The Rhyme: Korea's Leveraged Chip Crash & Its 2015 Echo

"History doesn't repeat… but it rhymes." — Mark Twain

South Korea's KOSPI index crashed 10% on Tuesday, closing at 8,203 after circuit breakers tripped once in a single session. The trigger was not a war, not a pandemic, not a bank failure. It was the regulator. FSS Governor Lee Chan-jin admitted publicly that his agency had rushed the approval of leveraged ETFs tied to Samsung Electronics and SK Hynix, and that the products had become dangerous. Samsung fell 12.3%, and SK Hynix dropped 12.5%, dragging the VanEck Semiconductor ETF down 7% in New York and pulling Micron 13.5% lower in its worst session since June.

KOSPI −10% to 8,203 | Samsung −12.3% | SMH −7% | S&P 500 7,365

A regulator greenlights leveraged products for retail investors chasing a parabolic rally, then admits it was a mistake. The crash that follows wipes out weeks of gains in a single afternoon. This has happened before. Almost exactly.

On the afternoon of Friday, June 12, 2015, the China Securities Regulatory Commission published a set of draft rules aimed at reining in margin lending by fintech platforms. It was a bureaucratic document, dry and technical, posted to a government website while most of Shanghai was thinking about the weekend. But the timing was a problem. The Shanghai Composite Index had closed that day above 5,166, its highest level since 2008, capping a run of 150% in twelve months. More than 30 million new brokerage accounts had been opened in the first five months of the year alone, most of them by retail investors who had never owned a stock before and were now buying on borrowed money.

The margin lending that powered the rally was staggering. Official brokerage margin debt had swelled to 2.27 trillion yuan. Shadow platforms that skirted the rules accounted for another estimated 2 trillion on top of that. People were borrowing against their apartments to buy shares. Cab drivers in Shenzhen were checking stock tickers between fares. State media had been cheerleading the whole thing, running editorials about how the bull market was good for China's economy and good for ordinary citizens trying to build wealth. Nobody in Beijing told them to stop. Not until it was too late.

The following Monday, June 15, the selling began. Not a slow pullback. A rout. The Shanghai Composite dropped 2% on Monday, then 3.5% on Tuesday. By the end of the week, it was down 13% from its peak. Margin calls went out by the hundreds of thousands, and when overleveraged retail investors couldn't meet them, the brokerages sold their positions into a falling market, which pushed prices lower, which triggered more margin calls. The feedback loop had started, and there was no off switch.

Within three weeks, the index had fallen 32%. More than $3.5 trillion in market value had evaporated. In Beijing, officials scrambled. They halted IPOs. They banned short selling. They created what traders came to call the "National Team," a collection of state-owned financial firms that poured 1.5 trillion yuan into the market, buying shares of thousands of companies. None of it worked for long. By August 24, the day the world would call "Black Monday," the Shanghai Composite had cratered to 3,209, and the eventual peak-to-trough decline would reach 43%.

The lesson was simple. When the regulator builds the ladder and then kicks it out from under you, the fall is worse than a normal selloff. Because the people at the top are the ones who shouldn't have been up there in the first place.

Both crashes started the same way: a regulator approved the fuel, a regulator lit the fuse, and ordinary people holding leveraged products they barely understood got burned first and worst.

The rhyme is tight, but the cracks in it matter more than the pattern.

Here is what happened in China after the first big drop. The Shanghai Composite fell 13% in the first week after the CSRC's June 12 announcement, then bounced. Traders who bought that bounce lost more money than the ones who sold into the crash. By July 8, the index was down 30%. Beijing threw everything it had at the problem. IPO freezes. Short-selling bans. The National Team is buying shares with public money. The market stabilized for about three weeks in late July, long enough for people to start saying the worst was over.

It wasn't. On August 24, Black Monday, the Shanghai Composite fell another 8.5% to 3,209. The total peak-to-trough decline would eventually reach 43%. It took until February 2016 for the selling to fully exhaust itself. Investors who bought "the dip" in late June 2015 sat through another five months of pain before the floor held.

Now look at Seoul. The KOSPI bounced 3.3% on Wednesday after Samsung floated a rumored 90 trillion won buyback. That bounce looks a lot like the dead-cat rallies that punctuated Shanghai's decline through the summer of 2015. The question isn't whether Samsung's earnings justify its stock price. They probably do. The question is whether the leveraged structures wrapped around those stocks have finished unwinding. Leveraged ETFs that have come to dominate daily trading in SK Hynix and Samsung don't disappear after one bad day. The rebalancing math works in both directions, and when the products are being sold by scared retail investors who hold 92% of the float, the selling pressure can last for weeks.

The smart money in 2015 didn't buy the first bounce or the second one. It waited for the leverage to clear the system, which took roughly two months. Then it bought the companies that actually made money. The patient capital that entered China's market in late October 2015 at Shanghai 3,300 made good returns by year-end as the index recovered to 3,539. The ones who chased the June bounce got buried.

In every leverage-driven crash, the first bounce is a trap. The underlying business can be excellent, and the stock can still fall another 20% because the products wrapped around it are mechanically unwinding. The pattern says: wait for the leverage to clear. Watch the daily rebalancing flows on Samsung and SK Hynix ETFs. When leveraged product flows subside, the real bottom is close. Until then, the bounce is the danger.

May PCE data hits at 8:30 a.m. this morning. April headline PCE came in at 3.8%, the highest since May 2023, and the consensus expects another month-over-month increase. If the number prints above 4%, the last time the Fed's preferred inflation gauge was there was early 2023, and the playbook from that era was to keep hiking. Watch how the 10-year yield reacts. A move above 4.60% would start to rhyme with the autumn of 2023, when yields briefly touched 5%, and equities rolled over hard.