The fiberglass sector in the A-share market experienced a collective surge. On February 11, International Composite Materials quickly hit the 20% daily limit-up, with Honghe Technology, Sinoma Science & Technology, Shandong Fiberglass, and several other stocks also reaching the limit-up. Over a longer period, the Wind Fiberglass Index has accumulated gains exceeding 74% since December 1, 2025.
Electronic fabric refers to the general term for electronic-grade glass fiber cloth used in the electronics industry, serving as a critical base material for printed circuit boards (PCBs). High-end electronic fabric is an indispensable core structural material for high-performance chip substrates and PCBs. Driven by explosive growth in demand for AI chips, the high-end electronic fabric market is facing increasingly significant supply constraints.
The immediate catalyst for this market surge stems from recent price hikes for electronic fabric. Supply of high-end electronic glass fiber cloth produced by Nitto Boseki is becoming increasingly tight, with market conditions even described as a situation where "fabric is hard to come by." To ensure stable supply of core raw materials, NVIDIA CEO Jensen Huang recently traveled to Japan to negotiate cooperation with Nitto Boseki, while Apple is also actively communicating with relevant Japanese government departments seeking supply chain support.
On February 4, leading companies like International Composite Materials and Guangyuan New Materials raised prices for electronic fabric, with standard electronic fabric prices increasing over 10%, breaking through the 10,000 yuan per ton threshold. This marks the fourth industry-wide price increase since early 2025, and this hike features the largest magnitude and shortest cycle, reflecting tight supply-demand dynamics for electronic yarn and fabric overall. Quotations for standard electronic yarn have reached 10,300-10,700 yuan per ton, while the mainstream price for 7628 electronic fabric has risen to 5.1-5.5 yuan per meter.
According to Huatai Securities calculations, this price increase brings approximately 0.5 yuan per meter in additional net profit, and is expected to contribute net profit increments of approximately 440 million yuan, 110 million yuan, and 110 million yuan to China Jushi, International Composite Materials, and Honghe Technology, respectively. These increments represent about 13%, 34%, and 50% of the three companies' 2025 net profits attributable to shareholders.
Behind this round of price increases lies a combination of supply constraints and demand growth. Demand for electronic fabric continues to rise, driven by the construction of computing infrastructure such as AI servers and high-speed network equipment. On the supply side, limited capacity expansion exists due to high investment costs and strong technical barriers. The industry has entered a new boom cycle, and the tight supply of electronic fabric may impose new supply chain constraints on the development of computing infrastructure.
Supply increments are limited, and rising costs are inhibiting capacity expansion. According to industry data forecasts, the net supply increase for the standard electronic yarn market in 2026 may be below 10%. SCI99 statistics show that from 2023 to the first half of 2025, aside from minor capacity replacements, no new production lines were commissioned in the standard electronic yarn sector. The main new capacity in 2026 will come from China Jushi's Huai'an project (Phase I with 50,000 tons expected to commence in Q1, Phase II with 50,000 tons planned for H2) and Kingboard's Qingyuan Line 7 (70,000 tons expected in H2). However, considering that two furnaces at International Composite Materials will undergo cold repairs during the same period, the actual net capacity increase as a percentage of current operating capacity is projected to be less than 10%.
The supply side also faces dual constraints of structural shifts and rising costs. Driven by tight supply-demand and higher profit margins for high-end electronic fabric, some standard electronic yarn production lines have been converted to produce higher-end products, further squeezing the supply of standard yarn. Concurrently, significant price increases for upstream precious metals since 2026 have pushed up costs. Platinum-rhodium alloy, a key material for bushings, has seen price climbs that increase both furnace investment and daily operating costs. Huatai Securities analysis indicates that precious metal bushings account for nearly 40% of the total investment in an electronic yarn production line, and this cost pressure may further inhibit industry capacity expansion.
Capacity expansion is also constrained by extremely high capital and technical thresholds. The basic investment for a single electronic yarn furnace exceeds 500 million yuan, with investment for high-end production lines even surpassing 1.5 billion yuan. The timeline from construction to production can extend beyond two years, and equipment investment for a single standard production line often exceeds 500 million yuan. Such large-scale, long-cycle, irreversible heavy-asset investments pose substantial barriers for new industry entrants.
AI computing upgrades are driving demand growth. Market demand is experiencing a structural recovery. As the core base material in the copper-clad laminate (CCL) and printed circuit board (PCB) supply chain, end-demand for electronic yarn/fabric is supported by an orderly recovery in traditional sectors like home appliances, new energy vehicles, and consumer electronics, spurred by policies like "replace old with new" and new national subsidies. Demand for standard electronic fabric is expected to steadily increase accordingly.
A more significant driving force comes from hardware upgrades in high-performance computing, represented by AI servers. As AI servers evolve from traditional CPU architectures to GPU cluster architectures, their PCB layers have generally increased to over 20 layers, demanding higher performance from base materials.
Taking the NVIDIA GB300 server as an example, its PCB layers have increased to over 16 layers, with quartz electronic fabric (Q-cloth) usage per server reaching 18 to 24 meters—approximately five times the usage in traditional servers. Furthermore, the evolution of next-generation high-speed optical modules (1.6T/3.2T) and switches (224G ports) also relies on high-performance electronic fabric to ensure signal integrity and stability.
At the material performance level, the dielectric properties of electronic fabric directly impact signal transmission efficiency. A 10% reduction in dielectric constant can enable a doubling of transmission speed. Third-generation quartz electronic fabric (Q-cloth) has reduced the dielectric constant to the 2.2–2.3 range, with a dielectric loss factor controlled between 0.001–0.003, and heat resistance exceeding 600°C. Such materials can significantly reduce attenuation and distortion in high-frequency signal transmission, providing a foundation for higher speed and more stable signal transmission in AI computing equipment.
Breakthroughs in the high-end market are accelerating domestically. Japanese companies still dominate the high-end electronic fabric market. Nitto Boseki, Asahi Kasei, AGC, and other Japanese firms collectively control nearly 70% of the global market share for high-end electronic fabric. Among them, Nitto Boseki, leveraging long-term accumulation in core glass material formulations like NE-glass and T-glass, has consistently invested in R&D since the 1990s. With over thirty years of technical沉淀 (sedimentation/accumulation), it holds a first-mover advantage in the current industry boom.
Citing analysis by Professor Zeng Huidan from East China University of Science and Technology, media reports indicate that Nitto Boseki holds a significant advantage in the second-generation low-dielectric (LDK/LCTE) electronic fabric field, having achieved stable mass production for forty years with a yield rate maintained above 95%. Although domestic companies have achieved technological breakthroughs in recent years, their mass production yield rates generally range from 70% to 80%, with production capacity scale less than 20% of Nitto Boseki's. She pointed out that while China already ranks among the global leaders in the mid-to-low-end electronic fabric market, it remains in a transitional phase from technological breakthrough to scaled mass production for the highest-end products.
Domestication efforts are accelerating. In 2021, Honghe Technology took the lead in achieving mass production of 9-micron ultra-thin electronic fabric, breaking foreign monopolies. In the same year, Henan Guangyuan accomplished mass production of low-dielectric fabric, becoming the first domestic company to break through in this category. In 2024, Taishan Fiberglass achieved mass production of second-generation low-dielectric fabric. In the more technically challenging quartz electronic fabric (Q-cloth) sector, after eight years of R&D, PhiChem successfully launched its M9-grade Q-cloth product in 2025 and obtained official NVIDIA certification.
Tianfeng Securities analysis highlights three main barriers in the Q-cloth industry: firstly, the complex wire drawing process, as quartz fibers are brittle and furnace temperatures need to exceed 2000°C; secondly, reliance on imports for core equipment like looms, which have long delivery cycles; and thirdly, strict customer certification processes, typically requiring 2-3 years. High barriers correspond to high value-added: first-generation fabric costs about 30 yuan per meter, second-generation about 120 yuan per meter, while Q-cloth prices can reach 200-400 yuan per meter, with gross margins exceeding 60%.
Looking at inventory levels, as of September 2025, the inventory balances for China Jushi, International Composite Materials, and Honghe Technology were 3.72 billion yuan, 1.92 billion yuan, and 170 million yuan, respectively. These figures represent 82%, 89%, and 77% of their respective inventory peaks since 2021, indicating that overall industry inventory is at healthy levels, providing room for price transmission.
Looking ahead, Professor Zeng Huidan believes that breakthroughs in high-end capacity before 2027 face challenges, but the long-term trend towards domestication is clear. China has achieved breakthroughs from 0 to 1 in key technologies. As mass production capabilities improve and customer cooperation deepens, it is expected that within the next five to six years, Chinese companies will achieve critical breakthroughs in some high-end niche markets and capture significant market share.
Comments