Goldman Sachs expects a sustained shortage of high-end copper foil for the next three years, initiating a 'Buy' rating for Jinju Development

AI server replacement is pushing PCB copper foil from a low-margin material to a high-end supply bottleneck. The latest report from Goldman Sachs estimates that the addressable market size for high-end copper foil HVLP3 and above will grow rapidly from $216 million in 2025 to $2.4 billion in 2028, with a compound annual growth rate of 122%.

As the requirements for signal integrity continue to rise with AI servers and high-end Ethernet switches, Goldman Sachs estimates that the monthly demand for copper foil HVLP3 and above will surge from 679 tons in 2025 to 5,206 tons in 2028, with a compound annual growth rate of 97%.

AI servers and switches upgrade drive demand

AI servers are the core engine for the explosion in demand for high-end copper foil. Goldman Sachs estimates that from 2025 to 2028, the demand for copper foil HVLP3 and above in AI servers will grow more than sixfold, propelling it quickly into a stage of large-scale application from marginal specifications.

High-end switch upgrades further accelerate the demand slope: 400G switches have been using M7-grade copper-clad laminates equipped with HVLP3 since 2023; 800G switches have switched to M8-grade copper-clad laminates with HVLP3 and HVLP4 since the second half of 2024; and the 1.6T switches, expected to be mass-produced in the second half of 2026, will use M9-grade copper-clad laminates with HVLP4 and HVLP5.

In this trend, Goldman Sachs estimates that the proportion of HVLP3 in the total demand for high-end copper foil will drop significantly from 76% in 2025 to 30% in 2028, while the penetration rate of HVLP4 will rise from 24% to 46%, and HVLP5 is expected to contribute significantly to demand starting from 2027.

Yield bottlenecks limit effective capacity severely

Although the industry's nominal capacity will increase from 1,057 tons per month in 2025 to 4,977 tons per month in 2028, the report believes that, affected by yield rates and line switching losses, effective capacity can only grow from 803 tons per month to 3,759 tons per month.

Goldman Sachs points out that the yield rates for major suppliers' products at HVLP3 and above generally remain at 70%-80%, and line switching can cause an additional 10%-20% capacity loss. Based on effective capacity, the industry's supply gap in the years 2026-2028 will reach 28%, 39%, and 38%, respectively, significantly higher than the moderate imbalance indicated by nominal capacity, and the supply tightness situation will intensify markedly.

Pricing power for high-end copper foiland significant increase in profit margins

Goldman Sachs believes that unlike low-end copper foil, which has long adopted a "copper price + processing fee" pricing model, high-end copper foil HVLP3 and above, due to the extremely limited number of suppliers capable of stable mass production, has seen pricing power increasingly determined by supply and demand relationships.

This structural change is even more evident in profit margins: the average price of high-end HVLP copper foil is more than double that of traditional HTE copper foil, with gross margins generally between 40%-60%, while the gross margin for HTE copper foil is only 0%-10%.

Goldman Sachs assigns "Buy" rating to Jinju Development for the first time

Among the suppliers, Mitsui Kinzoku continues to maintain the industry's leading position, but its capacity cannot fully cover global customer demand. Goldman Sachs estimates that the company's effective capacity for HVLP3 and above will be 718 tons, 870 tons, and 1,140 tons per month for the years 2026-2028, respectively.

This opens up significant space for the second supplier. Goldman Sachs covers the Taiwanese copper foil manufacturer Jinju Development for the first time, assigning a "Buy" rating, with a 12-month target price of NT$