Over the past month, the optical module sector has undergone a significant correction. Shares of ZJ INNOLIGHT (ASX: 03308), Eoptolink Technology, and T&S Communications (collectively referred to as the "Big Three") have all experienced varying degrees of pullback, and related companies such as Accelink Technologies and HGTECH have not been spared either. As of the October 30 close, cumulative declines for these companies mostly ranged between 30% and 50%.
Notably, this round of adjustment was not accompanied by weakening industry demand. Demand for 800G and 1.6T products continues to grow, with leading manufacturers' order visibility extending to 2027, and some customers even providing demand guidance for 2028. Brokerage earnings forecasts for the next one to two years have not been systematically revised downward either.
The third-quarter reports were also not poor. ZJ INNOLIGHT, Eoptolink Technology, and T&S Communications recorded Q3 revenue growth of 171.83%, 80.3%, and 31.5% year-over-year, respectively; net profit attributable to shareholders grew 173.74%, 61.15%, and 55.30% year-over-year, respectively. Yet even with such results, the sector's malaise could not be reversed.
Clearly, the market is no longer trading on current-period earnings, but rather on growth expectations for the next one to two years. Our assessment is that after this round of adjustment, the optical module sector has already digested a considerable portion of valuation and expectation pressure. In the short term, it is more likely to enter an L-shaped bottoming phase rather than a V-shaped rebound.
A large decline does not equal the exhaustion of negative factors, and a low valuation does not equal a stock price bottom. For the sector to regain strength, new positive signals are needed in orders, technology roadmaps, and earnings forecasts.
Therefore, this article mainly explores three questions: What exactly has this round of adjustment digested? What is still missing for a rebound? And among the three leading optical module companies, which one is better positioned to emerge first?
Two Concerns Weighing on the Market
The pressure from this round of adjustment mainly stems from two aspects: technology roadmap changes brought by CPO, and slowing growth in AI capital expenditure.
First, CPO. On October 21, NVIDIA announced the expansion of supply chain capacity, with partner Silicon Precision Industries planning to increase CPO optical engine capacity to 15 times the current level by mid-2027. This showed the market a possibility: CPO technology, which was still being gradually introduced, might achieve large-scale application faster than expected.
For optical module leaders like ZJ INNOLIGHT and Eoptolink Technology, what does this mean? Over the past few years, these companies have mainly benefited from the demand explosion for 800G and 1.6T pluggable optical modules, with revenue and profit growing at high speed continuously. CPO could change the product form of optical modules and even redistribute profits along the industry chain.
The biggest difference between CPO and traditional pluggable optical modules is that CPO moves the optical engine from the switch panel to near the switch chip, reducing power consumption and loss and increasing bandwidth density by shortening the electrical signal transmission distance. Especially against the backdrop of continuously expanding AI computing clusters, the power consumption of optical interconnects is becoming an important technical constraint. CPO's advantages in this regard are precisely why companies like NVIDIA and Broadcom are pushing its commercialization.
However, the ramp-up of CPO does not mean traditional pluggable optical modules will exit the market. On one hand, although CPO offers advantages in power consumption and bandwidth density, more complex packaging, fiber coupling, testing, and maintenance also bring new cost and reliability challenges. Even if related products have entered production, the cost and economics of large-scale deployment still require further validation.
On the other hand, traditional pluggable technology has not stopped upgrading. On October 12, NVIDIA and Cisco announced they would jointly promote a new high-density pluggable optical module form factor called USFP, supporting 400G per lane, with a future goal of increasing single-module bandwidth to 12.8 Tb/s (current mainstream 800G and 1.6T products have single-module bandwidths of 0.8 Tb/s and 1.6 Tb/s, respectively).
Meanwhile, NPO (Near-Package Optics) is also emerging as another important pathway. It places the optical engine close to the switch chip but does not co-package it with the chip, with power reduction falling between pluggable and CPO solutions, and offering more flexible maintenance.
This indicates that the industry has not bet on a single technology pathway. In the coming years, CPO, NPO, and pluggable optical modules are more likely to develop in parallel across different application scenarios, rather than one technology rapidly replacing others.
800G and 1.6T refer to speeds, answering how much data can be transmitted per second and solving the question of how fast. Pluggable, CPO, and NPO refer to packaging architectures, addressing how the optical engine and switch chip are positioned and how wiring is arranged. All three pathways will continue to upgrade in speed.
Therefore, we believe that CPO does increase uncertainty about the future competitive landscape of optical modules, but the accelerated development of CPO alone is not sufficient to conclude that the growth logic of existing leaders has fundamentally changed.
Compared to technology substitution, the other variable worthy of greater vigilance is the change in AI capital expenditure growth. While CPO affects the competitive landscape, capital expenditure directly determines demand itself.
Optical modules are a growth-cyclical business. The cyclicality comes from cloud providers' capital expenditure, data center construction, and customer inventory changes; the growth comes from continuous upgrades from 100G, 200G, 400G to 800G and 1.6T. The former determines whether demand fluctuates, while the latter continuously creates new product demand and value.
Over the past two years, AI computing infrastructure construction has driven both types of growth simultaneously. The expansion of GPU cluster scale has increased data transmission demand between servers and racks; at the same time, optical module speeds have been continuously upgraded from 400G to 800G and 1.6T, bringing product structure and value improvements.
However, recent forecasts of slowing AI investment growth by cloud providers have triggered market concerns about the sustainability of growth. According to relevant reports published by Morgan Stanley between late September and mid-October (including leasing investment), global major cloud service providers and emerging computing service providers' related investment is estimated at approximately $980 billion in 2026, approximately $1.58 trillion in 2027, and approximately $1.81 trillion in 2028; year-over-year growth rates will decline from about 100% to 60%, then to about 14%.
This is a slowdown in slope, not a peak in demand. But for the optical module industry, the investment growth rate of cloud providers is precisely the most noteworthy variable. It determines the pace of GPU cluster expansion and networking density, which in turn determines the demand elasticity for optical modules.
This is why the strong third-quarter reports did not directly trigger a stock price rebound. The Q3 reports reflect past operating results and cannot change the market's concerns about the growth slope two years out. The market is trading on growth expectations for 2027, 2028, and even further ahead.
This is the pricing characteristic of growth-cyclical stocks. When the industry is booming, earnings growth and valuation expansion can occur simultaneously; when the market begins to worry about the next capital expenditure cycle or competitive landscape changes brought by technological iteration, valuations may contract in advance even if current-period earnings are still growing.
Therefore, this round of adjustment does not mean industry demand has peaked; it merely means the market has begun to reassess the sustainability of future growth. CPO affects the long-term competitive landscape, while capital expenditure deceleration affects the near-term demand slope. With both pressures combined, the premium the market is willing to pay for long-term growth in optical modules will naturally narrow.
The key question is whether this round of adjustment has fully digested these concerns and when a rebound can occur.
Three Signals Still Needed for an Optical Module Rebound
After this round of adjustment, negative factors have indeed been released considerably. As of October 30, 2026, ZJ INNOLIGHT's PE-TTM (trailing twelve months price-to-earnings ratio) was 25.15 times, at the 18.99th percentile over the past five years; Eoptolink Technology's PE-TTM was 18.75 times, at the 14.78th percentile over the past five years; and T&S Communications' PE-TTM was 61.26 times, at the 60.28th percentile over the past five years.
From a forward valuation perspective, ZJ INNOLIGHT and Eoptolink Technology's expected PE for 2027 and 2028 have both fallen below 10 times, entering the valuation range of cyclical stocks; T&S Communications falls to 27 times for 2027 and further to 18 times for 2028, with the margin of safety at the valuation level accumulating.
Therefore, from a valuation standpoint, some leaders have entered a relatively attractive range, but a bottom still requires earnings expectations to stabilize. Moreover, cheap does not equal a safety cushion. If current mainstream 800G/1.6T product price declines exceed expectations, CPO substitution accelerates beyond expectations, or cloud providers slow procurement due to AI investment returns falling short of expectations, extend procurement cycles, or even cut 2027 and 2028 orders, any one of these materializing could turn order guidance into order cuts, triggering downward revisions in earnings forecasts, and current low valuations could be further revised down.
Meanwhile, when a rebound occurs cannot be judged by valuation alone. Going forward, three signals are truly worth tracking.
First, whether order and demand guidance can remain stable. Slowing AI investment growth does not necessarily mean optical module demand will decline. Cloud providers' investment includes not only optical modules but also spending on building data centers, power supply, cooling, and more. Therefore, as long as cloud providers' total investment continues to increase, GPU clusters continue to expand, and 800G and 1.6T products continue to upgrade, optical modules will still have incremental demand. Even if the overall growth rate slows, the proportion of optical modules in total capital expenditure may increase, giving orders a micro-level basis for remaining strong.
Conversely, if customers begin to cut procurement plans or order guidance is significantly lowered, it would signal a substantive weakening on the demand side, and the market could further revise down earnings expectations accordingly. Therefore, whether optical module manufacturers' order guidance for 2027 and 2028 remains stable and whether 800G and 1.6T shipments can continue to grow is the first signal worth tracking.
Second, next-generation products such as CPO and NPO need to move from technology deployment to volume delivery. For companies like ZJ INNOLIGHT and Eoptolink Technology, next-generation products such as CPO are both a threat and a potential new growth opportunity. The key is whether they can continue to secure customer orders in next-generation optical interconnect solutions and maintain customer share and profitability.
It is especially important to distinguish between samples, customer validation, order guidance, and volume delivery. Only when the latter two stages are truly realized will they contribute more directly to company performance.
Third, institutional earnings forecasts need to begin upgrading. This is the key to whether stock prices can move from bottoming to a sustained rebound. For growth-cyclical stocks, stock prices often reflect earnings expectations for the next one to two years in advance. If orders, product prices, or gross margins fall short of expectations and institutions lower their 2027 profit forecasts, then the currently low forward PE may be passively pushed higher.
Conversely, if subsequent orders remain stable, product ramps and gross margins do not deteriorate significantly, and institutions do not lower earnings forecasts or even begin to raise them, the market will have reason to reassess the company's fair valuation.
Among the three signals, orders determine demand, product delivery determines whether companies can generate new revenue, and earnings forecasts determine whether valuations can recover. What the sector truly needs is no longer grand industry stories, but concrete evidence supporting future earnings growth.
Which of the Big Three Will Present Opportunities First?
Based on the above, there is no need to be overly pessimistic about the optical module industry, as there may still be growth space ahead. But this does not mean all companies can maintain their past growth rates, nor does it mean all companies should enjoy the same valuation. During the bottoming phase, elasticity cannot be realized, so in stock selection, certainty should be weighted more heavily than elasticity.
Under these circumstances, among the three companies, if ranked by current earnings certainty, valuation level, and new product progress, I would prefer to observe ZJ INNOLIGHT first, followed by Eoptolink Technology, and then T&S Communications. Below, I will provide reasons in this order.
(1) ZJ INNOLIGHT: Relatively stronger earnings certainty. In the first three quarters of 2026, ZJ INNOLIGHT achieved revenue of 69.547 billion yuan, up 178.13% year-over-year; net profit attributable to shareholders was 22.238 billion yuan, up 211.81% year-over-year. 800G/1.6T pluggable products continued to ramp up, and 1.6T, being newly ramped with tight supply-demand dynamics, enjoyed new product pricing premiums, which was an important reason for both revenue and profit growth.
That is, ZJ INNOLIGHT's core advantage remains its customer base and large-scale delivery capability for high-speed pluggable optical modules. The next stage requires observing whether 1.6T products can continue to ramp up and whether next-generation products such as NPO and CPO can enter customer supply chains as planned. Management stated during an October 29 exchange that next-generation products are expected to gradually be released in the second half of 2027 and are expected to ramp significantly in 2028.
Additionally, ZJ INNOLIGHT's PE-TTM on October 30 was 25.15 times, at approximately the 19th historical percentile over the past five years; if 2027 earnings forecasts are realized, the expected PE could drop to 9.16 times, making the valuation somewhat attractive. ZJ INNOLIGHT's advantage is not betting on a particular new technology, but rather that its existing business still has a strong growth foundation while possessing the ability to extend into next-generation products. Therefore, if the sector experiences a fundamentals-driven rebound, it is more worthy of priority tracking.
(2) Eoptolink Technology: Lower forward valuation, rebound elasticity worth watching. Eoptolink Technology achieved revenue of 31.850 billion yuan in the first three quarters of 2026, up 92.97% year-over-year; net profit attributable to shareholders was 11.372 billion yuan, up 79.74% year-over-year. Performance was also solid.
However, although it also focuses on high-speed pluggable optical modules as its core business like ZJ INNOLIGHT, its 1.6T pluggable product ramp-up was slightly later, starting in Q2 2026 and only gradually ramping in Q3, about one quarter behind ZJ INNOLIGHT.
The next stage requires observing whether 1.6T pluggable product ramp-up can accelerate and whether NPO orders can transition from guidance to volume delivery. Management stated on October 29 that NPO progress is smooth and clear order guidance for 2027 has been seen.
Based on current market capitalization and institutional earnings forecasts, Eoptolink Technology's PE-TTM on October 30 was 18.75 times, at approximately the 15th historical valuation percentile; institutions expect net profit attributable to shareholders to grow about 84% in 2027, with an expected PE of approximately 7.44 times, lower than ZJ INNOLIGHT. If 1.6T ramp-up accelerates and NPO orders are realized in the future, it may have greater valuation recovery elasticity. But low valuation does not mean no risk; the ramp-up pace, pricing, and gross margins of 1.6T pluggable and NPO products remain key to whether earnings forecasts can be realized.
(3) T&S Communications: Greater CPO theme elasticity, but relatively insufficient valuation margin of safety. T&S Communications achieved revenue of 4.751 billion yuan in the first three quarters of 2026, up 21.25% year-over-year; net profit attributable to shareholders was 2.083 billion yuan, up 42.18% year-over-year. Revenue and profit growth rates are the lowest among the three, mainly due to different business models. Unlike the first two, T&S Communications mainly provides optical components and does not manufacture complete optical module units.
As CPO and NPO increase optical engine integration, components such as FAU fiber arrays and ELS external light sources are expected to see new demand. Company management stated that CPO supporting products will gradually increase volume from 2027, and NPO supporting products, if validation and customer introduction proceed smoothly, are also expected to achieve large-scale production in 2027.
But from a valuation perspective, T&S Communications' expected PE for 2027 is approximately 27.28 times, with a current PE-TTM of 61.26 times and a historical valuation percentile of approximately 60%. Even after a significant correction, its valuation digestion is still less thorough than the other two. If the market re-trades CPO industrialization, T&S Communications may gain better theme elasticity; but for a sustained rebound supported by earnings, CPO and NPO component orders need to truly translate into revenue and profit. Current valuations still demand a lot from future growth.
Overall, each of the three companies has its own highlights, and their rebound conditions differ. ZJ INNOLIGHT has relatively stronger certainty in existing business growth, making it suitable as a priority observation target; Eoptolink Technology has a lower forward valuation, and if product ramp-up accelerates, its valuation recovery elasticity is worth watching; T&S Communications has greater potential in CPO and NPO, but needs more earnings realization to justify its valuation.
Therefore, although this round of adjustment has already digested a considerable portion of concerns, the market's worries about future growth have not completely disappeared, and it will be difficult for the sector to return to its previous uptrend in the short term. Going forward, whether orders can maintain growth, whether next-generation products such as CPO can be delivered in volume, and whether institutional earnings forecasts can stabilize will determine whether the sector can move from L-shaped bottoming to a rebound.
Disclaimer: The content of this article is for reference only. The information or opinions expressed herein do not constitute any investment advice. Readers should make investment decisions cautiously.
Comments