CCL/PCB Update: High-End Copper-Clad Laminate Shortage Extends to 2027; M9 Materials and HDI Capacity Continue to Reprice
本内容基于公开资料和研报数据整理，不构成任何投资建议，不代表任何个人观点，仅供学习参考，请理性阅读
BofA’s report pushes the pricing-upcycle logic for high-end CCL/PCB out by another year: shortages may last until end-2027 or early 2028. The key bottlenecks are shifting from strong demand to M9 materials, HDI yields, equipment lead times, and the speed at which PCB makers can pass through costs. The next variables to track are yield, delivery, and pricing pass-through.
The most useful part of this report is that it breaks down “AI servers driving PCB price increases” into more specific supply constraints. BofA believes high-end CCL/PCB supply-demand tightness will continue until end-2027 or early 2028, for three reasons: AI servers continue to migrate toward HDI and higher-layer-count PCBs; yields and output for new material solutions remain unstable; and although PCB makers are increasing capex, capacity that can deliver stably is coming online slowly. The key point in this chain is not the familiar statement that “demand remains strong,” but that high-end materials, drilling, lamination, certification, and equipment lead times together determine the pace of supply.
CCL is more likely than PCB to extend the pricing cycle. BofA’s reasoning is straightforward: there are only 5-6 major suppliers of high-end CCL, and industry concentration is higher; suppliers are also more disciplined in expanding capacity, while lead times for key equipment suppliers are relatively long. This gives CCL makers more cushion when demand fluctuates. PCB makers have greater operating flexibility, but they are also more vulnerable to margin pressure from upstream material price increases. For PCB makers, the next two questions are therefore: whether orders are there, and whether they can pass CCL price increases through to customers smoothly.
The material roadmap is the first key divergence in this report. BofA is positive on Elite Material’s (EMC, 2383.TT) M9 quartz-based CCL, because it is superior to low-Dk2 solutions in electrical performance such as Dk/Df. In plain terms, the higher the transmission speed of AI server boards, the stricter the requirements for material loss, stability, and consistency. Commodity materials can benefit from price increases first, but the suppliers that capture high-end share will still be those that can deliver performance, yield, and mass-production cadence together.
The challenges on the PCB side are also rising, but the solution path is relatively clear. The report notes that PCB makers can improve yields and output for new products by using more laser drilling instead of mechanical drilling, and by replacing traditional high-layer-board processes with HDI technology. This view matters for high-end PCB makers such as Victory Giant Technology, WUS Printed Circuit, Shennan Circuits, and Dynamic Holding: over the next one to two years, the profit gap may come from “who can stabilize high-spec boards first.” Capacity announcements are only the starting point.
AI PCB and CCL Deep Dive: From GB300 to Rubin Ultra, Who Runs Short First in High-Layer Boards, M9 Materials, and Electronic Glass Fabric
Elite Material is the most obvious beneficiary of this round of estimate increases. BofA raised its 2026-2028 EPS forecasts for Elite Material by 12%-16%, and lifted its target price from NT$6600 to NT$7800, still using 34x 2H27-1H28E P/E. The increase is mainly driven by stronger revenue and gross margin after 2Q26, not by simply raising the multiple. BofA now expects Elite Material’s 2026-2028 revenue to reach NT$189.8bn, NT$295.4bn, and NT$416.8bn, respectively, with EPS of NT$107.45, NT$187.39, and NT$269.95. Its 2026-2028 EPS forecasts are about 6%-15% above market consensus. Beyond the valuation multiple, sell-side disagreement is also concentrated on the growth slope.
Taiwan Union Technology (TUC, 6274.TT) looks more like a case of the valuation window being pushed further out. BofA did not change its earnings forecasts, but raised the target price from NT$2250 to NT$2600, because the high-end CCL shortage cycle is becoming longer. The valuation base rolls from 2Q27-1Q28E to 2H27-1H28E, while the target multiple remains 30.5x. The implication is simple: if the shortage lasts until end-2027, the market is willing to look six months further into the earnings peak, so the valuation window can naturally roll forward.
Dynamic Holding (3715.TT) represents both the opportunity and pressure for PCB makers. BofA remains positive on its AI server PCB exposure, with a target price of NT$245, and expects AI server PCB revenue contribution to rise to 50% and 60% in 2027 and 2028. But the report also notes that AI project ramp-up may not arrive until 4Q26, while 3Q26 gross margin will first be affected by CCL price increases, with cost pass-through starting from 4Q26 at the earliest. This timing is critical for PCB makers: price increases happen first at the materials end, while PCB makers must wait for customer repricing, leaving a period of income-statement pressure in between.
Shengyi Technology (600183.SH) is the sharper point of disagreement. BofA raised its target price from RMB71 to RMB83, shifting the valuation base to 2027E and 24x P/E, and acknowledged that the high-end CCL shortage will benefit the company. But the rating remains Underperform, because valuation is already elevated, while large-scale adoption of pure PTFE PP/CCL in AI server systems is still some way off. In other words, Shengyi Technology can benefit from industry price increases and tight high-end CCL supply, but BofA is unwilling to fully pull forward share realization from the PTFE route.
Shengyi Technology Deep-Dive Update: AI CCL Enters Nvidia Share-Realization Phase; High-End Copper-Clad Laminates Move from Price Increases to Delivery
For the A-share and Hong Kong-listed materials chain, the main lesson from this report is to split “price increases” into three layers. The first layer is price increases in upstream materials such as electronic glass fabric, copper foil, resin, and CCL. The second layer is whether CCL makers can convert high-end materials into stable shipments and share gains. The third layer is whether PCB makers can pass material price increases through to end customers. The first two layers benefit more directly; the third requires tracking customer certification, order visibility, and repricing cadence. From an investment perspective, it is not enough to look at who is closest to AI servers. Investors also need to assess who sits closer to the bottleneck in the pricing chain.
Kingboard Laminates Deep-Dive Update: Citi Raises Target Price Seven Times in Two Months; July CCL Price Up Another 15%, Copper-Foil Profits Expanding
The four most important variables to track next are: first, customer validation progress for M9, quartz-based materials, and low-Dk2 solutions, and who can secure stable mass-production share first; second, equipment lead times and new capacity release for high-end CCL, and whether supply discipline is being disrupted; third, whether PCB makers can pass CCL price increases through to AI server customers after 4Q26; and fourth, whether the valuation of A-share materials leaders such as Shengyi Technology has already reflected PTFE and high-end CCL share in advance. As long as these four variables do not loosen at the same time, the high-end CCL/PCB trade will continue to revolve around supply bottlenecks.
Victory Giant Technology Update: Thailand Plant, Overseas Capacity, and AI PCB Effective-Capacity ValidationCCL/PCB Update: High-End Copper-Clad Laminate Shortage Extends to 2027; M9 Materials and HDI Capacity Continue to Reprice
本内容基于公开资料和研报数据整理，不构成任何投资建议，不代表任何个人观点，仅供学习参考，请理性阅读
BofA’s report pushes the pricing-upcycle logic for high-end CCL/PCB out by another year: shortages may last until end-2027 or early 2028. The key bottlenecks are shifting from strong demand to M9 materials, HDI yields, equipment lead times, and the speed at which PCB makers can pass through costs. The next variables to track are yield, delivery, and pricing pass-through.
The most useful part of this report is that it breaks down “AI servers driving PCB price increases” into more specific supply constraints. BofA believes high-end CCL/PCB supply-demand tightness will continue until end-2027 or early 2028, for three reasons: AI servers continue to migrate toward HDI and higher-layer-count PCBs; yields and output for new material solutions remain unstable; and although PCB makers are increasing capex, capacity that can deliver stably is coming online slowly. The key point in this chain is not the familiar statement that “demand remains strong,” but that high-end materials, drilling, lamination, certification, and equipment lead times together determine the pace of supply.
CCL is more likely than PCB to extend the pricing cycle. BofA’s reasoning is straightforward: there are only 5-6 major suppliers of high-end CCL, and industry concentration is higher; suppliers are also more disciplined in expanding capacity, while lead times for key equipment suppliers are relatively long. This gives CCL makers more cushion when demand fluctuates. PCB makers have greater operating flexibility, but they are also more vulnerable to margin pressure from upstream material price increases. For PCB makers, the next two questions are therefore: whether orders are there, and whether they can pass CCL price increases through to customers smoothly.
The material roadmap is the first key divergence in this report. BofA is positive on Elite Material’s (EMC, 2383.TT) M9 quartz-based CCL, because it is superior to low-Dk2 solutions in electrical performance such as Dk/Df. In plain terms, the higher the transmission speed of AI server boards, the stricter the requirements for material loss, stability, and consistency. Commodity materials can benefit from price increases first, but the suppliers that capture high-end share will still be those that can deliver performance, yield, and mass-production cadence together.
The challenges on the PCB side are also rising, but the solution path is relatively clear. The report notes that PCB makers can improve yields and output for new products by using more laser drilling instead of mechanical drilling, and by replacing traditional high-layer-board processes with HDI technology. This view matters for high-end PCB makers such as Victory Giant Technology, WUS Printed Circuit, Shennan Circuits, and Dynamic Holding: over the next one to two years, the profit gap may come from “who can stabilize high-spec boards first.” Capacity announcements are only the starting point.
AI PCB and CCL Deep Dive: From GB300 to Rubin Ultra, Who Runs Short First in High-Layer Boards, M9 Materials, and Electronic Glass Fabric
Elite Material is the most obvious beneficiary of this round of estimate increases. BofA raised its 2026-2028 EPS forecasts for Elite Material by 12%-16%, and lifted its target price from NT$6600 to NT$7800, still using 34x 2H27-1H28E P/E. The increase is mainly driven by stronger revenue and gross margin after 2Q26, not by simply raising the multiple. BofA now expects Elite Material’s 2026-2028 revenue to reach NT$189.8bn, NT$295.4bn, and NT$416.8bn, respectively, with EPS of NT$107.45, NT$187.39, and NT$269.95. Its 2026-2028 EPS forecasts are about 6%-15% above market consensus. Beyond the valuation multiple, sell-side disagreement is also concentrated on the growth slope.
Taiwan Union Technology (TUC, 6274.TT) looks more like a case of the valuation window being pushed further out. BofA did not change its earnings forecasts, but raised the target price from NT$2250 to NT$2600, because the high-end CCL shortage cycle is becoming longer. The valuation base rolls from 2Q27-1Q28E to 2H27-1H28E, while the target multiple remains 30.5x. The implication is simple: if the shortage lasts until end-2027, the market is willing to look six months further into the earnings peak, so the valuation window can naturally roll forward.
Dynamic Holding (3715.TT) represents both the opportunity and pressure for PCB makers. BofA remains positive on its AI server PCB exposure, with a target price of NT$245, and expects AI server PCB revenue contribution to rise to 50% and 60% in 2027 and 2028. But the report also notes that AI project ramp-up may not arrive until 4Q26, while 3Q26 gross margin will first be affected by CCL price increases, with cost pass-through starting from 4Q26 at the earliest. This timing is critical for PCB makers: price increases happen first at the materials end, while PCB makers must wait for customer repricing, leaving a period of income-statement pressure in between.
Shengyi Technology (600183.SH) is the sharper point of disagreement. BofA raised its target price from RMB71 to RMB83, shifting the valuation base to 2027E and 24x P/E, and acknowledged that the high-end CCL shortage will benefit the company. But the rating remains Underperform, because valuation is already elevated, while large-scale adoption of pure PTFE PP/CCL in AI server systems is still some way off. In other words, Shengyi Technology can benefit from industry price increases and tight high-end CCL supply, but BofA is unwilling to fully pull forward share realization from the PTFE route.
Shengyi Technology Deep-Dive Update: AI CCL Enters Nvidia Share-Realization Phase; High-End Copper-Clad Laminates Move from Price Increases to Delivery
For the A-share and Hong Kong-listed materials chain, the main lesson from this report is to split “price increases” into three layers. The first layer is price increases in upstream materials such as electronic glass fabric, copper foil, resin, and CCL. The second layer is whether CCL makers can convert high-end materials into stable shipments and share gains. The third layer is whether PCB makers can pass material price increases through to end customers. The first two layers benefit more directly; the third requires tracking customer certification, order visibility, and repricing cadence. From an investment perspective, it is not enough to look at who is closest to AI servers. Investors also need to assess who sits closer to the bottleneck in the pricing chain.
Kingboard Laminates Deep-Dive Update: Citi Raises Target Price Seven Times in Two Months; July CCL Price Up Another 15%, Copper-Foil Profits Expanding
The four most important variables to track next are: first, customer validation progress for M9, quartz-based materials, and low-Dk2 solutions, and who can secure stable mass-production share first; second, equipment lead times and new capacity release for high-end CCL, and whether supply discipline is being disrupted; third, whether PCB makers can pass CCL price increases through to AI server customers after 4Q26; and fourth, whether the valuation of A-share materials leaders such as Shengyi Technology has already reflected PTFE and high-end CCL share in advance. As long as these four variables do not loosen at the same time, the high-end CCL/PCB trade will continue to revolve around supply bottlenecks.
Victory Giant Technology Update: Thailand Plant, Overseas Capacity, and AI PCB Effective-Capacity Validation