EN / 中文

Zhongji InnoLight Valuation Debate: Can AI Optical Module Growth Sustain Profit Margins by 2029?

by aerfagongchang·November 19, 2026

Lead-in: A set of assumptions regarding 2029 compressed the future profits of a leading AI optical module manufacturer to CNY 30 billion, bringing the divergence in market views on AI industry chain valuations back into the spotlight.

On November 17, a bearish valuation report on Zhongji InnoLight rapidly circulated within the investment community.

The most striking aspect of this material was not simply providing a bearish target price, but attempting to derive the reasonable value of Zhongji InnoLight for the coming years using a comprehensive long-term model.

It selected global AI capital expenditure of USD 800 billion in 2029 as the benchmark, assuming that optical interconnect value accounts for 7% of AI hardware capital expenditure, Zhongji InnoLight captures a 50% market share, and the net profit margin falls back to 15%.

According to this model, Zhongji InnoLight's net profit in 2029 would be approximately CNY 30 billion.

Applying a P/E ratio of 10 to 15 times, the company's reasonable market cap at that time would be approximately CNY 300 billion to CNY 450 billion. Considering discount factors, even under the most optimistic scenario, the calculation shows that the current stock price still has about 25% downside potential.

The final recommendation in the material suggests that allocation-oriented funds could consider designing a left-side grid low-buy strategy when the market cap is below CNY 400 billion.

This is not a simple bearish conclusion. It attempts to answer a more difficult question: If the high-speed growth of the AI industry eventually returns to normal levels, what should Zhongji InnoLight actually be worth?

However, this calculation quickly sparked controversy.

On one hand, the market is reassessing the sustainability of AI infrastructure investment; overseas research institutions have lowered capital expenditure forecasts, and AI hardware stocks have undergone a noticeable adjustment. On the other hand, Zhongji InnoLight's recently released operational data remains quite strong, with orders, product prices, and gross margins showing no obvious deterioration.

Why did a calculation based on 2029 assumptions trigger such a strong reaction at this specific moment?

How Was the CNY 30 Billion Net Profit Calculated?

First, let's look at the core calculation process of this material.

Instead of starting directly from Zhongji InnoLight's existing revenue and profits to forecast performance over the next three years, it chose to start from the scale of global AI infrastructure investment, working backward to deduce the optical interconnect market space, and then calculating the potential revenue and profits for Zhongji InnoLight.

The entire process can be broken down into four steps.

Step 1: Determine global AI capital expenditure.

The report sets 2029 AI capital expenditure at USD 800 billion.

This figure is not entirely without external basis. In its model updated on November 13, semiconductor research firm SemiAnalysis significantly lowered its forecast for global AI accelerator capital expenditure for the coming years, reducing the 2029 forecast from the previous USD 1.352 trillion to USD 796 billion.

This indicates that expectations for growth in global AI infrastructure investment are indeed changing.

However, attention must be paid to a scope issue here: This data from SemiAnalysis refers to global AI accelerator capital expenditure, which is not equivalent to total AI industry capital expenditure. The two cannot be directly equated.

Step 2: Determine how much investment optical interconnects capture.

This calculation assumes that after 2029, pluggable optical modules and NPO combined account for 7% of AI hardware capital expenditure.

Specifically, the value of pluggable optical modules corresponding to Scale-out scenarios accounts for about 5%, while in Scale-up scenarios, as demand for optical interconnects inside racks increases, technologies like NPO account for about 2% of the value.

Based on USD 800 billion, the value of the global optical interconnect market is approximately USD 56 billion.

Step 3: Determine Zhongji InnoLight's market share.

The report assumes that Zhongji InnoLight can secure a combined 50% market share in the fields of pluggable optical modules and NPO.

Calculating accordingly, the company's corresponding revenue scale is approximately USD 28 billion.

Converted at an exchange rate of 6.75, the relevant revenue is approximately CNY 189 billion.

If calculated again using a 15% net profit margin, Zhongji InnoLight's net profit in 2029 would be approximately CNY 28.3 billion, which the report rounds up to CNY 30 billion.

Step 4: Determine valuation.

The report provides a valuation range of 10 to 15 times P/E.

Based on a net profit of CNY 30 billion, the corresponding year-end market cap in 2029 is CNY 300 billion to CNY 450 billion.

At this point, the entire bearish model is complete.

If further considering the time value of money, the report also uses a discount rate of 8% to 12% to discount the 2029 year-end valuation.

The most optimistic scenario involves a market cap of CNY 450 billion, a 15x P/E ratio, and an 8% discount rate. Discounted to the present, the valuation is approximately CNY 357.2 billion.

In the neutral scenario, the valuation is approximately CNY 270.5 billion; in the pessimistic scenario, it is only CNY 213.5 billion.

This is why the report concludes that even under the most optimistic assumptions, Zhongji InnoLight still faces about 25% downside potential.

Mathematically, this model is not complex; it is quite intuitive.

What truly warrants discussion are several key parameters within it.

Because if any one parameter changes, the final valuation could be completely different.

The Core of the Controversy

The most direct reason this calculation triggered a market backlash is the assumption of a 15% net profit margin.

In the first three quarters of 2026, Zhongji InnoLight achieved operating revenue of CNY 69.547 billion, a year-on-year increase of 178.13%; net profit attributable to parent company shareholders was CNY 22.238 billion, a year-on-year increase of 211.81%.

Based on these figures, the company's net profit margin attributable to parent company shareholders in the first three quarters was approximately 32%.

The 15% assumed in the report is less than half of the company's current net profit margin.

If a company's revenue is still growing rapidly, product prices and gross margins remain stable, and customer orders continue to increase, why compress the net profit margin directly to 15% when forecasting the next three years?

This is exactly where the market raises doubts.

Of course, the 15% net profit margin is not entirely without room for discussion.

The optical module industry is not an industry that can maintain high profit margins forever.

In recent years, demand for 800G products has grown rapidly, and 1.6T products have begun entering mass production phases. Due to increasing demand for high-speed optical interconnects in AI data centers, enterprises with technological, customer, and capacity advantages have gained strong bargaining power.

But this supply-demand relationship cannot last forever.

As more companies enter the high-end optical module market, capacity gradually releases, and customers will re-evaluate pricing and costs among suppliers. Even if industry demand continues to grow, declining product prices, intensified competition, and regression of profit margins may all occur simultaneously.

Therefore, the real question bearish analysts need to answer is not how high today's profit margins are, but whether the industry can sustain its current profitability structure three years from now.

The problem lies in the fact that the circulating calculation materials do not adequately explain why 15% represents Zhongji InnoLight's long-term steady-state profit margin.

The rationale provided is that as supply-demand balance restores and industry supply increases, profit margins will be significantly suppressed.

This is a possible scenario, but it does not equal a proven industry law.

Especially since Zhongji InnoLight's profitability does not stem solely from tight industry supply and demand.

In the optical module industry, product iteration speed, R&D investment, customer certification, delivery capability, and supply chain management all influence how much profit an enterprise can ultimately earn.

If the industry enters a new round of competition, will all companies' profit margins decline synchronously, or will leading enterprises with technology and customer advantages retain higher profit margins?

These two scenarios would have completely different impacts on Zhongji InnoLight's long-term valuation.

Another controversy is the 50% market share.

The report combines the market share of pluggable optical modules and NPO, assuming Zhongji InnoLight can capture half of it.

However, the future optical interconnect market is not just a simple extension of existing products.

Scale-out and Scale-up have different demands for optical connections, and there are competitive and substitution relationships between NPO, CPO, pluggable optical modules, and other technical solutions.

A company having strong competitive advantages in 800G and 1.6T products does not mean it will necessarily maintain the same market share in the next-generation technology roadmap.

Conversely, if Zhongji InnoLight can continuously participate in new product R&D and maintain leadership in customer introduction and mass production phases, the future market space may not necessarily be calculated based on today's product structure.

This is the most easily overlooked aspect of this model: It places an industry that is still changing rapidly into a mature competitive framework prematurely.

As for the 10 to 15 times P/E ratio, there is also controversy behind it.

The report compares future Zhongji InnoLight with mature communication infrastructure enterprises and the cyclical storage industry, arguing that as AI capital expenditure growth slows, optical module enterprises will eventually return to the valuation levels of mature industries.

This logic is not without merit.

If AI hardware investment enters a low-growth phase in the future and optical module enterprises' profit margins also decline significantly, then the market indeed has no reason to continue granting them valuations associated with high-growth stages.

But the question is, will Zhongji InnoLight have entered a mature cycle by 2029?

If next-generation products following 800G and 1.6T are still ramping up at that time, and optical interconnect demand driven by Scale-up continues to grow, whether a 10 to 15 times P/E ratio is sufficient to reflect the company's growth potential remains debatable.

In other words, the valuation multiple given by the report is itself an assumption requiring argumentation, rather than a naturally established conclusion.

The True Divergence in the AI Industry Chain

If we place this debate back into the context of the entire AI industry chain, we find that it is not just a problem for Zhongji InnoLight alone.

Over the past two years, the most important basis for the market's valuation of AI hardware companies was demand growth.

As long as NVIDIA continued to increase GPU shipments, and Microsoft, Google, Amazon, and Meta continued to expand capital expenditure, the market was willing to believe that most enterprises in the AI industry chain could win more orders.

But by the second half of 2026, this logic began to change.

Overseas research institutions successively lowered their forecasts for AI accelerator shipments and capital expenditure for the coming years. At the same time, the market began focusing on the investment returns of large cloud computing companies and the financing costs borne by AI infrastructure construction.

This does not mean AI demand has disappeared.

More accurately, the market has started recalculating: How much revenue can these investments ultimately generate, and how much profit can be retained across various links in the industry chain?

For optical module enterprises, this shift is particularly important.

Because optical modules are both a crucial component of AI data center construction and a manufacturing segment that requires continuous technological iteration and capital investment.

Demand growth is certainly important, but demand growth does not automatically equal profit growth.

If customers continue to expand procurement, but the number of suppliers increases and product prices persistently fall, optical module enterprises may face a situation where revenue grows but profit margins gradually decline.

If the next-generation technology roadmap changes, enterprises need to continuously invest in R&D and capacity building to maintain their existing market share.

This is why the market has shifted focus from orders to delivery, product prices, gross margins, and cash flow.

At least judging from currently disclosed information, Zhongji InnoLight's operational status has not yet shown the obvious deterioration described in this bearish model.

The company's investor relations record dated November 3 showed that the vast majority of clients had not revised down their demand or orders; some clients had adjusted product structures, and some clients' 1.6T demand had increased compared to before. The company stated that most orders for 2027 already cover the full year, at least one key client has provided clear demand guidance for 2028, and demand is strong.

On November 6, the company further disclosed that orders for products such as 800G and 1.6T in Q3 2026 achieved good quarter-over-quarter growth compared to Q2, 1.6T product revenue accelerated further, and product prices and gross margins remained stable.

The company also stated that demand for next-generation optical interconnect products such as NPO and XPO is expected to gradually release in the second half of 2027.

This information at least indicates that current industry demand and the company's order status still exist at a distance from a long-term pessimistic model built on shrinking demand.

But this does not mean Zhongji InnoLight can maintain its current growth rate forever.

The company's current order status can help the market judge operational trends for the next one or two years at most, but it cannot directly answer questions about market share, product prices, and profit margins in 2029.

This is the true dilemma investors need to face.

A company having high profit margins today does not mean they won't decline in the coming years; a company having orders covering the next year does not mean it will necessarily maintain the same profitability three years later.

Similarly, a report compressing the net profit margin from 32% to 15% cannot prove, based solely on a number, that the company's future value is capped at CNY 300 billion to CNY 450 billion.

Truly valuable research needs to explain why profit margins will decline, why market share will change, why industry capital expenditure will slow down, and at what speed these changes will occur.

Rather than giving a long-term valuation first and then deriving the conclusion through a few parameters.

For Zhongji InnoLight, a net profit of CNY 30 billion may be an overly pessimistic assumption, or it may be a stress scenario the company needs to face after intensified industry competition in the future.

But either way, it should not become the sole basis for judging a company.

In the past, the market was willing to pay a premium for AI industry chain growth; now, the market is beginning to require these enterprises to prove that growth can be converted into sustained profits.

This is the true reason this bearish calculation sparked controversy.

On the surface, it discusses what Zhongji InnoLight will be worth in 2029, but in reality, it touches upon a valuation transition currently undergoing in the entire AI hardware industry chain.

When capital expenditure is no longer viewed as perpetually growing, and when industry competition begins to impact profit margins, should the market price enterprises based on the speed of growth, or based on how much profit remains after growth ends?

This question clearly belongs to more than just Zhongji InnoLight.