The Story
If you only watch headline index numbers, the markets in May 2026 look like one story: stocks are up, AI is still the engine, everyone’s fine. But the index number is the least interesting thing on the screen right now. The interesting part is underneath it — where the money is actually moving. And once you look there, you find something genuinely odd: the US market and the Korean market are both rallying, and they’re rallying in completely opposite shapes.
Start with the US. As of late May 2026, the S&P 500 is sitting near record territory after a sharp run-up — the index climbed roughly 13% from its late-March low, one of its sharpest rallies in years. The obvious assumption is that the usual suspects did the lifting: the Magnificent Seven, the mega-cap tech names, Nvidia and friends. That assumption is now wrong, and that’s the whole point.
For the first time in this cycle, the Mag Seven are lagging. At multiple points in 2026, all seven have underperformed the S&P 500. Their internal correlation — how much they move together — has collapsed to around 25% on a six-month basis, the lowest since at least 2019. Translation: they’ve stopped behaving like a single bloc and started trading on their own individual stories. The “Mag Seven” as a unified trade is quietly coming apart.
So who’s carrying the index? The other 493 companies. Call them the “Impressive 493” — they’re up around 2.9% on the year while the headline index is roughly flat. Look at sector performance since the start of 2026 and the rotation jumps out: Energy is up about 21%, Materials about 17%, Consumer Staples about 15%, Industrials about 12%. Technology, the supposed leader, has been faltering. Money is flowing out of expensive growth and into cyclicals and value — banks, energy, industrials, even boring staples. Caterpillar, Walmart, Exxon: these are the names doing the work now, not the chip darlings.
There’s a name for this and it’s not exotic: it’s a “broadening” rally. Instead of a narrow handful of stocks dragging everything upward, the gains are spreading out across more sectors. Most strategists treat that as healthy — a market resting on 200 stocks is sturdier than one resting on seven. The gap between tech’s earnings growth and everyone else’s is closing, and as it closes, leadership naturally spreads.
Now flip to Korea, because Korea did the exact opposite.
The KOSPI has been the hottest major market on the planet in 2026 — up somewhere in the neighborhood of 50% year-to-date, depending on the measurement date, with intraday spikes that briefly tagged the 8,000 level before pulling back. On May 21 it jumped 8.42% in a single session, a 606-point gain, the largest one-day point gain in its history. Spectacular. But here’s the uncomfortable detail: that rally is the most concentrated the Korean market has ever been. Samsung Electronics and SK Hynix together made up a record 42.2% of the entire KOSPI in May. Two stocks. Forty-two percent of the index. Samsung is up around 82% on the year; SK Hynix around 70%. The KOSPI isn’t really a stock market right now — it’s a leveraged bet on AI memory chips wearing a stock market’s clothing.
So you have two rallies with opposite skeletons. The US rally is getting wider — leadership spreading from seven names to hundreds. The Korean rally is getting narrower — leadership collapsing into two names. Same direction on the chart, opposite structure underneath.
The capital-flows data tells the same split story from another angle. In Korea, foreign investors have been heavy net sellers — they dumped roughly 20 trillion won of KOSPI shares between May 7 and 12 alone, and cumulative foreign net selling since last November runs to something like 80 trillion won. The May 21 record bounce wasn’t foreigners rushing back in; it was led by domestic institutions, with foreigners only mild net buyers and individuals actually net sellers that day. Brokerages largely chalk the foreign exit up to “rebalancing” — after a market runs up 50%, big global funds trim it back to target weight. That’s not panic. It’s arithmetic.
(All figures as of market activity through May 21, 2026. Year-to-date percentages vary by measurement date — treat them as ballpark, not precise.)
The Takeaway
When we wrote about the May market whipsaw a few days ago, the frame was a chain reaction — oil to inflation to rates to a record KOSPI bounce. That piece answered “why did the market move.” This one is about a quieter question that the headline numbers actively hide: “what is the market made of right now.” And the answer is two very different things wearing the same green color.
Here’s the first thing worth holding onto. The US “broadening” is the genuinely encouraging development of 2026, and it’s underrated precisely because it’s boring. For two years the standard worry about American stocks was concentration — too much of the index resting on too few names, so a stumble by Nvidia or Apple could take the whole market down. A broadening rally is that risk unwinding in slow motion. When energy, industrials, and staples start pulling their weight, the index stops being a single bet. It’s not exciting. Healthy markets rarely are.
The second thing is the mirror image, and it’s the part Korean investors should sit with. The KOSPI’s record-breaking year is real, but the shape of it is fragile in a way the index number won’t tell you. When 42% of your market is two semiconductor companies, you don’t own a diversified Korean economy — you own AI memory demand with extra steps. If the memory cycle stays hot, the KOSPI keeps flying. If it cools — and chip cycles always, eventually, cool — there is no second engine underneath to catch the fall. The US spent 2026 fixing its concentration problem. Korea spent 2026 deepening its own.
Which leads to the valuation twist, and this is the one that should make you pause. The old story about Korea was the “Korea discount” — Korean stocks chronically cheap versus global peers because of governance and geopolitical worries. After a 50%-plus run, that story is getting harder to tell. By some trailing measures the KOSPI’s price-to-earnings ratio has pushed up toward the expensive end of major markets, flipping Korea from the cheap half to the pricey half of the global table. (Forward P/E estimates are far lower and more flattering, which is exactly why the bull-versus-bear argument is so loud right now — the two camps are quoting different numbers.) The point isn’t to call a top. The point is that the easy part of the Korea trade — buying something cheap that the world had overlooked — is largely over. What’s left is a bet on whether the AI memory boom keeps delivering, at prices that no longer assume much can go wrong.
So if you’re trying to read the market in May 2026, ignore the index print for a second and ask the structural question instead. In the US, the rally is spreading out, and that’s a quiet vote of confidence. In Korea, the rally is narrowing into two chip stocks at full-price valuations, and that’s a setup that rewards you handsomely while the cycle holds and offers no cushion when it turns. Two rallies, same color, opposite skeletons. The skeleton is the part that matters.
This article is for informational purposes only and is not investment advice.
Photo: Maxim Hopman / Unsplash
댓글 남기기