📊 Samsung Electro-Mechanics: Could the AI Component Shortage Translate Into Results? - 2026.10.26

Well, well~ Onion’s here with a cup of coffee. ☕ You can’t just shrug off talk that AI-driven component shortages are spreading to substrates and MLCCs! 📈

The shortage has moved beyond memory

According to a ChosunBiz report, AI demand is driving component shortages beyond memory semiconductors and into MLCCs and substrates ^[2]. It also covered why the earnings outlooks for Samsung Electro-Mechanics and LG Innotek differ. The important thing here is that these are forecasts from securities firms and others. They aren’t plans confirmed by the companies, but views from third parties that should be read as conditional. After 30 years in the market, I’ve learned to first check who’s making claims like these and what evidence they’re using.

A scramble for substrates as the market grows

The picture gets bigger when we turn to FC-BGA substrates. Big tech companies are moving to secure K-substrates, and the news is that supply shortages are persisting because substrates have high technological barriers to entry ^[1]. Another key point is that domestic companies are drawing attention in a market led by Japan and Taiwan. LG Innotek is also reportedly pushing to begin mass production of substrates for AI accelerators next year, beyond its PC products. If competitors are jumping in too, doesn’t that signal just how hot demand is? Still, this is also reported information, so it’s too early to treat it as confirmed fact.

The contract term is short, so why are expectations high?

Naver Premium Contents covered why Samsung Electro-Mechanics chose a short contract term even as big tech companies are posting strong results ^[3]. But without details in the article, I can’t speculate about the contract terms or the reasons behind the choice. Still, one question remains: Are longer contracts always better, or is it better to keep pricing and volumes flexible with shorter contracts? The title of a Seoul Economic TV video emphasized that a massive contract could land and change the first digit of the share price ^[6]. A headline is just a headline, so the answer is to weigh the video itself alongside company disclosures.

The share price is still fairly quiet

Looking at the price action, the closing price on October 26 was ₩1,119,000 and the change was -0.27%. The previous closing price, on October 23, was ₩1,111,000, down -3.05%. News headlines reported that the share price was rising ^[5], while there was also news of a decline at the close that same day ^[4]. The detailed reasons can’t be confirmed because the article text isn’t available. The mismatch—good news piling up while the share price fails to follow—is the most interesting part today. 🤔

Onion thinks the bullish scenario is clear if demand for AI components feeds through to earnings. But quarterly results and the pace at which contracts convert into revenue will show whether that premise holds. 🚀

There are three key factors: how quickly long-term substrate contracts show up in earnings, how short contract terms affect pricing and supply-demand conditions, and whether securities firms’ forecasts are borne out by actual numbers. Doesn’t having so many questions make this even more interesting? That’s the point, exactly!