoption might prove temporary.

Ligent enters Hong Kong with real scale, accelerating sales, and a defined investment plan. It also enters behind a rival whose larger and faster-growing business received a cautious market welcome.

That is why the Ligent Hong Kong IPO matters beyond its fundraising total. It measures how investors value the less visible hardware connecting AI systems after enthusiasm meets manufacturing reality.

Watch the September 22 debut first, then follow margins and high-speed product shipments. Those signals will show whether Ligent’s smaller base creates valuable growth or a more expensive chase.