Is the Korean market dizzying lately? In times like these, you need to study companies and seek out alpha in the Great AI Era. You have to research stocks that are stronger than the index, fell less than Samsung Electronics or Hynix, and have living momentum. You need to dig in, taste them, and obsess over them to earn the right to brag during the next rotation.
Many of you probably sold off after seeing news about AI data center cancellations. But few actually read the full article. Less than a third of the data centers scheduled to operate in the US by 2026 are actually under construction, and the reason isn't a collapse in demand—it's a shortage of power equipment like transformers and switchgear. It's not that customers aren't coming to the store; it's that we can't mine the gold because we don't have pickaxes. In this game, the accounts of those who fled after seeing the headline and those who read the reason for cancellation will diverge sharply.
Let's look at the reality of the US power grid. Import dependency for transformers is over 80%, lead times have exploded from 1 year to a maximum of 4 years, and 70% of the installed transformers are "old-timers" installed around the year 2000. On top of that, AI is consuming electricity at an incredible rate. It's a black comedy of 2026 that the Achilles' heel of US hegemony isn't aircraft carriers, but the fuse box.
What about tariffs? In February this year, the Supreme Court struck down Trump's reciprocal tariffs in a 6-3 ruling. There's even talk of refunding collected tariffs. But interestingly, while Trump was rebuilding the tariff system under Section 232, he personally cut the tariff on power equipment from 25% to 15% first. Though it's temporary until the end of 2027, even amidst the chaos of courts destroying and Trump rebuilding, transformer tariffs only went one way: down. The fact that the man who bullied the world with tariffs is making an exception for Korean power equipment is the real X-File of the 2026 trade war. Demand is also guaranteed by the government. The Fed's Beige Book officially certified that AI data center investment is boosting US manufacturing, and when New York State paused data center permits, Trump himself shouted for them to be released.
Now, for today's protagonist... LS Electric. Why is this guy always in my picks? Let's find out. This is all useful info, so read carefully. If you can't even read this, how are you going to trade stocks?
Starting as Goldstar Instrument & Electric in 1974, passing through LG Industrial Systems, and becoming LS Electric in 2020, it's a 52-year-old veteran of the domestic power equipment industry. From your home's fuse box to factory switchboards, anyone using electricity in Korea has used this company's products. The business is largely three-fold: power equipment like circuit breakers and switchboards; power infrastructure including ultra-high voltage transformers and HVDC; and factory automation. But right now, the protagonist is overwhelmingly power. Here, the positioning of the domestic "Big 3" differs: while HD Hyundai Electric and Hyosung Heavy Industries are the ones building the highways from power plants to data centers with 765kV ultra-high voltage, LS is the one building the capillaries inside the data centers with switchboards and low-to-medium voltage equipment. The point is that this is where Big Tech directly opens its wallet.
So, what do the numbers look like? That's what matters. Q1 revenue was 1.3766 trillion KRW with an operating profit of 126.6 billion KRW, growing 33% and 45% year-over-year, respectively. For Q2, with one-time costs gone and price hikes kicking in, operating profit is projected at 155.5 billion KRW with a 10.8% margin. Annually, the consensus sees revenue at 6.1 trillion KRW and operating profit between 650 billion and 700 billion KRW. The order backlog is even prettier.
They reached 75% of their annual guidance early by stacking 3 trillion KRW in orders in the first half alone, with a total backlog of 7.2 trillion KRW. North American Big Tech revenue hit 1.2 trillion KRW in just half a year, exceeding all of last year. This company's weapon is a delivery time 30% faster than competitors. In a shortage market with 4-year lead times, the one who delivers fast has the moat. Moreover, Big Tech usually signs contracts for supply slots of 3 years or more, so once you're in, it's repeat business. They are serious about the US, too: the Bastrop campus in Texas is complete, the Utah plant is undergoing a 6-fold expansion with a 250 billion KRW investment, and Chairman Koo Ja-kyun personally stated the goal to become a Top 4 player in the US by spending a total of $240 million by 2030.
After reading this far, you probably want to go all-in... you're opening your trading app, aren't you? Now, let's pour a bucket of cold water on that.
Valuation. With the stock price around 250,000 KRW and a market cap of about 37 trillion KRW, the forward PER is about 70x based on the 2026 consensus net profit of 500 billion KRW. The PBR is in the double digits. Historically, this industry received a PER of 12x to 20x, and during last year's correction, the average for the Big 3 reset from 22x to 14x. This is a price where Yeouido analysts openly admit it's hard to explain the current stock price with EPS growth alone. Calculating backward is even scarier. For this market cap to be justified at a mature multiple of 15x, net profit would need to be 2.5 trillion KRW—five times the current level. Even growing at a 25% compound rate, that takes 7 years. So, buying at 250,000 KRW now is like paying upfront for a scenario where LS grows without a single hiccup until the early 2030s.
But why am I on the side of the pickaxe seller? Because in a shortage market, the stock price isn't a multiple of earnings; it's the scalper's price for a spot in line. When a concert is sold out, the guy complaining about the face value of the ticket is the last one to enter. In this market, there are overwhelmingly more spectators than seats, and the fact that this imbalance will last for years is proven by the order backlog. However, even scalper tickets have an expiration date. At the end of 2027, the end of temporary tariff cuts and the global ramp-up of production capacity will coincide. The moment seats increase, the scalper price drops—that's the rule. So, I see the time until then as the duration of this party. The alarm starts with the Q2 results in mid-August.
Conclusion: When others run away seeing headlines about cancellations, I read the reason for the cancellation. If the reason is a lack of equipment, that's not a reason to sell; it's a reason to buy the pickaxe seller. I know it's expensive. But at a party, the most expensive liquor is always the first to run out. You get the feeling, right?
This is not a stock recommendation but my personal opinion. Protect your own account.
