Caitong: AI servers and automotive electronics drive increasing demand for MLCCs; domestic replacements and K-shape cycle recovery in the wave of high-end transition.

Caitong released a research report stating that AI servers and automotive electronics are becoming structural incremental sources for MLCC. The domestic substitution path is expected to evolve step by step along the lines of "mature specification substitution in consumer electronics, entry into the domestic automotive electronic supply chain, cooperation of AI servers with domestic computing power chain to achieve from 1 to N" evolution. Recommended attention to domestic MLCC industry chain enterprises with core material/process/equipment barriers, deep integration with AI servers and automotive electronics in high prosperity downstream, and products that have entered the customer introduction or volume phase. Caitong's main points are as follows: The demand side logic is reconstructed, with MLCC shifting from consumer electronics cyclical products to hard technology growth products such as AI computing power and automotive electronics. According to Dataintelo, the global MLCC market size is expected to grow from approximately $14.8 billion in 2025 to approximately $28.6 billion in 2034, at a CAGR of approximately 7.6%. AI servers and automotive electronics have become structural incremental sources. AI servers have evolved from single-machine motherboards to rack-level high-density computing platforms, with MLCC usage approximately 8-10 times that of traditional servers. Murata expects the average capacitance load of AI server computing boards to increase from 10,000-20,000 pieces to 15,000-25,000 pieces; the electrification, intelligence, 800V high-voltage platform, and advanced driving of automotive electronics continue to increase the single-vehicle MLCC usage and reliability requirements. The supply-side high-end capacity rigid constraints are prominent, overseas giants' strategic shift opens the window for domestic substitution. High-end MLCC barriers are reflected in material formulation, ultra-thin laminating, thousand-layer stacking, co-firing consistency, and automotive/server-level customer validation. The China Electronic Components Industry Association and the prospectus of Chengdu Hongming Electronics disclosed that by 2024, the global MLCC market CR5 reached 77.3%, with Murata and Samsung Electric Motors accounting for 31.8% and 22.9% respectively. Headquartered in Japan and South Korea still dominate the high-end AI servers, automotive specifications, and ultra-miniature products; at the same time, overseas leading manufacturers continuously shift capacity to high-value-added categories, creating substitution windows in the mid-to-low-end and some mid-to-high-end markets. Mainland manufacturers are still in the catching-up stage, with Chaozhou Three-Circle (Group), Guangdong Fenghua Advanced Technology, and Weiyong Technology having a global market share of 2.5%, 1.9%, and 1.5% respectively by 2024. The domestic replacement path is expected to evolve step by step along the lines of "mature specification substitution in consumer electronics, entry into the domestic automotive electronic supply chain, cooperation of AI servers with domestic computing power chain to achieve from 1 to N". Price cycles are reshaping, with high-end AI/vehicle specifications products and general consumer products showing a K-shaped recovery. Since Q4 2025, Taiwanese and Japanese manufacturers have successively raised prices. According to media reports, Murata has raised prices for high-end products by 15% to 35%, signaling the establishment of the upward cycle for high-end MLCC in this round. The bank judges that this round of the cycle presents a significant "K-shaped differentiation": high-end AI/vehicle specifications products benefit from high demand growth, rigid capacity, and cost transmission, with price sustainability; general consumer products benefit from cost transmission and order overflow, with mild repair potential, but price elasticity and sustainability are expected to be weaker than high-end categories. Risk factors: AI servers and new energy vehicle demand lower than expected; high-end MLCC price transmission lower than expected; domestic manufacturers' high-end product yield climbing and customer certification lower than expected; domestic computing power chain volume and domestic MLCC import progress lower than expected; overseas giants expand production beyond expectations and industry competition intensifies; significant fluctuations in raw material prices.