An image symbolizing the "risk window" for the Korean stock market in September and October, where foreign profit-taking, pressure from National Pension Service (NPS) rebalancing, retail margin trading, currency volatility, and U.S. trade pressure overlap. Following the surge in the KOSPI, foreign profit-taking, retail buying, NPS rebalancing, and currency instability are mounting simultaneously. If U.S. trade pressure ahead of the midterms is added to the mix, the Korean stock market could enter a risk window. [Photo=Hanmi Ilbo Graphic]Is the risk window for the Korean stock market opening in September and October? 

Was the surge in the KOSPI the result of an influx of new foreign capital, or was it the result of retail investors and the National Pension Service (NPS) absorbing the supply as foreign investors sold off their holdings to realize gains? 

Recent supply and demand trends, exchange rates, NPS asset allocation, and the U.S. trade agenda all point to the same question.

If what is to come must come, the market's timetable points to September and October. The moment trade pressure ahead of the U.S. midterm elections meets foreign capital flight and currency instability is the most dangerous. 

This Hanmi Data Lab report follows four numbers hidden behind the KOSPI surge: foreign holdings, retail margin debt, the NPS allocation ratio, and foreign exchange intervention amounts.

Data ① Foreigners sold, but still profited

According to the "Trends in Foreign Securities Investment" for May 2026 released by the Financial Supervisory Service, foreign investors net sold 47.019 trillion won in listed stocks during the month of May. Foreign net selling of stocks has continued for five consecutive months. Despite this, as of the end of May, the value of foreign-held listed stocks reached 2,852.3 trillion won, accounting for 35.3% of the total market capitalization.

What these figures show is clear: foreigners did not buy massive amounts of new Korean stocks; rather, they sold stocks while the valuation of their existing holdings had skyrocketed. As of the end of December 2025, foreign-held listed stocks were valued at 1,326.8 trillion won. By the end of May 2026, this increased to 2,852.3 trillion won. In terms of holdings alone, this is an increase of over 1,525 trillion won in just five months.

Of course, this is not realized profit. Foreign investors' actual trading gains or losses can only be determined by checking the buy and sell prices of individual stocks. However, looking at the balance of holdings and the flow of net selling, foreign investors are closer to being entities that cashed in some of their holdings after securing valuation gains during a bull market, rather than being the drivers who pushed the market up through new buying.

Reuters reported that in the first half of 2026, foreign investors net sold $137.36 billion in Asian stocks, with $70.8 billion exiting from Korea alone. This report interpreted the trend as foreign investors moving toward profit-taking and rebalancing because the AI and semiconductor rallies in Korea and Taiwan had become too steep.

The key point is this: foreign investors did not grow the market; they sold into a grown market. Yet, their total holdings still increased. This is the first optical illusion of the current market trend.

Data ② Retail investors and the NPS were the ones absorbing the supply

If foreigners were selling, who was buying? The answer is retail investors and domestic institutions. In particular, retail investors were the primary supply-and-demand agents that absorbed the volume shed by foreigners during the rally. The problem is that a portion of that buying was funded not by their own money, but by debt.

According to the Korea Financial Investment Association, the balance of margin trading loans hit a record high of 38.4786 trillion won on June 19, 2026. The margin balance on the KOSPI market alone reached 29.3977 trillion won. As the KOSPI crossed the 9,000 mark, leveraged investment by retail investors also reached an all-time high.

In a bull market, margin debt amplifies returns. However, it works in reverse in a bear market. When collateral ratios collapse, securities firms force liquidation through margin calls, even if investors want to hold on. While foreign selling is profit-taking and NPS selling is rebalancing, retail "debt-investing" (bit-too) selling is forced liquidation.

The NPS is another pillar. This year, the NPS raised its target allocation for domestic stocks from 14.9% to 20.8%. The Ministry of Health and Welfare explained that this adjustment was intended to mitigate the market impact of increasing the actual domestic stock weight and conducting rebalancing. The target domestic stock ratio takes effect starting at the end of June, when the rebalancing grace period ends.

Viewing this measure from the perspective of the stock market, it effectively built a higher dam. It bought time so that the NPS would not immediately begin mechanical selling even if domestic stocks surged. However, dams are not infinite. The NPS is not an institution meant to defend the stock market; it is a fund that manages the nation’s retirement assets. If the domestic stock portion exerts pressure on its targets and permissible ranges again, the NPS will eventually face rebalancing pressure.

Therefore, while the NPS acts as a dam at first, it becomes a floodgate at some point. We are currently standing at the inflection point where the dam is turning into a floodgate.

Data ③ Exchange rates have already sounded an alarm

Stock market selling can end as a simple correction. But the story changes if foreign investors convert the won they received from selling stocks into dollars and leave the country. At this point, stock market selling turns into dollar demand in the foreign exchange market. If the won weakens further, foreign investors may sell even faster to avoid currency translation losses.

According to Reuters, Deputy Prime Minister of Economy Koo Yun-cheol assessed that the won-dollar exchange rate staying in the mid-1,500 range is excessive compared to economic fundamentals. He noted that even though Korea's exports and current account balance are strong, the reason the won remains weak is due to foreign investors offloading approximately 140 trillion won in stocks for portfolio rebalancing during the rapid price rally.

Foreign exchange authorities have already spent a significant amount of dollars. According to the breakdown of market stabilization measures for the first quarter of 2026 released by the Bank of Korea, authorities net sold $13.628 billion to stabilize the market in the first quarter. In the fourth quarter of last year, the net selling amount was $22.467 billion. In half a year, over $36 billion has been poured into defending the exchange rate.

The fact that exports are continuing is the greatest defense line for Korea's foreign exchange market.

However, export proceeds are not immediately available as the government's "ammunition" for currency defense. The dollars earned by companies must be supplied to the domestic FX market to serve as a defense for the won. If foreign stock selling, dividend remittances, overseas investments, foreign debt repayments, and dollar-hoarding demand all increase simultaneously, it is difficult to fully block exchange rate instability with trade surpluses alone.

When foreigners sell stocks, it is a correction. But when they convert won into dollars and leave, it is an exodus.

Data ④ Why September and October?

September and October are not dates predicted for a crash. They are a "risk window" where supply/demand, exchange rates, the NPS, margin debt, and U.S. political schedules overlap.

The U.S. midterm elections will be held on November 3, 2026. This election involves all 435 seats in the House and 35 of the 100 seats in the Senate. The 4 to 8 weeks before the election are the period when messages regarding manufacturing, jobs, trade deficits, and external trade pressure are at their most intense.

Added to this is the U.S. Trade Representative’s (USTR) Section 301 investigation. In March 2026, the USTR launched Section 301 investigations into several economies, including Korea, concerning structural overproduction and manufacturing output.

The USTR cited various manufacturing sectors—such as automobiles, batteries, electronics, semiconductors, shipbuilding, and steel—and included Korea as one of the economies under investigation.

If Section 301 targets automobiles, the shock will quickly spill over into the real economy. Automobiles and auto parts affect prices, margins, sales volume, parts suppliers, and employment. If it targets semiconductors, it is even more dangerous. This is because the core engine that has driven the Korean stock market this year could simultaneously become the source of currency instability.

Of course, it cannot be concluded that Section 301 will immediately lead to comprehensive tariffs. The U.S. also needs Korea’s semiconductor, battery, and automotive supply chains.

Therefore, the actual pressure is likely to take a mixed form of tariffs, quotas, investment demands, rules of origin, blocking circumvention through China, and supply chain security logic.

However, the market reacts to the justification before the final form of the measures. It gives foreigners a justification to sell, gives the exchange rate a justification to jump, and puts pressure on the NPS that makes it harder to hold on.

Conclusion of the Hanmi Data Lab

The essence of this market trend is not just a simple bull market. Foreigners sold while profiting, and retail investors absorbed the supply. The NPS bought time by raising the rebalancing dam. However, margin debt has grown, the exchange rate is already shaken, and foreign exchange authorities have poured significant dollars into market stabilization.

If trade pressure ahead of the U.S. midterm elections overlaps with this structure, September and October will become a risk window. The moment foreign profit-taking, NPS rebalancing, retail margin calls, sharp currency hikes, and the U.S. Section 301 card align in the same direction, the market could shift from a correction phase to a liquidation phase.

So the question is this:

“Is the Korean stock market in an extension of a bull market, or is it heading toward the September-October risk window?”

The bill for the KOSPI surge has not yet arrived. However, the numbers are already pointing in one direction.

If what is to come must come, the market's timetable points to September and October.