In late July, a price increase notice was posted on the bulletin board of the electronic components industry: Given the continuous rise in the prices of raw and auxiliary materials for zirconia powder production, the company has decided, after careful consideration, to raise the selling price of zirconia powder from July 27, with an increase ranging from 10% to40%. The issuer of this notice is Sinocera, a globally leading producer of dielectric powder for MLCCs (Multi-Layer Ceramic Capacitors).
After the passive components market had been flooded with price hike news for over half a year, the reason this notice deserves to be singled out for scrutiny is simply this—it officially pushes this round of price increases to the very upstream of the industry chain.
Starting with this set of figures, this article will first conduct a comprehensive analysis of three sets of data: the magnitude of the price increase, sales volume, and capacity pace. It will then extend layer by layer along the data towards downstream demand, upstream landscape, and long-term industry trends, attempting to answer one question: When the tide reaches the very upstream, who exactly holds the confidence to raise prices?
01. Data Breakdown: A Complete Map Pieced Together by Three Sets of Figures
Let's look at the magnitude first. The 10% to 40% price increase range for zirconia powder is itself an information-rich expression. Powder is not a standardized commodity with a single specification; differences in purity, particle size distribution, and crystal form control determine its grade at the application end. From the perspective of industry pricing rules, powder for high-end automotive-grade and AI server components, which has stringent requirements for consistency and reliability, usually sees price increases at the upper end of the range, while powder for conventional applications experiences relatively moderate increases.
This gradient structure of "high-end leading the rise, low-end following" is completely isomorphic to the downstream MLCC price increase sequence. Murata Manufacturing Co., Ltd. took the lead as early as late March by raising prices for AI server and high-end automotive-grade MLCCs by 15% to 35%; Samsung Electro-Mechanics has uniformly increased the shipping prices of all product categories by 30% starting August 1; Taiyo Yuden issued a price adjustment notice to customers on July 23, adjusting the prices of some products from September 1. The three price hike letters are arranged sequentially from high-end to all categories, forming a clear transmission sequence, and the zirconia powder letter is exactly the endpoint of this sequence.
Let's look at sales volume next. The 2026 semi-annual report disclosed by Sinocera shows that in the first half of the year, the company achieved an operating revenue of 2.513 billion RMB, a year-on-year increase of 16.64%; net profit attributable to shareholders of the parent company was 363 million RMB, a year-on-year increase of 9.36%; net profit after deducting non-recurring gains and losses was 361 million RMB, a year-on-year increase of 12.54%; and net cash flow from operating activities was 456 million RMB, a year-on-year increase of 36.78%. Notably, the growth rate of cash flow is significantly higher than that of profit—this means the true value of the growth lies not in the recognized revenue on the books, but in the solid quality of cash collection and order fulfillment.
Sinocera explicitly stated in its semi-annual report that, benefiting from the recovery of downstream demand and the rapid growth in demand for emerging application fields such as AI servers and automotive electronics, the sales volume of MLCC dielectric powder and electronic paste has grown rapidly. This is not an isolated piece of evidence. JM Technology, engaged in the release film business for MLCCs, revealed in a survey that the sales volume of its release film business has surged since March this year, growing by more than 100% year-on-year, and it is expected to be in short supply in the first half of next year. The company has proactively adjusted its product structure, abandoned some low-end products, and focused on mid-to-high-end products.
Finally, let's look at capacity. Recently, executives from domestic MLCC material-related companies have repeatedly stated: "Customer orders continue to increase, and actual demand has exceeded current delivery volumes, with order visibility also significantly improved." "Despite operating at full capacity, many orders have to be delayed." Full production with delayed delivery is the golden yardstick for judging real prosperity—because it eliminates the price illusion created by channel hoarding and points to actual production line congestion. Industry calculation data shows that in the second half of 2026, the shortage of high-end high-capacity MLCCs is about 15%-20%, and the shortage may further expand to 30% in 2027. Only when the three sets of figures—magnitude, sales volume, and capacity—are pieced together can a complete picture be formed: the price increase is not just asking for a higher price, but a true expression of order backlog; what the material end receives is the escalating demand transmitted from component manufacturers.
02. Where Does the Confidence Come From: The Three-Layer Squeeze of Demand, Supply, and Capacity
To understand this picture, we need to return to the supply and demand structure itself. MLCCs are known as the "rice of the electronic industry." They are small in size and large in usage, performing functions such as decoupling, voltage regulation, and instantaneous power replenishment, and are present in almost all electronic devices. Past rounds of prosperity cycles were mostly driven by the recovery of consumer electronics and the bottoming out of channel inventory, with price increases often spreading synchronously across all categories and regions. The most essential difference in this cycle is that the first to light up on the demand side is not mobile phones, but AI servers.
The data at the demand layer is highly impactful. Public reports and institutional calculations generally point in the same direction: the MLCC usage per rack in AI servers can be several to over ten times that of traditional servers. Institutions such as Northeast Securities give a magnitude judgment that "the high-capacity demand for AI servers is 5 to 10 times that of traditional servers." Goldman Sachs' latest research report shows that the AI server MLCC segment is expanding rapidly at a compound annual growth rate of about 80%. As GPU power consumption continues to rise and power supply chains become increasingly complex, MLCCs have leaped from "standard auxiliary materials" to critical components that cannot be easily replaced in the computing power bill of materials—they are no longer capacitors that can be replaced at any time, but infrastructure that determines whether the system can output stably.
The squeeze at the supply layer is equally intense. According to TrendForce's latest MLCC industry survey, thanks to robust AI demand, in June 2026, shipments from the three major Japanese and South Korean manufacturers—Murata, Samsung Electro-Mechanics, and Taiyo Yuden—simultaneously hit a five-year high for a single month. Meanwhile, as Japanese and South Korean manufacturers accelerate the shift of consumer-grade X5R capacity to high-end specifications such as X6S/X7R for AI, the capacity for consumer-grade mid-to-high capacity MLCCs (1uF to 22uF) has been squeezed, and the inventory levels of mainstream components have generally fallen below 30 days. Under this crowding-out effect, urgent orders from channels, distributors, and second- and third-tier customers have surged. Distributors who have stocked up in advance have taken the opportunity to raise prices, with an average increase of 20% to 25%, and spot market prices have even been pushed up to two to three times their original levels.
The material layer is the most difficult link to loosen in the entire chain. The purity control, particle morphology, paste formulation, and sintering process of ceramic powder are the products of decades of process accumulation. Core production equipment mostly relies on in-house R&D by leading enterprises, and the release of effective capacity often takes 18 to 24 months. This means that even if downstream component manufacturers decide to expand capacity immediately, the supply elasticity at the material end remains extremely small.
Steepened demand, contracted supply, and rigid capacity—the three-layer squeeze jointly explains the same thing: why the transmission of this round of price hikes is so smooth, and why it can ultimately reach the very upstream of the industry chain.
03. From "Rice" to "Rice Barn": The Value Reassessment of the Material Segment
Continuing upstream along the path of the price hike, a more profound industrial change is taking place: MLCCs are the "rice of the electronic industry," while ceramic powder, electronic paste, release film, and zirconia beads are exactly the "rice barn" for producing this bowl of rice. Over the past few decades, the control of this rice barn has been highly concentrated. According to relevant reports by the China Electronic Components Industry Association, global MLCC capacity is highly concentrated in Japan, South Korea, and the Taiwan region of China. The share of mainland Chinese manufacturers is about 10%, and even lower in the high-capacity field. The concentration in the upstream material segment is even higher than that in the finished product segment.
However, the change is precisely starting to loosen from the most difficult segment. As a globally leading producer of MLCC dielectric powder, Sinocera has, in recent years, through horizontal expansion and layout, mastered various key raw materials for MLCC preparation, such as dielectric powder, inner and outer electrode pastes, and zirconia beads for grinding. It can provide systematic technical solutions and product services for customers in the electronic components field. In the first half of this year, the company focused on MLCC dielectric powder for AI servers and automotive-grade applications, continuou sly developing corresponding products and advancing verification. JM Technology, on the other hand, cut into the segmented release film segment and completed a proactive switch to a mid-to-high-end product structure, relying on the prosperity of doubled sales. In its latest research, AVIC Securities pointed out that upstream materials in the MLCC industry chain possess multiple attributes of "electronic materials + high-reliability military supporting + domestic substitution."
More noteworthy than market share is the shift in roles. In past rounds of price hike cycles, domestic material manufacturers were mostly passive followers—international giants raised prices, and domestic manufacturers followed, earning beta returns on the tailwind of the cycle. This time, however, the price hike letters were proactively issued first by domestic enterprises. Behind this is a concentrated verification of the bargaining power and price transmission capability of Chinese material enterprises: when a Chinese company can independently price the rise in upstream raw and auxiliary material costs, it means it already has the bargaining confidence to go toe-to-toe with international rivals, and it means customers cannot do without its capacity and quality. From "following price hikes" to "independent pricing" is a leap from rice to rice barn, and also a turning point for China's electronic materials industry from moving from a cost item to a value item. This bowl of rice in the computing power era is becoming increasingly expensive, and the rice barn is being built more and more on Chinese land.
04. After the Tide Recedes: What Will Be Left Behind
Every round of price hikes is asked the same question: How long can it last? A rational answer is that any price has a cycle, but there are two fundamental differences between this round of price hikes and the old cycle. First, the source of demand is different—the driving force of the old cycle was the replacement and restocking of consumer electronics, which had strong emotional components and high reversibility; the driving force of the new cycle is the rigid expansion of AI capital expenditure, which has lower price sensitivity to high-capacity components, making high-end prices easier to hold. Second, the capacity sequencing is different—according to public information, Samsung Electro-Mechanics has disclosed multiple supply contracts for AI server-specific MLCCs this year, with the coverage period extended to around 2027, and long-term agreement volume locking is becoming the industry norm. When supply shifts from "short-term procurement" to "medium- to long-term supply security," the price discovery mechanism also changes accordingly. Fluctuations in the spot and channel markets reflect the squeezed marginal demand more than a unified clearing of the entire market.
For domestic manufacturers, this means a divergence is unfolding: one path is to capture the market share expansion brought by the price increase of imported components and uncertain delivery times—when Murata and Samsung Electro-Mechanics prioritize supply for cloud manufacturers and major automotive-grade customers, the mid-to-high-end demand spilling over to second- and third-tier customers is exactly the window for domestic manufacturers to take over; the other path is to continue catching up in the yield rate of high-end high-capacity products, and the independence of powder and equipment—this requires time, R&D investment, and customer verification cycles, but only by successfully navigating this path can enterprises move from the position of being "lifted by the tailwind" of price hikes to a position of independent pricing. The two paths are not mutually exclusive; the prosperity cycle gives speed to the first path, while industry trends give depth to the second.
The tide will rise and fall, but where the tide washes over, it will always leave behind something that cannot be washed away: verified processes, accumulated customer relationships, and a redefined supply order. Returning to the price hike notice at the beginning—10% to 40% is not a simple percentage range; it is China's materials industry, after years of accumulation in the high-end segment, standing at the very upstream of the industry chain, proactively writing its own price. MLCCs remain the rice of the electronic industry, but the direction of the rice barn is quietly changing. This is perhaps the most memorable part of this round of price hikes—not the price itself, but who holds the confidence to set the price.
- End -