KOSPI Crashes 11%: Why SK Hynix, Samsung Stock Fall is Shaking Korea, and What Happens Next

South Korea's KOSPI index closed roughly 11% down on July 28, 2026, its worst single session in years, as Samsung Electronics fell 13% and SK Hynix fell around 15%. That headline is already everywhere. What is missing from almost every article covering this crash is that it is not really one crash.

It is the third time this month that SK Hynix alone has fallen by roughly 15% in a single session, and by Korean brokerages' own math, both stocks were already down 30 to 40% from their 2026 highs before today added anything at all.

Let's break down what actually moved the market today, why Korea fell so much harder than Japan or Taiwan even though they share the same AI chip exposure, whether the math supports a real correction or just a loud pause in an intact rally.

Why Did the KOSPI Crash 11% Today?

The session did not fall in one clean move. It fell in stages, each one worse than the last, which is its own kind of information.

Three sidecars and circuit breakers in one morning is not a normal Tuesday. The index lost roughly 730 points from Monday's close, and it did so while most of Asia had a far calmer day. Japan's Nikkei 225 fell about 4%, Taiwan's Taiex fell 3.9%, Hong Kong's Hang Seng was roughly flat, and Australia's ASX 200 actually gained 0.6%.

The proximate triggers are easy to list. 

• A Wall Street Journal report said Nvidia is in talks to guarantee roughly 250 billion dollars in financing for a 10 gigawatt OpenAI data centre project in Ohio, which sent Nvidia down about 4.5 to 5% on Monday and revived old worries about circular financing in AI infrastructure. 
• Chinese memory maker CXMT closed its Shanghai debut up 466% after pricing its IPO at 8.66 yuan, giving it a market value near $485 billion, and a separate report said a Chinese state-backed firm has begun making its own immersion DUV lithography tools for SMIC, Hua Hong and CXMT. 
• Kiwoom Securities analyst Han Ji-young told Reuters that chip stocks sold off because investors have stopped rewarding a good quarter on its own. They want proof the spending keeps going.

SK Hynix Stock Crash: Three 15% Drops in One Month

Here is the part almost nobody covering today's move is putting in context. SK Hynix has now fallen by roughly 15% in a single session three separate times in July alone.

Average those three sessions and you get 14.89%, with barely a percentage point of spread between the worst and mildest of the three. By local brokerages' own reporting, SK Hynix had already fallen as much as 32% from its 2026 high before today, and Samsung had fallen as much as 40%, driven by worries over NAND pricing and leveraged retail positioning rather than any single news event. Today's drop lands on top of that, not at the start of it.

That reframes the question. The headline asks whether an 11% day ends a bull run. The more useful question is why a stock can lose a third of its value across a month of trading in a market that, on paper, is still forecasting a record quarter.

Why SK Hynix and Samsung Control Nearly Half of the KOSPI

This is the structural piece, and it is the reason Korea's crash was so much sharper than its neighbours' today.

Samsung Electronics and SK Hynix now make up roughly half of the KOSPI's total weight. eToro market analyst Zavier Wong noted this was closer to a quarter at the end of last year, meaning the concentration has doubled in under seven months. Call this the ‘two-engine problem’. 

Most large stock indices are built like planes with many engines: if one sputters, dozens of others keep it in the air. The KOSPI, as of today, effectively flies on two engines. When both cough at once, the whole plane loses altitude. It is not that the other 900-plus listed Korean companies did something wrong today. It is that they were never generating enough lift to matter.

The practical takeaway from this framework is that the real question is not whether the Korean stock market is broken. With this much of the index riding on two names, that question has effectively collapsed into a narrower one: is the AI memory chip cycle broken. Everything else, the sidecars, the circuit breakers, the 10.8% headline, follows from how that one question gets answered.

Can AI Memory-Chip Cycle Recover After the KOSPI Crash?

Memory chips are a famously boom-bust business, and both companies have lived through the bust side of that cycle inside the last decade. SK Hynix posted a full year operating loss of 7.73 trillion won in 2023, with a net margin around negative 28%, at the bottom of its most recent down cycle. Its finance chief at the time called the pace of the price collapse the worst since late 2008. Further back, in 2003, the company traded close to 135 won a share and carried the nickname "Dongjeon-ju," or penny stock, following a near-bankruptcy and creditor takeover.

That history matters for two opposite reasons. It shows the industry can and does overbuild capacity, flood the market, and crush its own pricing, which is precisely the bear case skeptics are making about the current AI-driven buildout. It also shows that when SK Hynix kept building HBM capability quietly through the last downturn instead of retreating from it, that bet is what let it overtake Samsung on market capitalisation for the first time in June 2026. 

Shin Jae-yong, a business administration professor at Seoul National University (SNU), described the reversal as something few in the industry would have predicted, given how hard it is for a challenger to catch a market leader in a business this capital intensive. Cycles end. In this industry's own recent history, they have also rewarded whoever kept building through the downturn.

SK Hynix vs Samsung: Analyst Ratings and Share Price Targets

If professional analysts had a clear, shared view of fair value here, that alone would tell you something. They do not.

Look at the spread on SK Hynix alone: 1.85 million won at the cautious end, 4.2 million at the bullish end, a gap of more than double on the exact same stock at the exact same time. This reflects genuine disagreement about whether current profit levels are the new normal or a peak about to reverse. 

Zooming out, data compiled from FnGuide shows the average gap between broker targets and actual prices across the KOSPI's top 20 stocks by market cap has widened from about 25% at the start of the year to roughly 64% now, with Samsung and SK Hynix among the widest gaps in that group. 

When target prices and market prices drift that far apart across an entire market's largest stocks, it is a sign that models have not caught up with sentiment, in either direction.

Are SK Hynix and Samsung Stocks Cheap After the KOSPI Crash?

Current earnings for both SK Hynix and Samsung are running at a genuine cyclical peak, Samsung's preliminary guidance for the June quarter showed operating profit near 89.4 trillion won, close to a 19-fold jump from a year earlier, which makes any multiple built on today's or next year's profit look artificially cheap if that peak does not hold. That is the central health warning on every number below.

With that caveat stated plainly, here is where the two stocks stood on a forward multiple basis just before today's crash, and a rough, clearly-labelled estimate of where that leaves them after it.

Two readings are possible from this same table, and both are defensible. 

• The bull reading: a forward P/E in the low single digits for two companies guiding record profits is genuinely rare, and if AI infrastructure spending keeps growing the way KB Securities projects (industry-wide AI capex rising from about 800 billion dollars this year to 1.5 trillion by 2028), today's price becomes a gift in hindsight. 
• The bear reading: Forward earnings estimates are themselves built on assumptions, specifically HBM4 ramp timing and DRAM pricing, that Korea Investment and Securities already had to cut once this month. If those estimates fall the way they did in every prior memory downturn, the multiple that looks like 2 to 5 times earnings today could look a lot closer to 10 or 15 times a revised, lower earnings base a year from now. 

Neither reading is provably correct today. That is precisely why the earnings due from SK Hynix within the next day and from Samsung on July 30, with full segment detail, matter more than anything written about this stock this week, including this article.

5 Risks Facing SK Hynix, Samsung and KOSPI

How to Invest in SK Hynix, Samsung and Korean Stocks From India

Under India's Liberalised Remittance Scheme, buying SK Hynix or Samsung's Korea Exchange listed shares directly is not an option most retail platforms offer. But one can invest in Korean ETFs like the iShares MSCI South Korea ETF (EWY) and Franklin FTSE South Korea ETF (FLKR) from INDmoney. These ETFs hold SK Hynix and Samsung in majority. These two stocks together make up somewhere close to 40-45% of these funds’ weight, similar to their share of the KOSPI itself. 

An investor buying EWY or FLKR for diversified exposure to Korea's economy, its carmakers, its banks, its entertainment companies, is in practice buying a concentrated bet on the same two chip stocks driving today's headlines, plus a long tail of much smaller positions.

Aside from Korean ETFs, SK Hynix listed American depositary shares on the Nasdaq under the ticker SKHY are also an investment option for Indians via platforms like INDmoney.

What Happens Next for SK Hynix, Samsung and KOSPI?

The bull case: Forward earnings multiples of roughly 2 to 5 times for two companies both guiding record profits are unusual by any historical standard. HBM demand from Nvidia's own roadmap is contracted years out, industry-wide AI capex is projected to nearly double by 2028, and Morgan Stanley, Mirae Asset and KB Securities have all called the recent weakness a buying opportunity rather than a reason to exit.

The bear case, in the numbers that support it: Both stocks have now fallen 30 to 40% from their 2026 highs before accounting for today. The industry has a well documented history of overbuilding into exactly this kind of enthusiasm before prices break for two or three years at a time, as it did in 2018-2019 and again in 2022-2023.

The debate to actually track: Whether SK Hynix's results, and Samsung's full quarterly report with segment detail later this month confirm the HBM4 ramp and DRAM pricing assumptions that analysts have already had to cut once this month. That single data point will do more to answer whether the bull run is over than any single day's index move, including today's.

For investors weighing exposure to Korean stocks, the more immediate decision is less about timing this specific dip and more about which instrument matches the intended bet: SK Hynix's own Nasdaq listing or EWY and FLKR for something close to the same concentration wrapped in a country label. Neither is a call on whether Korean stocks are cheap or expensive right now. That call, on the evidence available today, is genuinely not settled, and anyone telling you it is has skipped the parts of this story that do not fit their headline.