The Real Reason Behind JPMorgan's Skyrocketing Price Target for Samsung Electro-Mechanics

I recently checked my stock app and thought I had misread the numbers.

Samsung Electro-Mechanics was up over 15% in a single day. It’s not exactly a semiconductor bellwether; it’s a components company that makes capacitors and substrates.

At first, I thought it was just a thematic reaction. But the more I looked into it, the more I realized this wasn’t a flash-in-the-pan event driven by some speculative group, but a trend that had been building for months. And right at the center of that trend was JPMorgan.

Why JPMorgan Picked Samsung Electro-Mechanics as a Top KOSPI Pick

In a Korean equity strategy report released last May, JPMorgan raised its KOSPI target to 10,000 in its bull case scenario. The base case was 9,000, and the bear case was 6,000. It was the first time I’d ever seen a 10,000-point KOSPI target in a brokerage report.

The rationale was memory semiconductors.

JPMorgan’s logic was that Korea is the only country besides the U.S. to have both Samsung Electronics and SK Hynix in its top 20 market cap rankings, and that in a phase where rising raw material prices from the Middle East are fueling stagflation pressure, they prefer markets with high exposure to AI and security. So far, this is all about semiconductors.

However, if you look at their top pick list, names like Samsung Electro-Mechanics, Hyundai Motor, Samsung C&T, Hanwha Aerospace, and Samsung Life Insurance are listed right alongside Samsung Electronics and SK Hynix.

A month later, in June, JPMorgan revised its KOSPI target again—to 12,500 for the base case and 15,000 for the bull case—and Samsung Electro-Mechanics remained on the list. It felt strange at first to see a components company mentioned consistently in the context of semiconductors.

*

*Just how many capacitors go into a single server?

MLCCs (Multi-Layer Ceramic Capacitors) are tiny components that store and regulate electrical current inside electronic devices. With hundreds going into a single smartphone and thousands into a car, they aren’t exactly glamorous parts.

One report JPMorgan released last autumn stands out now. At the time, they raised Samsung Electro-Mechanics’ target price from 165,000 KRW to 230,000 KRW, citing an analysis that the amount of MLCCs in one AI server rack is 8 to 10 times that of a regular server. Considering the stock is now in the 1.4 million KRW range, those old numbers feel like they belong to a different era. Of course, that’s largely because the market environment has changed entirely.

The reason for the multi-fold increase in adoption is simple.

AI servers consume vastly more power and perform far more calculations than regular servers, and these small capacitors are what keep the power supply stable. The higher the performance of the product, the higher the required quantity and specifications.

Therefore, as investment in AI data centers increases, it’s not just semiconductor chip orders that are rising—the demand for the supporting components attached to them is growing as well. Of course, given that estimates of adoption rates vary by manufacturer, it’s probably best to view the “8-10x” figure as a broad range rather than an absolute value.

If you just look at the numbers, it might seem trivial, but you can see traces of this in their actual earnings reports.

In the first quarter of this year, Samsung Electro-Mechanics reported revenue of 3.2091 trillion KRW, an 11% increase from the previous quarter and a 17% increase year-over-year. Operating profit reached 280.6 billion KRW, up 17% from the previous quarter and 40% from the previous year. Even after accounting for one-time expenses related to retirement benefits, the figures reportedly exceeded market expectations.

The fact that operating profit grew much faster than revenue suggests that it wasn’t just volume that increased, but also the proportion of high-value products. That said, it’s still too early to declare a trend based on a single quarter.

Why did MLCC prices jump 30% right now?

In late July, Samsung Electro-Mechanics notified clients that it would raise prices on all MLCC products by 30%, effective August 1st. Around the same time, rumors emerged that Japan’s Taiyo Yuden was considering a similar hike for September 1st.

This means it’s not just one company—the entire industry is moving in the same direction.

While the reason for the price hike is often explained as “explosive demand,” it’s more accurate to frame it as a supply issue.

High-performance lines are already at capacity, and it typically takes nearly two years to build a new production line.

The book-to-bill ratio—the ratio of orders received to shipments—for the three major players (Murata, Samsung Electro-Mechanics, and Taiyo Yuden) reportedly rose to between 1.25 and 1.31 as of the end of June, the highest level since the COVID-19 pandemic. Since orders are outpacing production, there’s little room for inventory to accumulate.

The company itself announced that it would more than double its capital expenditure this year compared to last year, which seems like further evidence of how severe the current bottleneck is.

Long-term contracts are a notable development.

This year, Samsung Electro-Mechanics signed a series of silicon capacitor supply contracts with Big Tech clients, including a 450 billion KRW one-year deal and a two-year deal worth over 1.5 trillion KRW. Given that short-term ordering was the norm in the MLCC industry, this is considered unusual, which might signal that clients are feeling anxious about securing supply.

The situation is similar for FC-BGA (semiconductor package substrates), where it was mentioned during earnings calls that demand from servers and data centers alone has already exceeded production capacity.

How is this MLCC boom different from 2017?

“MLCC supercycle” isn’t a new term.

There was a similar boom between 2017 and 2018, followed by a sharp decline. Investors who remember that time probably aren’t feeling very comfortable with the current atmosphere.

Those who argue that “this time is different” generally base their logic on the following:

If the 2017 boom was the result of a mix of inventory shortages and a spike in short-term orders, the current situation is tied to a much longer-term trend of data center investment. Above all, they argue that supply capacity itself is physically constrained.

Estimates suggest that the portion of Samsung Electro-Mechanics’ revenue derived from AI-related business will grow from the 20% range in 2026 to the 30% range by 2027. In August, Morgan Stanley shifted its top pick for Korean tech stocks from Samsung Electronics to Samsung Electro-Mechanics, raising its target price to 2.62 million KRW and leaving the door open to 3 million KRW in a bull case scenario.

However, there are voices pointing out valuation burdens.

One domestic brokerage report noted that the 12-month forward P/E ratio implied by the target price is 92x, which is significantly higher than the historical ceiling of around 30x for this stock. Even Morgan Stanley, which raised its target price, noted a bear case scenario of 1.35 million KRW. Structural factors and valuation pressures are present simultaneously, and it will likely take the next few quarters of earnings reports to clarify which force will prevail.

Ultimately, I’m not sure whether this is a true structural shift or a cycle that will revert like it did in 2018 once the heat dies down. Still, I never expected that talking about semiconductors would eventually lead me to talk about capacitors.