Semco at 1,757,000 Won: The MLCC Content Story the Multiple Hasn’t Caught

An AI server carries several times the multilayer ceramic capacitor content of a conventional server, and that single ratio is what the price at 1,757,000 won is still underwriting incompletely. Samsung Electro-Mechanics crossed three trillion won in quarterly revenue for the first time in the most recent quarter, reporting 3.2091 trillion won, up 17 percent year on year, with operating profit of 280.6 billion won, up 40 percent. The operating profit growing more than twice as fast as revenue is the figure that matters. It is the signature of a mix shift toward high-end MLCCs and FC-BGA substrates, where the company is selling into demand that exceeds what it can currently produce.

The reason this matters for value rather than momentum sits in the gross margin trend. Margin ran 26.3 percent in 2021, then compressed to 19.3 percent in 2023 and 19.0 percent in 2024 as the consumer-electronics cycle bottomed and capacity sat underutilized. The most recent full year came in near 20.1 percent. A reporting unit whose margin has been clawing back off a trough, while its highest-value end market accelerates, is not the same business the trailing multiple was assigned to. Return on capital recovers fastest when fixed-cost lines that were running below capacity start filling with premium product, and that is the mechanical state management described when it said strong demand should persist for AI servers and data centers.

The market that prices passive components has historically made the same error at the same point in the cycle. Capacitors and substrates get treated as commodity throughput, valued on a cost base that assumes pricing stays flat, right up until utilization tightens enough that the supplier sets price rather than accepts it. In prior supply-constrained stretches for high-end MLCCs, the pricing inflection arrived before the consensus margin estimates moved, and the stocks were re-rated only after the higher margin printed. What was missed earlier each time was that the content-per-unit step-up in the new end market changes the volume base permanently, not for one or two quarters. The same dynamic is visible now: management has discussed price increases into full utilization projected for the second half, which is the pricing-before-margin sequence repeating.

The backdrop reinforces this. Industrial policy across the US and allied economies is pushing AI compute and data-center buildout as a strategic priority, and that demand lands directly on the component layer that Semco supplies. The same geopolitical contest carries a cost: the shift to annual US export licenses, replacing indefinite approvals, for chipmaking-equipment shipments into China facilities raises the compliance and disruption risk premium across the Korean electronics supply chain that Semco feeds. Korea’s position as the second-source of high-end MLCC capacity outside Japan is what makes the demand spillover land here specifically rather than dissipating. Full-utilization economics in this segment convert incremental volume to cash at a high rate, because the capital is already in the ground. That is the honest test ahead: whether the operating-profit acceleration shows up as FCF conversion rather than working-capital buildup.

On price, the move has already been large, so the setup has to be stated plainly: the stock has run sharply over recent months as the quarterly beat confirmed the demand and pricing story. The point of a constructive case here is not the move that happened. It is that the content-per-server step-change and the second-half pricing have a multi-quarter tail, and the P/E versus a growth-adjusted multiple still reflects a business priced closer to its trough-margin identity than its recovery one. Capital allocation is the lever that decides whether this compounds: a company that funds the high-end expansion without bloating the cost base, and that resists the temptation to chase low-margin commodity volume to fill lines, earns the higher multiple. If management chooses instead to divest weaker package or optical lines and concentrate capital on the MLCC and substrate cash engine, the return-on-capital gap closes faster.

The risks are specific, and they attach to numbers. The first is that gross margin stalls below the low-20s instead of continuing toward the mid-20s the 2021 base showed was achievable. If utilization fills with mix that does not carry premium pricing, the operating leverage that drove 40 percent profit growth flattens, and the recovery thesis weakens with it. The second is concentration: a large share of revenue routes through the Samsung electronics supply chain, so the annual-licensing regime on China facilities is not abstract. A disruption that idles part of that downstream demand would hit utilization directly, and utilization is the entire mechanism here. The third is the cyclicality of MLCC pricing itself, which has historically given back gains as fast as it took them once new capacity arrives.

If second-half operating margin fails to hold above the roughly 20 percent the most recent year delivered, and instead drifts back toward the 19 percent trough of 2023 and 2024, this bull case breaks down, because the entire argument rests on the mix shift converting to durable margin rather than a single-quarter spike. Until that shows up in the print, the numbers support a business whose earnings power the multiple is still discounting against its weakest years.

Figures reflect the most recent available data and may differ slightly from live market prices. · © Mathstock