The market has already written the script: Korea topped out, foreigners sold, leverage got too playful, and now the whole thing is coming apart. That is the consensus. Reality is less cinematic. A market can fall into bear territory while the actual earnings engine is still running hot, and that is exactly what a positioning unwind looks like when it starts with the most crowded names.
Let’s use the hard stuff first. Reuters reported that the KOSPI dropped more than 20% from its late-June record close, which is the textbook definition of a bear-market break. But Korea Customs Service data showed June exports up 4.3% year over year to $58.2 billion, with semiconductor exports up 14.7% to $14.97 billion; that is not the profile of an economy suddenly losing its core cash generator. The market is selling price like it saw a fundamental collapse, while the operating data says the machine is still humming.
The flow story is Korea-specific, and it matters because the index does not trade in a vacuum. Reuters reported foreign investors pulled $137 billion from Asian stocks in the first half of 2026, and South Korea was one of the biggest pressure points in that move. When the marginal buyer turns into a seller, you do not need a recession to get a violent chart; you just need crowded ownership and a lot of people heading for the same exit.
Now the valuation piece, which is where the market starts telling itself fairy tales. Korea’s rally had already been compressed into a narrow leadership trade built around semis and a few large names, so the setup was fragile before the selloff even began. Once price starts falling faster than earnings, the market is not discovering a broken business model — it is repricing a crowded one. That distinction matters because valuation compression can chop 20% off an index without a matching hit to profits.
Earnings revisions are the cleanest truth serum here. If the bear market were fundamentally earned, you would already see repeated estimate cuts across the major exporters and the index itself. Instead, the current evidence points the other way: exports are still growing, semis are still growing, and the burden of proof sits with the bears to show actual revision damage, not just scary tape action. That is why the thesis stays narrow: this is a positioning event unless earnings revisions, exports, and guidance all roll over together.
The operating picture outside semis still needs to hold up, and that is where the next test lives. If Korea’s export base broadens rather than contracts, and if major exporters keep guidance intact, then the selloff is mostly a function of who owns the stock, not whether the business has cracked. That is the market’s blind spot right now: it keeps trying to turn a cleanout in ownership into a verdict on industrial reality.
Here is the deadpan fact bomb: Korea posted $58.2 billion of exports in June while the index was already down more than 20% from its late-June peak. A market can lose 20% on a valuation reset and 0% on a factory shutdown. Prices are emotional; shipments are rude.
So cut the drama. The consensus wants to call this the start of a structural break because that sounds smarter than admitting the trade got crowded and then unwound. But the evidence in front of you is simpler: Korea is getting hit by foreign selling, valuation compression, and a reset in positioning, while exports and semis are still putting up numbers that do not belong in a true earnings recession. If the next two export prints stay positive and the next revision cycle does not meaningfully cut consensus, the bear market will look a lot less like a regime change and a lot more like a very expensive flush-out.
My call is clean: buy the dislocation. If Korea’s next two export releases stay positive, semis keep growing, and consensus revisions do not roll over, this selloff gets remembered as a positioning event that the market mistook for a fundamental break.