Invest1 distinct publisher3 min readUpdated
Foreigners swung from heavy net sellers to heavy net buyers of KOSPI shares in a single week as Samsung and SK hynix rebounded. The small-cap record run was a flow event.
The Investor · Invest desk

Compiled by The InvestorSomething wrong?How this is made
The Korea Exchange said on the 16th that the KOSPI closed at 6,977.94 on the 14th, up 2.42% in a session, while the KOSDAQ finished the same day at 864.65, up 0.38% [1][2]. That one-day spread is the visible end of a fortnight in which leadership in Korean equities moved back to large-cap semiconductors, which suggests the KOSDAQ's record August was a plumbing event rather than a rotation with staying power.
The weekly numbers are stark. In the first week of the month, the 3rd to the 7th, the KOSDAQ rose 10.98% and ranked first for returns among major global indexes while the KOSPI fell 5.10% [3][4]. In the second week, the 10th to the 14th, the KOSPI gained 11.49% and took the top spot among major global indexes, with the KOSDAQ up 8.24% and second [5][6]. Month to date the KOSDAQ still leads at 20.13% against the KOSPI's 5.80%, a gap of 14.33 percentage points [7][8][9]. Compounding those two weeks accounts for essentially all of each index's monthly return, so what looks like a month of small-cap outperformance is two weeks of opposite-signed flow [10][11].
The flows say who was driving. Foreign investors net sold nearly 6 trillion won of KOSPI shares in the first week and net bought 6.5741 trillion won in the second, a swing of roughly 12.6 trillion won [12][13][14]. Institutions added 771.9 billion won of KOSPI shares in the second week; retail investors net sold 7.0846 trillion won [15][16]. On the KOSDAQ the foreign bid never showed up: net selling of more than 1 trillion won in the first week and 618.7 billion won in the second, with institutions selling 40.3 billion won and retail absorbing the lot with 646.0 billion won of net buying [17][18][19][20]. Retail carried the KOSDAQ. Foreigners set the KOSPI.
On the cause, Samsung Electronics and SK hynix rose for four consecutive sessions from the 11th to the 14th, which Sedaily attributes to eased concerns about a slowdown in US AI infrastructure investment combined with expectations for new shareholder-return policies [21][22]. The KOSDAQ side of the trade has a more administrative origin: funds shifted there after deposit requirements on single-stock leveraged products were tightened late last month [23]. KOSDAQ trading value went from 4.4929 trillion won on the 30th of last month to 5.4357 trillion won on the 31st, the first day the rules took effect, then 7.1234 trillion won on the 10th and 7.5587 trillion won on the 14th, a rise of about 68% [24][25][26][27][28]. Turnover kept building while the index stalled, which reads less like conviction in small caps than like leverage-seeking money with nowhere else to go.
Watch three things. Whether the KOSDAQ can find the 15.7% it needs to reclaim 1,000 from 864.65 [29]; Kang Jin-hyuk of Shinhan Securities describes the index as consolidating in the mid-800s while it waits for additional momentum [30]. Second, the design of the proposed top tier tentatively called KOSDAQ Select, where industry sources say the argument is over how many names to include: stricter criteria than the existing KOSDAQ 150 would attract institutions, but too narrow a list concentrates funds in a few stocks and raises volatility [31][32]. Third, whether foreign buying of chips persists, because the KOSPI is only 0.32% from 7,000 and needs no new story to get there [33].
Follow any of these and your For You feed starts watching them — no settings page required.
Ranked by verification strength, evidence, and original report placement.
The Korea Exchange said on the 16th that the KOSPI closed at 6,977.94 on the 14th, up 2.42% from the previous session, putting a return to the 7,000 mark within reach.
The KOSDAQ ended the 14th at 864.65, up 0.38%.
For the month as a whole, the KOSDAQ remains in first place with a 20.13% return.
For the month as a whole, the KOSPI has gained 5.80%, narrowing the gap to the KOSDAQ.
Foreign investors net sold nearly 6 trillion won on the KOSPI in the first week of August.
Foreign investors net bought 6.5741 trillion won on the KOSPI in the second week.
Evidence-backed comparisons of source perspectives and observed adoption signals. Read the methodology
Which Builder, Operator, and Investor concerns the observed source mix emphasized—not a truth score.
Evidence, demonstrated adoption, hype gap, incentives, and confidence are assessed independently, each on its own current evidence. How these are measured.
Exchange data, one publisher
The quantitative core - index closes, weekly and month-to-date returns, investor-category net flows, and dated KOSDAQ trading values - is attributed to the Korea Exchange and is internally consistent: the published weekly returns compound exactly to the published month-to-date figures for both indexes. That consistency and the presence of a named sell-side analyst raise evidence quality above assertion level. It is capped by the cluster having a single publisher, no primary KRX document or link, and no independent verification of the flow totals.
Not applicable in supplied material
The cluster contains no releases, deployments, benchmarks, pricing or licence changes, or disclosed usage of a product or technology. It reports secondary-market index performance and investor flows, which the supplied material does not frame as adoption of anything, so no adoption observation can be recorded without inventing one.
Numbers solid, causation asserted
The measured facts are not overstated - the flow and return figures are specific and self-consistent - but the explanatory layer runs ahead of the evidence. The claim that eased U.S. AI-capex slowdown concerns and shareholder-return expectations pulled foreigners back into chip leaders is the outlet's own attribution with no supporting data, and the leveraged-product rotation between boards is asserted rather than quantified at fund or product level. The KOSDAQ Select tier is presented as a potential source of further gains while resting on unnamed sources and a tentative name. Hence a modest positive gap rather than a large one.
Exchange data plus sell-side and anonymous sourcing
The primary data provider is the Korea Exchange, which has a structural interest in presenting active, rising markets, and the outlet is a domestic financial daily whose readership rewards bullish market wraps. The forward-looking commentary comes from a sell-side researcher at Shinhan Securities and from unnamed industry sources promoting a KOSDAQ Select tier designed to attract institutional money - parties with an interest in higher turnover and index inflows. Offsetting this, the backward-looking figures are exchange statistics that are hard to shade and the article also reports heavy foreign selling on the KOSDAQ and a stalled index, which is not a promotional detail.
Checkable figures, single voice
Confidence is moderate: the factual spine is specific, dated, exchange-attributed and arithmetically self-consistent, which supports the cluster's central reading that the KOSDAQ's August strength was flow-driven and is now reversing toward KOSPI large-caps. It is held down by there being exactly one publisher, no linked primary data, an unquantified causal narrative, and no adoption dimension available as an independent cross-check.
invest
Korea's biggest fund houses answer 91 trading halts with bonds, gold and covered calls1 distinct publisher
invest
Five brokers cut JYP targets because two groups' availability is the entire thesis1 distinct publisher
invest
Four sessions, 8.07 trillion won: the KOSPI is now a foreign positioning trade1 distinct publisher
invest
Korean mortgage renewals reset 0.5 to 0.8 points higher, and the ETF desks smell an opening1 distinct publisher
Distinct publishers with included, body-backed reporting in this cluster.
en.sedaily.com
1 article · August 16, 2026