Optical fiber stocks are staging a frantic catch-up rally.

Fiber optic stocks rallied collectively today. Yangtze Optical Fibre and Cable hit the 10% daily limit immediately. Hengtong Optic-Electric's intraday gain exceeded 8%, with its trading volume breaking 100 billion yuan, ranking first across the whole market. ZTT rose by nearly 6%, and Hangdian Co., Ltd. notched its second consecutive daily limit.

Some attribute this rally to the outstanding semi-annual performance released by Hengtong Optic-Electric the previous night, but the story is far from that simple.

If we combine Hengtong Optic-Electric's financial data with Zhang Jianping's move of buying stakes in two fiber optic companies in the second quarter, we will find that the market is re-evaluating an industry that has been overlooked for many years.

In the first half of the year, Hengtong Optic-Electric posted operating revenue of 42.026 billion yuan, up 31.13% year on year. Its net profit attributable to shareholders stood at 3.12 billion yuan, a 93.38% year-on-year increase, and its non-recurring profit-deducted net profit rose by more than 100%.

A notable figure is that the operating revenue of the company's optical communication business increased by over 130% year on year, with its gross margin exceeding 60%.

In the manufacturing sector, a gross margin above 60% is usually seen in fields such as branded Baijiu, innovative pharmaceuticals, or high-end chip design. Few people would associate such a profitability level with a company that produces optical fibers and cables.

Over the past two decades, the gross margin of China's domestic optical fiber and cable industry has long fluctuated between 15% and 25%. During the overcapacity cycle from 2018 to 2020, the gross margin of some enterprises even fell to single digits. The market has long regarded the optical fiber and cable sector as a capital-heavy, highly cyclical, low-value-added infrastructure industry, whose valuation has long been benchmarked against steel and cement. A price-to-earnings ratio of 10 times is considered high for the sector.

Now Hengtong Optic-Electric's optical communication business has achieved a gross margin of over 60%, a figure that actually indicates the underlying logic of the industry has fundamentally changed.

Why has such a huge change taken place? It all traces back to the brutal industry reshuffle from 2018 to 2022.

Back then, the optical fiber and cable industry suffered from severe overcapacity, and the price war drove many small and medium-sized enterprises out of business. A large number of manufacturers exited the market, and industry concentration rose rapidly.

After 2024, the demand side is driven by three forces at the same time: massive fiber interconnection is required inside AI data centers, operators are accelerating the construction of computing power networks, and the fiber-to-the-home (FTTH) construction in overseas markets has entered a period of accelerated growth.

With shrinking supply and expanding demand, the market has naturally shifted from a buyer's market to a seller's market.

There is another detail here: the capacity expansion cycle of optical fibers and cables is very long. It usually takes 18 to 24 months from planning and construction to the formation of effective production capacity.

Hardly any enterprise dared to invest in new production capacity against the trend during the industry trough, which means the current tight supply situation will last for at least another two years. The gross margin's rise from 20% to 60% is not a short-term price fluctuation, but a concentrated reflection of the mismatch between supply and demand structure.

Therefore, when we look at Zhang Jianping's operations in the second quarter, things become more interesting.

He became the fifth largest shareholder of Hengtong Optic-Electric and the third largest shareholder of Hangdian Co., Ltd. as a new investor. Most people's first reaction is that the top hot money chases the market trend. But if we observe these two targets in the coordinates of the industrial chain, we will find that this is not a simple short-term move.

Hengtong Optic-Electric is a core midstream enterprise in the optical fiber and cable industrial chain, while Hangdian Co., Ltd. produces wires, cables and optical fiber composite products, and also undertakes part of the supply for upstream supporting links. Heavy positions in both the midstream and upstream of the industrial chain essentially bet on the logic of profit redistribution across the entire industrial chain.

When the industry shifts from a buyer's market to a seller's market, the bargaining power of upstream supporting enterprises usually improves later, but their profit elasticity may be more dramatic, because they were suppressed more severely in pricing in the past, leaving greater room for profit recovery. Zhang Jianping's real-money investment expresses the view that optical communication is not a separate story of a single company, but an inflection point emerging for the entire industrial chain.

And this leads to a greater cognitive bias.

Over the past two years, the market's main investment line for AI computing power has been highly concentrated, with capital almost all pouring into the GPU, server, optical module and CPO tracks. Optical fibers and cables are classified as traditional infrastructure, and almost completely excluded from the AI narrative.

But there is a fundamental fact that cannot be avoided: no matter how strong the GPU performance is, how high the switch rate is, or how advanced the optical module solution is, data ultimately needs to be physically transmitted through optical fibers. Optical modules are responsible for photoelectric conversion, while optical fibers are responsible for carrying signals. The two are equally important in AI infrastructure. The market has revalued optical module enterprises, with the market value of Innolight and Eoptolink increasing by 5 to 10 times, but it has not revalued optical fiber enterprises synchronously.

The reason is that most people take it for granted that optical fibers are standardized products with transparent prices and low technical content, which do not deserve a valuation premium.

This default premise has been completely broken by three structural changes from 2025 to 2026.

The first change is the jump in demand for optical fibers from AI data centers.

The cabling volume of a traditional data center is usually at the level of tens of thousands of core-kilometers, while the cabling volume of a large AI training cluster can reach hundreds of thousands or even millions of core-kilometers.

Because parallel computing between GPUs requires extremely low latency and ultra-high density interconnection, the optical fiber cabling density is 5 to 10 times higher than that of traditional architectures.

This is not a gradual growth, but a leap in order of magnitude.

The second change comes from overseas markets.

The FTTH construction in Europe, the Middle East and Southeast Asia has accelerated significantly in 2025. Overseas fiber prices are generally 30% to 80% higher than domestic prices. A considerable part of the 130% year-on-year revenue growth of Hengtong Optic-Electric's optical communication business in the first half of the year came from high-priced overseas orders.

Overseas orders not only contribute to revenue scale, but also directly drive up the overall gross margin.

The third change is that the product structure itself has begun to stratify.

New types of low-loss, large-effective-area optical fibers such as G.654.E are the standard configuration for 400G and 800G long-distance transmission, with much higher technical barriers than traditional G.652.D products. Only a few enterprises around the world can supply them stably.

Optical fibers are no longer undifferentiated standardized commodities. High-end products begin to enjoy technical premium, and clear price stratification has emerged within the industry.

The combination of these three changes means that the optical fiber and cable industry is undergoing a revaluation of its pricing system.

If the strong performance only comes from the better-than-expected semi-annual report, it can be regarded as a nice performance-driven rally that will end after the price goes up. But if the market begins to price the industry with a new framework, there may be much more room for growth ahead.

At present, the A-share market still values leading fiber optic enterprises under the framework of cyclical stocks. Even after this round of rally, Hengtong Optic-Electric's TTM price-to-earnings ratio is just over 20 times, ZTT's is less than 15 times, and Yangtze Optical Fibre and Cable's is slightly higher. For an industry whose gross margin has jumped from 20% to 60%, the migration of its valuation center from 10-15 times to 25-35 times is itself a correction of the pricing logic.

The market is used to labeling industries with past experience, but when profit margin and sustainability have undergone fundamental changes, the old labels will gradually fail.

Of course, any judgment needs to take risks into account. A considerable part of the current high gross margin comes from the rising centralized procurement prices of operators and the increasing proportion of high-priced overseas orders.

If leading enterprises launch capacity expansion one after another in the next two years, while domestic demand growth slows down, a new round of supply-demand rebalancing may emerge in the second half of 2027, and fiber prices will face downward pressure. But that is a situation two years later, and the capital market rarely waits that long to start pricing.

The more critical variable is the trend of operators' centralized procurement prices. If the centralized procurement prices remain high or even continue to rise in 2027, the sustainability of this round of revaluation will far exceed most people's expectations.

This article is from the WeChat public account "Dongzhen Strategy", author: Dongzhen Strategy, published with authorization from 36Kr.