According to a research report released by Kaiyuan Securities, global original manufacturers will continue raising prices throughout 2026, and the supply-demand tightness across the entire MLCC industry will keep deepening. Beyond passing through costs to ease margin pressure, this round of MLCC price increases reflects a deliberate decision by original manufacturers to curb mid- and low-capacitance orders and free up capacity for high-end supply, in response to surging demand for high-capacitance and ultra-high-capacitance products used in AI servers. Domestic original manufacturers may see price hikes materialize in Q3, and the MLCC industry is expected to sustain a long period of prosperity. Overall, a widening supply-demand gap combined with clear price-increase expectations is expected to drive the global MLCC supply chain into a high-growth channel featuring simultaneous gains in both volume and price. The key views of Kaiyuan Securities are as follows: Samsung has won a major server-grade long-term agreement, Murata has structurally discontinued certain production, and the high-end MLCC gap may continue to widen. According to a report by Shanghai Securities News, Samsung Electro-Mechanics recently disclosed that it signed a supply contract for MLCCs and inductors worth more than 285 billion Korean won (about 212 million US dollars) with a major global customer, marking another breakthrough following the 1.07 trillion won (about 796 million US dollars) order for AI server MLCCs for the full year of 2027 received on September 1. So far this year, Samsung Electro-Mechanics has locked in four long-term MLCC agreements extending to the end of 2027, with a total value of 2.11 trillion won; it has also secured silicon capacitor orders worth about 1.5 trillion won. As global AI computing demand grows, the capacity gap for high-capacitance and above MLCCs may continue to expand. On one hand, Samsung has repeatedly signed LTAs extending to 2027 to lock in its 2027 capacity, confirming that demand for AI server-grade MLCCs remains tight. On the other hand, Murata launched product line optimization in its 2026 fiscal year, explicitly discontinuing certain part numbers in consumer-grade and automotive-grade lines such as GRM and GCM, potentially freeing up capacity for ultra-high-capacitance products. However, its official timetable shows that customer confirmation will not be completed until the end of 2027, last-time buy (LTB) will be extended to March 2028, and final shipments will run until 2029. This lengthy EOL timeline proves that shifting and expanding capacity from consumer-grade and automotive-grade to AI server-grade requires a relatively long cycle. Global original manufacturers will continue raising prices throughout 2026, and supply-demand tightness across the entire MLCC industry will keep deepening. Since the start of 2026, the global MLCC price increase wave among original manufacturers has continued to deepen, with price increase letters being implemented one by one starting in February. In February, Yageo took the lead in raising prices for high-voltage and high-capacitance MLCCs by 10%-20%; in March, Murata adjusted prices for AI server and high-end automotive-grade products by 15%-35%; in April, Taiyo Yuden raised prices for consumer low-capacitance and automotive MLCCs by 6%-13%, with Samsung Electro-Mechanics following suit; in June, Walsin Technology followed. In July, Yageo successively launched across-the-board price increases to ease cost pressure from high upstream raw material prices; in August, Samsung Electro-Mechanics uniformly raised shipment prices across all product categories by 30%, triggering precautionary stockpiling by distributors and creating a supply-demand divergence in which "actual ODM orders declined while channel orders climbed." According to TrendForce, it is expected that consumer-grade mid- and high-capacitance quotes will still rise by 10%-20% in Q4, and Samsung Electro-Mechanics plans to raise prices for consumer-grade X5R MLCC customers by an average of 25% to 30% in 2026 Q4, partly to reduce orders and release capacity for high-end production. The firm believes that, beyond passing through costs to ease cost pressure, this round of MLCC price increases is also a decision by original manufacturers to actively curb mid- and low-capacitance orders and free up capacity to secure high-end supply in the face of surging demand for high-capacitance and ultra-high-capacitance AI server products. Domestic original manufacturers may realize price increases in Q3, and the MLCC industry is expected to maintain long-term prosperity. The firm believes that domestic MLCC original manufacturers were previously in a global price trough, and the pricing benefits from manufacturers' adjustments in June-August had not yet been reflected in their 2026 first-half financial reports. As Q3 price increases are concentrated and realized, domestic original manufacturers' third-quarter results are expected to achieve strong year-on-year and quarter-on-quarter growth; looking ahead, from 2026 Q4 to 2027 H1, original manufacturers at home and abroad may set off a new round of price increases. From the perspective of the supply-demand landscape, under the dual effects of Samsung locking in capacity through 2027 long-term agreements and Murata still needing time to ramp up transferred production, the global capacity gap for high-capacitance and above products may further intensify and is expected to continue expanding in 2027, creating a golden window for domestic original manufacturers to accelerate customer penetration and share gains. Overall, the widening supply-demand gap and clear price-increase expectations are forming a resonance that is expected to drive the global MLCC supply chain into a high-prosperity development channel of "rising volume and price." Investment recommendations, beneficiaries: (1) domestic MLCC original manufacturers: Sanhuan Group, Fenghua Advanced Technology, Sunway Communication, Torch Electron, Yunjia Technology, etc.; (2) upstream material manufacturers: Jiemei Technology, Boqian New Materials, Sinocera Materials, etc. Risk warnings: AI industry development falling short of expectations, raw material price increase risks, capacity expansion falling short of expectations, and technology R&D falling short of expectations.