Snowshagal · INVESTMENT NOTE
NO.09 · 2026.10.08
Snowshagal ReportKOREA EQUITY · BUYBACK EXIT · FLOWS
INVESTMENT NOTENO.092026.10.08
FOLLOW-UP TO NO.04 · SAMSUNG BUYBACK ENDED OCT. 6 · OCT. 8 RESULTS INCLUDED

When the Buyback BidSteps Away

When KRW 55tn of buybacks stop, who takes the other side—and what should investors watch?

Who sold, who absorbed · What is finished and what remains · What the historical average hides · Three numbers that may reveal the next bid

Samsung Electronics · KRW 15tn Ended Oct. 6 Filed window ends Nov. 21 SK hynix · KRW 40tn Remaining 84.2% by value as of Oct. 6 Around Oct. 20 Filed window ends Nov. 19 Published Oct. 8 8.20 11.21
Dates are plotted to scale · Dashed lines show the disclosed acquisition windows · The projected SK hynix finish around Oct. 20 is a media estimate assuming the current pace and share price hold1, 2
OTHER CORPORATIONS · AUG. 24–OCT. 2
+43.4tn
Largest net buyer across KOSPI and KOSDAQ · Foreign investors −KRW 31.5tn, individuals −KRW 16.5tn3
AVG. DAILY BUYBACK ÷ MEDIAN DAILY VOLUME
21% SK hynix
Samsung about 12% · Approximation for September (from Sep. 7)4, 5
SK HYNIX BUYBACK BUDGET REMAINING
6.3tn
84.2% completed by value · Daily request slowed to 400,000 shares2, 6
OCT. 8 · OPERATING PROFIT KRW 107.4tn (PRELIM.)
−2.42% Samsung
KOSPI −2.62% · Options expiry and ETF rebalancing hit on the same day7, 8

Since late August, the core of the heavy buying by the KRX 'other corporations' category was the two companies’ buybacks—Samsung for employee compensation, SK hynix for cancellation. Even after Samsung reported preliminary operating profit of KRW 107.4tn on Oct. 8, the KOSPI closed −2.62%; with options expiry and ETF rebalancing hitting at the same time, the effect of the buyback ending cannot be isolated cleanly. The variables that matter now are how fast foreign selling continues—and what domestic capital is willing to absorb it

Published Oct. 8, 2026 · Information cutoff: after the Korean market close on Oct. 8 · Includes Samsung Electronics’ preliminary 3Q results and Oct. 8 investor flows

snowshagal.com

AT A GLANCE

QUESTION
When the buyback bid stops, who takes its place?
THE BID SO FAR
Aug. 24–Oct. 2: other corporations +KRW 43.4tn · From Sep. 7, average daily company purchases were roughly 12% of Samsung’s median daily volume and 21% of SK hynix’s3, 5
NOT THE SAME CAPITAL
Both appear as 'other corporations' buying in investor-flow data, but Samsung’s KRW 15tn program funds employee stock compensation; SK hynix’s KRW 40tn program is to be fully cancelled1
PROGRESS
Samsung ended Oct. 6 · SK hynix has KRW 6.3tn left, has slowed to 400,000 shares a day, and is estimated to finish around Oct. 202, 6
HISTORY
Daishin Securities cases since 2020 · 33 periods with the same flow mix averaged +6.4% over the 60 trading days after buybacks ended · The median appears near 0% in the chart3, 5
10/8
Samsung 3Q operating profit KRW 107.4tn (prelim.) · KOSPI −2.62% · Foreign investors about −KRW 2.0tn, institutions about −KRW 1.7tn · Options expiry and ETF rebalancing coincided7, 8, 9
WHAT TO WATCH
The next bid is still unknown · Daily foreign net selling in the KOSPI; net buying by individuals and financial investment firms plus brokerage customer deposits; and how sharply other-corporation net buying falls · Filter out ETF rebalancing and expiry days before reading the signal
PAGE 01Who Sold—and Who Absorbed It
Who sold, who absorbed

KRW 43.4tn
The Buyer Behind the Flow

Net buying by investor type, KOSPI + KOSDAQ · Aug. 24–Oct. 2 · Daishin Securities compilation

From Aug. 24, when Samsung Electronics began its buyback, through Oct. 2, the largest net buyer was other corporations (+KRW 43.4tn) · Foreign investors and individuals were net sellers of KRW 31.5tn and KRW 16.5tn, respectively3

Same KRX category, very different capital
Samsung’s KRW 15tn is for employee stock compensation; SK hynix’s KRW 40tn is to be fully cancelled Both show up as buying by 'other corporations' in KRX flow data, but they are different forms of capital allocation1 · In Korea Capital Market Institute’s study of 282 KOSPI buyback decisions from 2024 through September 2026 (280 analyzed), cumulative abnormal returns over the three trading days around the announcement were 3.42% for cancellation-purpose buybacks and 0.59% for employee-compensation buybacks · Those figures measure the announcement effect, not what happens when the buying ends10
Net buying by investor type · Aug. 24–Oct. 2KRW tn · KOSPI + KOSDAQ
Other corporations +43.4 Fin. investment +7.3 Other institutions −2.7 Individuals −16.5 Foreign investors −31.5
All institutions: +KRW 4.6tn, including financial investment firms at +KRW 7.3tn · The five rows sum to zero · Daishin Securities / FnGuide compilation3
HOW TO READ IT
Other-corporation net buying and the two buyback programs are of the same order of magnitude But net buying is a net flow, while buyback value is gross executed purchase value, so the two cannot be subtracted from each other. The data support only the narrower statement that they are similar in scale—the same standard used in Note No.04. Daishin Securities and press reports identify the two buybacks as the main driver of other-corporation buying3, 11, 12 · Investor-category net flows do not identify the actual counterparty to each trade. They only show that, over the same period, one group was selling while other corporations were buying
Samsung Electronics8/24~10/2
SELLERS
Individuals −KRW 7.6tn · Foreign investors −KRW 7.7tn · Institutions ex-financial investment −KRW 2.9tn3
BUYERS
Other corporations about +KRW 14–15tn (read from chart) · Financial investment firms +KRW 3.8tn3
FULL WINDOW
Aug. 24–Oct. 6: foreign investors −KRW 8.2995tn and individuals −KRW 7.2932tn in Samsung, a combined KRW 15.6tn by this note’s calculation—close in scale to the KRW 15tn buyback, though the measures are not directly comparable2, 5
SK hynix8/20~10/2
SELLERS
Individuals −KRW 13.8tn · Foreign investors −KRW 16.4tn · Institutions ex-financial investment −KRW 2.0tn · Financial investment firms −KRW 0.4tn3
BUYERS
Other corporations about +KRW 32tn (read from chart)3

Market-wide and Samsung flows are measured from Aug. 24; SK hynix stock-level flows start on Aug. 20, when its buyback began. Do not add the two stock-level other-corporation figures and try to reconcile them with the KRW 43.4tn market total: the original Daishin chart uses different start dates, so the periods cannot be put on a single like-for-like basis3

HOW LARGE WAS THE SUPPORTING BID? · SEPTEMBER (FROM SEP. 7)
about 12%Samsung Electronics · Avg. daily buyback about 2.0m shares ÷ median daily volume 16.08m
about 21%SK hynix · Avg. daily buyback about 622,000 shares ÷ median daily volume 2.95m

This note’s calculation · Average buyback volume is inferred from cumulative progress reports between Sep. 4 and Sep. 30/Oct. 1; trading volume uses Investing.com daily data and the median September trading day. NXT inclusion is not specified, so these are approximations4, 5, 13, 14, 15

PAGE 01 · TAKEAWAYS
  1. Since late August, the core of other-corporation buying (+KRW 43.4tn) that absorbed heavy net selling was the two companies’ buybacks · From Sep. 7, average daily buyback volume was roughly 12% and 21% of median daily trading volume, respectively
  2. The same 'other corporations' category masks different uses of capital: Samsung’s program funds compensation; SK hynix’s program retires shares
  3. The selling was led by foreign investors and individuals · Financial investment firms bought alongside Samsung, but not SK hynix
PAGE 02What Is Finished—and What Remains
What ended, what remains

The Finish Line Was Set by
Value, Not Shares

Samsung ended Oct. 6 · SK hynix 84.2% complete by value · Note No.04 estimate vs. actual

By share count, Samsung Electronics had already exceeded its disclosed target on Oct. 1, but purchases continued through Oct. 6 · The reason is simple: Samsung disclosed a KRW 15tn purchase budget, not a hard share-count ceiling2, 14, 16

Progress Against the Disclosed TargetsAs of Oct. 6 · 100% = disclosed target
Samsung Electronics · KRW 15tn Ended Oct. 6 Value 100% Share count 106.7% SK hynix · KRW 40tn KRW 6.3tn remaining Value 84.2% Share count 80.0% Disclosed 100%
Samsung: 56,856,305 shares · KRW 14.9999tn executed (completed Oct. 6) · SK hynix: 19.25m shares · KRW 33.6895tn executed · Hatched areas show the remaining budget and the range to the reported final share-count estimate of about 22.9m shares, or roughly 95% of the original share target · Share-count percentages are calculated in this note1, 2, 5

Note No.04 Estimate vs. What Actually Happened

Samsung ElectronicsNOTE NO.04: OCT. 1 → ACTUAL: OCT. 6
ESTIMATE
Note No.04 divided the remaining disclosed share count by the average daily purchase pace and arrived at Oct. 111
ACTUAL
The share-count estimate was right in one sense: Samsung reached 100.96% of the disclosed share count on Oct. 1. But because the program was budget-based, the company kept buying for two more trading days—Oct. 2 and Oct. 6—and finished on Oct. 62, 14
WHY
The average purchase price, KRW 263,823, was 6.3% below the KRW 281,500 reference price used at the board resolution. That let Samsung buy 6.7% more shares with the same KRW 15tn budget2, 5
SK hynixNOTE NO.04: OCT. 16 → MEDIA ESTIMATE: AROUND OCT. 20
ESTIMATE
Using the same method, Note No.04 estimated Oct. 1611
PRICE EFFECT
SK hynix’s average purchase price is about KRW 1.75m by this note’s calculation, above the KRW 1.662m reference price. That reduces the number of shares needed to spend the budget to roughly 22.9m. On price alone, the program would have ended sooner1, 2, 5
WHAT DELAYED IT
The company slowed the daily purchase request from 600,000–650,000 shares to 500,000 and then 400,000 in October · KRW 6.3149tn remains2, 6

The binding constraint was the amount of money, not the number of shares. When the market price is below the reference price, the same budget buys more shares—as with Samsung. When it is above the reference price, it buys fewer—as with SK hynix. Note No.04 therefore called Samsung’s finish too early on a share-count basis; for SK hynix, the October slowdown mattered more than the price effect

What Does Not Immediately Become Supply When the Buying Stops

KNOWN

Confirmed by filings and executions

  • All shares acquired by SK hynix are scheduled to be cancelled · Once cancelled, they do not return to the tradable float1, 3
  • Samsung’s buyback is intended to fund employee stock compensation, including performance incentives and special management-performance bonuses1
  • Samsung plans a roughly KRW 30tn 3Q cash dividend, including the regular quarterly dividend · The board is expected to finalize it at the end of October (Aug. 21 disclosure)17
  • On Oct. 7, the first day after Samsung’s buyback stopped, other corporations were net sellers of KRW 7bn in Samsung, versus average daily net buying of KRW 507.3bn during the purchase period2
ESTIMATE

Numbers built on assumptions

  • SK hynix could finish around Oct. 20 if the current pace and share price hold, according to press estimates2
  • Estimated daily buying while the program remains: about KRW 0.7tn · 400,000-share purchase request × Oct. 7 close of KRW 1.723m5, 6, 18
  • A possible additional SK hynix buyback in 4Q appears as an unattributed line in an article discussing an IBK Investment & Securities report · It is not a company announcement19
UNKNOWN

Still unknown

  • When Samsung’s compensation shares will be granted or sold, and whether they will carry transfer restrictions · Restricted stock awarded in January this year carried a one-year lockup20
  • Samsung’s remaining shareholder-return plan · The company estimates roughly KRW 90–110tn of 2026 shareholder-return capacity remains. About KRW 30tn of 3Q cash dividends is due to be finalized at the end of October; the size and form of the remainder are scheduled for a board decision at the end of January 202717
  • When foreign selling will ease

On Samsung’s capital-allocation calendar, the sequence is: KRW 15tn compensation buyback completed Oct. 6 → roughly KRW 30tn dividend finalized at end-October → remaining shareholder return decided at end-January 2027. A dividend is not open-market buying, so it does not directly replace the bid that just disappeared2, 17

A buyback ending is not the same thing as new sell supply appearing. Shares earmarked for cancellation disappear; compensation shares cannot be sold before they are granted, and any transfer restriction pushes potential selling out further. Whether the shares from this KRW 15tn program will carry such restrictions has not been confirmed3, 20

PAGE 02 · TAKEAWAYS
  1. Samsung finished later than the share-count estimate because the program was defined by value; SK hynix is likely to finish later than Note No.04 projected because it slowed the pace
  2. Samsung’s buying is finished. SK hynix still has roughly two weeks of demand left at about KRW 0.7tn per day, based on the current requested volume
  3. A buyback ending does not instantly create new sell supply · What disappears first is buy demand
PAGE 03Average vs. Median
Average vs median

The Average Is +6.4%,
but the Distribution Is All Over the Map

Daishin Securities · Buyback cases since 2020 · 33 periods with the same flow mix · 60 trading days after completion

Daishin Securities found that in 33 periods when financial investment firms were net buyers while other institutions were net sellers—the flow mix now seen in Samsung—the average return over the 60 trading days after a buyback ended was +6.4%, better than the +0.1% average in 34 periods when both groups were net sellers. But the same chart shows the median only as a bar, without a printed figure, and it appears to sit around 0% by visual read3, 5

HOW TO READ IT
When the average is far above the median, a small number of large winners may be doing much of the work Daishin’s original PDF shows the median only graphically and does not make clear whether that chart covers all 285 observations or the 105 non-financial direct-buyback observations. This note therefore uses the median only as a directional check, not as a precise statistic
60-Trading-Day Return After Buyback Completion, by Flow Mix% · Mean vs. median
SK hynix Samsung Electronics Mean Median +6.4 ≈0 ≈−2.1 ≈−8.6 +0.1 ≈−0.6 ≈+5.0 ≈+2.6 Fin. inv. buyOther inst. sellFin. inv. buyOther inst. buyFin. inv. sellOther inst. sellFin. inv. sellOther inst. buy
'Institutions' excludes financial investment firms · +6.4% and +0.1% are figures stated in the source; values marked ≈ are approximate readings of Daishin Securities Figure 7 · The 33 and 34 periods split cases in which other institutions were net sellers according to the direction of financial-investment flows · The source does not clearly state whether Figure 7 covers all 285 observations or the 105 non-financial direct-buyback cases · SK hynix’s financial-investment selling was modest at −KRW 0.4tn3, 5

What the Historical Cases Can—and Cannot—Tell Us

The evidence that completion itself determines direction is weak. In the 34 periods when both groups sold, the mean was +0.1% and the median about −0.6%—both essentially flat. SK hynix currently fits that mix, with financial investment firms at −KRW 0.4tn and other institutions at −KRW 2.0tn. Even Samsung’s apparently stronger setup has a median near zero3, 5 · Daishin itself makes the outlook conditional not on the fact that the buyback has ended, but on whether financial-investment buying persists and whether earnings and shareholder-return policy become more concrete3

COUNTERPOINT · The Flow Gap May Be Smaller Than It Looks

Daishin argues that institutional sellers may already have cleared much of their inventory during the buyback, leaving a lighter overhang afterward3 · Eugene Investment & Securities also argues that fears of a post-buyback flow vacuum are unlikely to be realized21 · Shinhan Investment Corp. described Samsung’s preliminary 3Q operating profit of KRW 107.4tn as slightly above consensus, while an FnGuide consensus reported on Oct. 5 was KRW 110.2803tn—so the verdict depends on which benchmark is used7, 8, 22 · Samsung’s foreign ownership at 46.41%, below its 10-year average of 52.92%, can be read as suggesting less room for incremental foreign selling than in the past. It does not mean the selling is over22 · Foreign KOSPI net selling on Oct. 6, the completion day, and Oct. 7 (including NXT) was larger than the September daily average. Foreign investors also sold roughly KRW 2tn on Oct. 8, though that session was distorted by expiry and rebalancing. So far, the flow data have not confirmed the counterargument9, 18, 23

PAGE 03 · TAKEAWAYS
  1. The historical flow mix that resembles Samsung averaged +6.4% after completion, but the median appears close to 0%
  2. The historical cases do not make a strong case for either a post-buyback drop or a post-buyback rebound
  3. The question is not simply that the buyback ended. It is who buys next
PAGE 04The Next Bid
Who takes the next bid

Three Numbers That May Reveal
Who Buys Next

Three days since Samsung finished · Filter mechanical flows · Three indicators · A three-week calendar

On Oct. 8, the day Samsung Electronics reported preliminary 3Q operating profit of KRW 107.4tn, the KOSPI fell 2.62%; Samsung closed −2.42% and SK hynix −2.44%7, 8 · Semiconductor ETF rebalancing, monthly options expiry, risk reduction ahead of a market holiday, and profit-taking after the earnings print all landed on the same session, making it difficult to isolate any buyback-completion effect8, 24, 25 · The same reports also cited weaker buyback demand as one contributor8, 9

Three Sessions Since Samsung’s Buyback Ended
  • OCT. 6 · KOSPI −0.89% · Samsung’s final buyback session · Foreign investors −KRW 1.7672tn · KOSDAQ rose23
  • OCT. 7 · KOSPI −1.98% · Individuals +KRW 3.2616tn, foreign investors −KRW 3.1095tn (including NXT) · Samsung −1.29%, SK hynix −2.82% · KOSDAQ, which was not directly exposed to Samsung’s buyback ending, also closed −2.34%18 · Foreign investors sold a net KRW 24bn of Samsung Electronics26
  • OCT. 8 · KOSPI −2.62% · Foreign investors about −KRW 2.0tn, institutions about −KRW 1.7tn, individuals about +KRW 2.9tn, other corporations +KRW 754bn · KOSDAQ −0.69% · Samsung’s preliminary-results day, ETF rebalancing day and monthly options expiry9

The market fell for three straight sessions, but each day carried variables unrelated to the buyback ending. The data bases also differ: Oct. 7 includes NXT; inclusion is unspecified for Oct. 6 and Oct. 8

KOSPI Net Buying by Investor Type · Oct. 8KRW tn · InfoStockDaily report
Individuals +2.91 Other corporations +0.75 Institutions −1.67 Foreign investors −1.99
InfoStockDaily report; data provider and NXT inclusion not specified · The four investor categories sum to approximately zero · Foreign net-selling figures differ slightly across reports: KRW 1.9921tn, KRW 1.9885tn citing KRX, and KRW 1.9902tn · This note does not use a separate figure said to include NXT because the underlying source could not be checked8, 9, 24

Even on the earnings day, foreign investors and institutions sold while individuals and other corporations bought. The +KRW 754bn from other corporations is close in scale to the roughly KRW 672bn of SK hynix purchases implied by assuming the same 400,000-share request as Oct. 6–7 at the Oct. 8 close. But the 'other corporations' bucket also contains buybacks by other listed companies and ordinary corporate trading, so the figures cannot be treated as the same buyer; the comparison is only for scale5, 6, 8, 9

EVENT FILTER · A DAY TO READ WITH CAUTION

On Oct. 8, regularly scheduled rebalancing across 24 domestic equity ETFs worth KRW 23.2tn was expected to concentrate in one session because of the holiday calendar; KRW 18.1tn of that total was semiconductor-related. Samsung Securities estimated that the process could create roughly KRW 200bn of Samsung Electronics selling and KRW 90bn of SK hynix buying (reported Oct. 6)25 · Monthly options expiry also fell on the same day, and press reports cited both events as contributors to the decline8, 9 · Daishin Securities noted that financial investment firms were also net sellers that day, alongside foreign investors and institutions24 · For that reason, this note does not treat net buying by financial investment firms or institutions on ETF rebalancing days, futures/options expiry days, or scheduled index-reconstitution days as a directional signal; foreign net selling on those days is also used only as context. The flows are still recorded, but the signal is checked again using a two- to three-session average after the event has passed—for Oct. 8, that means Oct. 12–14

SCALE CHECK
The bid fading out · September KOSPI other corporations +KRW 31.3992tn Potential replacement capital · September domestic equity funds +KRW 10.1tn; brokerage customer deposits +KRW 1.9tn as of Sep. 23 · These are measured on different bases and are shown only to establish scale, not as a direct offset12, 21
NUMBER 1 · FOREIGN INVESTORS

Daily Foreign Net Selling in the KOSPI

NOWSelling above the September average on both non-filter days
  • Baseline · September average about −KRW 1.1tn per day (−KRW 21.5108tn ÷ 20 trading days; this note’s calculation)5, 12
  • Current · Oct. 6 −KRW 1.7672tn; Oct. 7 −KRW 3.1095tn including NXT · Both non-filter days were worse than the baseline18, 23 · Oct. 8 was about −KRW 2.0tn, but because it was a filter day this note records the figure and waits for the Oct. 12–14 average9
  • Kiwoom Securities argued before the release that investors should watch the immediate share-price reaction, foreign flows and changes to 2027 earnings estimates more closely than whether preliminary earnings beat consensus27 · Eugene Investment & Securities argues that foreign selling is more likely to ease once the rise in U.S. yields loses momentum21
  • Watch · Does foreign net selling fall below the baseline after the Oct. 14 U.S. CPI?
NUMBER 2 · DOMESTIC CAPITAL

Individual and Financial-Investment Buying, Plus Brokerage Cash

NOWIndividuals bought about KRW 3tn on each of two consecutive sessions
  • Individuals: Oct. 7 +KRW 3.26tn including NXT; Oct. 8 +KRW 2.91tn · Net buyers on both days9, 18
  • September through Sep. 23: brokerage customer deposits +KRW 1.9tn; domestic equity funds +KRW 10.1tn versus bank deposits +KRW 43.6tn · Eugene notes that a 3.0% policy rate is close to 2.9% inflation, limiting the absolute appeal of deposits21
  • Daishin Securities treats continued buying by financial investment firms as a sign of market interest; they bought KRW 3.8tn during Samsung’s buyback period3 · That category also contains ETF creations/redemptions and arbitrage activity, so this note excludes the filter days above when reading it as a signal
  • Watch · Brokerage customer-deposit trend and the Oct. 22 Bank of Korea meeting
NUMBER 3 · OTHER CORPORATIONS

How Fast Does Other-Corporation Net Buying Fall?

NOWSK hynix is still buying · Estimated finish around Oct. 20
  • Oct. 8 KOSPI other corporations +KRW 754bn9 · While SK hynix’s program remains active, its implied contribution is about KRW 0.7tn per day based on the requested share count5, 6
  • Around Oct. 20, roughly KRW 0.7tn a day of SK hynix buyback demand could disappear, if the current pace holds. That would not eliminate buyback demand from the 'other corporations' category altogether: on Oct. 6, 49 other KOSPI stocks besides Samsung and SK hynix also submitted buyback purchase requests2, 6
  • As Note No.04 argued, the daily other-corporation flow is likely to reveal the practical end of the buying before the formal completion report does11
  • What comes next · Samsung expects to finalize roughly KRW 30tn of dividends at an end-October board meeting, then decide the remaining shareholder return at the end of January 202717 · The possible 4Q SK hynix buyback appears only as an unattributed sentence in an article discussing an IBK report; it is not a company announcement19
MARKET CLOSED
10.09FRI
Hangul DayKorean market closed28
Foreign investors
10.14WED
U.S. September CPI21:30 KST · Number 129
ESTIMATE
10.20AROUND
SK hynix buyback endsAssumes current pace and share price hold · Number 32
KOREA
10.22THU
Bank of Korea policy meetingDeposits vs. equities · Number 230
RATES
10.28WED
FOMC decision03:00 KST on Oct. 2931
SHAREHOLDER RETURN
OCT.END
Samsung Electronics boardExpected to finalize roughly KRW 30tn 3Q dividend · Date not yet set17

These three numbers are the monitoring framework used in this note, not an index forecast · The signal becomes more meaningful when all three point in the same direction

PAGE 04 · TAKEAWAYS
  1. The KOSPI fell for three straight sessions after the buyback ended, but KOSDAQ also fell on Oct. 7 and expiry/rebalancing distorted Oct. 8, so the completion effect cannot be cleanly isolated
  2. Individuals absorbed the selling for two consecutive sessions; foreign net selling was larger than the September average on both non-filter days, Oct. 6 and Oct. 7
  3. The next test is around Oct. 20, when roughly KRW 0.7tn a day of SK hynix buyback demand may also disappear
CLOSING NOTE
One Bid Has Stepped Away.
The Next One Is Still Unclear

The buybacks were undeniably large sources of demand, so their disappearance changes the flow structure. But price direction is not determined by one missing buyer; it is the net result of foreign, individual, institutional and passive flows arriving on the same day

DECISION RULES
  1. Does daily foreign KOSPI net selling fall below the September average of roughly KRW 1.1tn?
  2. After roughly KRW 0.7tn a day of SK hynix buying disappears, do individuals and financial investment firms repeatedly fill the gap?

Record ETF-rebalancing and expiry-day flows, then check the two- to three-session average afterward

This document examines the flow structure around the end of large corporate buybacks · It is not a forecast of any stock or index and is not investment advice · Item-specific dates, calculations and estimates are documented in the sources below

GLOSSARYQuick Glossary
Other corporations
A KRX investor category for ordinary corporations that are not financial institutions, pension funds, foreign investors or individuals · Listed companies buying back their own shares are recorded here
Share buyback
A company purchasing its own shares · After board approval, the company discloses the size and purchase window and files a completion report when the program ends · Both programs in this note were defined by purchase value, so the final share count changes with the market price
Share cancellation
Retiring repurchased shares so they are removed from shares outstanding · Cancelled shares do not return to the tradable float
Restricted Stock Award (RSA)
Equity compensation granted to employees with restrictions that prevent sale for a specified period
Financial investment firms
A KRX institutional-investor subcategory that includes securities firms’ proprietary trading · Daishin Securities interprets net buying by this group as a sign of market interest and cumulative selling by other institutions as evidence that their inventory overhang has eased
Average vs. Median
The mean adds all observations and divides by the count; the median is the middle observation when values are ranked · A few large outcomes can pull the mean far above the median
Liquid bank deposits
Demand and savings deposits that are readily available for use · In Korean market commentary, this measure is often watched for signs that money is moving back into bank deposits rather than risk assets
Brokerage customer deposits
Cash held in securities accounts and available for stock purchases · A commonly watched gauge of retail buying capacity
Cumulative abnormal return (CAR)
A stock’s return in excess of the market return, accumulated across several days around an event · Used to gauge how the market reacted to an announcement
ETF rebalancing and options expiry
Rebalancing days generate trades to reset index and ETF weights; options-expiry days can concentrate arbitrage unwinds · Because these are mechanical flows that need not express a directional view, this note filters financial-investment and institutional flows on those sessions
NXT
Korea’s alternative trading system, Nextrade · Investor-flow figures can differ slightly depending on whether NXT trades are included
SOURCES (31)

Information cutoff: after the Korean market close on Oct. 8, 2026 · Includes Oct. 8 closes and investor flows plus Samsung Electronics’ preliminary 3Q results · Published Oct. 8, 2026

  • REFERENCE DATEVaries by item · The relevant date is shown at the end of each source entry
  • CHART READINGSStock-level cumulative other-corporation flows, historical medians and several flow-mix averages are read from Daishin Securities charts
  • ESTIMATESK hynix finish around Oct. 20 and about KRW 0.7tn of daily buying · Assumes the current pace and share price hold
  • THIS NOTE’S CALCULATIONSShare-count progress, average purchase prices, combined KRW 15.6tn net selling, September daily average foreign selling, buyback volume equal to 12%/21% of median volume, and the Oct. 8 implied KRW 672bn SK hynix purchase · Formulas are documented in Source 5
  • DATA DIFFERENCESInvestor net-flow figures differ by several billion won across reports because the underlying data provider and treatment of NXT are not always the same
  • NOT A COMPANY ANNOUNCEMENTPossible additional SK hynix buyback in 4Q, cited in an article discussing an IBK report · Samsung’s dividend size and timing are company forward-looking information
1. Edaily 2026.08.21 · Samsung Electronics treasury-share acquisition disclosure · KRW 15tn · 53,285,968 common shares · Acquisition period Aug. 24–Nov. 21, 2026 · Open-market purchases · Purpose: employee stock compensation including performance incentives and special management-performance bonuses / Financial News 2026.08.19 · SK hynix treasury-share acquisition disclosure · KRW 40.00434tn · 24.07m common shares, about 3.3% of shares outstanding · Aug. 20–Nov. 19, 2026 · All acquired shares to be cancelled after completion · Amount calculated from the Aug. 18 close of KRW 1.662m / Samsung Electronics IR disclosure Decision to Acquire Treasury Shares Aug. 21, 2026 · Estimated acquisition amount KRW 14,999,999,992,000 · 53,285,968 shares · Aug. 24–Nov. 21, 2026 · Open-market purchases through Samsung Securities, Shinhan Securities and KB Securities · 'The company plans to acquire treasury shares based on the amount stated under Item 2, Estimated Acquisition Amount, and the actual number of shares acquired may change with future share-price movements.'
2. Woman Times (Seong Min) · Samsung Completes KRW 15tn Buyback Early; SK hynix Slows the Pace Oct. 7, 2026 · Based on KIND execution records · Samsung bought 56,856,305 shares over 29 trading days from Aug. 24 to Oct. 6 for KRW 14,999,999,899,750 · Average purchase price KRW 263,823 (−6.3% versus the KRW 281,500 close on the board-resolution date) · During the purchase period, foreign investors sold a net KRW 8.2995tn and individuals KRW 7.2932tn; the article says the buyback 'absorbed 96%' of that selling, a comparison of gross acquisition value with net selling that this note does not use in the main text · Completed about 46 days before the Nov. 21 disclosed deadline · SK hynix bought 19.25m shares for KRW 33.6895tn from Aug. 20 to Oct. 6, equal to 84.2% of the KRW 40.0043tn planned amount · KRW 6.3149tn remains · Daily buying slowed from 600,000–650,000 shares to 500,000 and then 400,000 in October · At the current pace and share price, the article estimates completion around Oct. 20 · Final share count estimated at about 22.9m, more than 1.1m below the original plan, reducing the cancellation ratio from 3.3% to 3.1% · On the first day without Samsung buying, other corporations were net sellers of KRW 7bn in Samsung, versus average daily net buying of KRW 507.3bn during the buyback period
3. Daishin Securities, 'Soon to Know · Who Sold Into All Those Semiconductor Buybacks?' Oct. 7, 2026 (Soon-ho Kwon, Quant) · Buyback progress by value as of Oct. 6: SK hynix 83%, Samsung Electronics 97% (Samsung had completed the originally planned share count) · Aug. 24–Oct. 2 KOSPI + KOSDAQ net flows: other corporations +KRW 43.4tn, foreign investors −KRW 31.5tn, individuals −KRW 16.5tn, all institutions +KRW 4.6tn (financial investment firms +KRW 7.3tn; other institutions −KRW 2.7tn) · Samsung purchase window, Aug. 24–Oct. 2: individuals −KRW 7.6tn, foreign investors −KRW 7.7tn, other institutions −KRW 2.9tn, financial investment firms +KRW 3.8tn · SK hynix purchase window, Aug. 20–Oct. 2: individuals −KRW 13.8tn, foreign investors −KRW 16.4tn, other institutions −KRW 2.0tn, financial investment firms −KRW 0.4tn · Stock-level cumulative other-corporation figures in Figures 3 and 4 are read from charts because exact values are not printed · Since 2020, 285 observed buyback periods among top-300 market-cap stocks; semiconductor comparison focuses on 105 direct buybacks by non-financial companies · In 33 periods with financial-investment net buying and other-institution net selling, the average 60-trading-day return after completion was 6.4%, versus 0.1% in 34 periods when both groups sold · Seven periods in which other institutions accounted for at least 50% of selling while financial investment firms bought averaged 22.5%; four were auto-related and dispersion was wide · Flow-mix means and medians in Figure 7 and 20-trading-day post-completion results by seller type in Figure 5 are read from charts because exact figures are not printed · 'The end of a buyback means reduced buying capacity, but it does not automatically mean an increase in sell supply' · The report argues that after preliminary earnings, sensitivity may shift toward earnings durability and concrete shareholder-return policy · The source labels the progress date as Oct. 6, but Samsung’s 97% figure appears to reflect executions only through Oct. 2 · This note uses the Oct. 6 KIND execution figures reported by Woman Times for current progress
4. Investing.com daily data · Samsung Electronics · SK hynix · Median September daily volume: Samsung Electronics 16.08m shares (Sep. 8–30, 15 trading days) · SK hynix 2.95m shares (Sep. 7–30, 16 trading days) · NXT inclusion not specified · Some closing prices differ from exchange closes, so only volume is used here as an approximation
5. This Note’s Calculations · During Samsung’s Aug. 24–Oct. 6 buyback, foreign and individual net selling in the stock totaled KRW 8.2995tn + KRW 7.2932tn = KRW 15.5927tn, similar in scale to the KRW 15tn acquisition value but measured on a different basis: acquisition value is gross executed buying, while net selling is a net flow · Samsung final share count 56,856,305 ÷ disclosed 53,285,968 = 106.7%, meaning the same KRW 15tn bought 6.7% more shares · SK hynix share-count progress 19.25m ÷ 24.07m = 80.0%; versus the reported final estimate of about 22.9m shares, current purchases equal about 95.1% · SK hynix average purchase price KRW 33.6895tn ÷ 19.25m shares ≈ KRW 1.75m · Daily SK hynix purchase scale: 400,000-share request × Oct. 7 close KRW 1.723m ≈ KRW 689.2bn · Remaining KRW 6.3149tn ÷ KRW 689.2bn ≈ 9.2 trading days; counting from Oct. 7 and allowing for the Oct. 9 holiday implies Oct. 20–21 · Had the 600,000–650,000 daily pace continued, the roughly 3.65m shares remaining to the estimated 22.9m final total would take about six trading days, around Oct. 14–15 after allowing for the Oct. 9 holiday · September foreign KOSPI net selling KRW 21.5108tn ÷ 20 trading days, excluding Sep. 24–25 holidays, ≈ KRW 1.08tn per day · Figure 7 chart readings: financial-investment buy / other-institution sell mean +6.4% (source) and median about 0%; buy / buy mean ≈−2.1%, median ≈−8.6%; sell / sell mean +0.1% (source), median ≈−0.6%; sell / buy mean ≈+5.0%, median ≈+2.6% · Average September daily buyback volume: Samsung (53.80m − 19.80m) ÷ 17 trading days from Sep. 7 to Oct. 1 ≈ 2.0m shares; SK hynix (17.75m − 7.80m) ÷ 16 trading days from Sep. 7 to Sep. 30 ≈ 622,000 shares · Median September daily volume from Investing.com: Samsung 16.08m shares, SK hynix 2.95m shares · Buyback as a share of volume: Samsung 2.0m ÷ 16.08m ≈ 12.4%; SK hynix 622,000 ÷ 2.95m ≈ 21.1%; approximate, and buyback shares are themselves part of reported trading volume · Implied Oct. 8 SK hynix purchase: 400,000 shares, assuming the same requested amount as Oct. 6–7, × Oct. 8 close KRW 1.681m ≈ KRW 672.4bn · Oct. 8 sum of four investor groups: +KRW 2.9061tn + KRW 754bn − KRW 1.6714tn − KRW 1.9921tn ≈ −KRW 3.4bn
6. NewsTNT · '[Disclosure] Oct. 6: Buyback Purchase Requests for 51 KOSPI Stocks' 10/6 KOSPI 51 stocks, including Samsung Electronics and SK hynix · Samsung direct request 1,058,790 shares; SK hynix 400,000 shares · '[Disclosure] Oct. 7…' 10/7 SK hynix direct request 400,000 shares · Requested quantities may differ from actual executions
7. Samsung Newsroom · Samsung Electronics Announces Preliminary 3Q 2026 Results Oct. 8, 2026 · Consolidated revenue KRW 195tn, +13.70% QoQ and +126.59% YoY · Operating profit KRW 107.4tn, +20.01% QoQ and +782.50% YoY · Preliminary K-IFRS figures and may differ from final results
8. Good Morning Economy · Oct. 8, 2026 article · Samsung Electronics KRW 262,000 (−2.42%) · SK hynix KRW 1.681m (−2.44%) · Foreign investors −KRW 1.9885tn citing KRX · Reported reasons for the decline: profit-taking, options expiry, ETF rebalancing, risk reduction ahead of a market holiday, and weaker buyback demand · Shinhan Investment Corp.’s Jin-hyuk Kang: results were slightly above consensus, but options expiry and ETF rebalancing weighed on the session · Eugene Investment & Securities’ Jun-young Lee: lower buyback demand and lighter positioning ahead of the holiday
9. InfoStockDaily · [Oct. 8 Closing Check] Oct. 8, 2026 · KOSPI 6,625.93 (−2.62%), third consecutive decline · KOSDAQ 892.27 (−0.69%) · KOSPI investor flows, with data provider and NXT inclusion unspecified: foreign investors −KRW 1.9921tn, institutions −KRW 1.6714tn, individuals +KRW 2.9061tn, other corporations +KRW 754bn · Article cites ETF rebalancing, options expiry and concern about buyback completion as reasons for the decline
10. Korea Capital Market Institute, Jong-eun Lee · 'Recent Characteristics of Listed-Company Share-Buyback Disclosures and Market Reactions' Capital Market Focus 2026-20 · Oct. 5, 2026 · 282 KOSPI buyback decisions from 2024 through Sep. 28, 2026; 280 analyzed and 278 with valid CAR observations · Cumulative abnormal return CAR(−3,+3) around the announcement: 4.13% overall, 3.42% for cancellation-purpose buybacks, 0.59% for employee-compensation buybacks · Under the amended Commercial Act, shares are in principle to be cancelled within one year after acquisition, except where a retention/disposal plan is approved by shareholders
11. Snowshagal INVESTMENT NOTE NO.04, 'The Supporting Bid Ends First,' Sep. 19, 2026 · Using Sep. 18 progress of 74.7% for Samsung and 58.8% for SK hynix plus actual average daily share purchases, the note estimated depletion on Oct. 1 for Samsung and Oct. 16 for SK hynix · Share-count basis, assuming the then-current pace continued
12. Seoul Economic Daily · Foreign Investors Sold KRW 21.5tn Again in September—Samsung and SK hynix Led the Selling Sep. 30, 2026 · KRX September KOSPI data: foreign investors −KRW 21.5108tn; individuals −KRW 14.237tn; institutions +KRW 4.4178tn; other corporations +KRW 31.3992tn, influenced by Samsung and SK hynix buybacks · Foreign selling: SK hynix KRW 11.5805tn and Samsung KRW 4.892tn, together 76.6% of foreign KOSPI net selling · The two stocks’ share of KOSPI market cap rose from 48.81% to 50.85% during September / same KRX figures cited by Korea Times, Oct. 1, 2026 Korea Times 2026.10.01
13. SmartToday · Sep. 6, 2026 article · Through Sep. 4, Samsung had bought 19.80m shares over 10 trading days and SK hynix 7.80m shares over 12 trading days
14. Segye Ilbo (Eun-a Song) · Samsung and SK hynix Buybacks Set to End in October—What Comes Next for the KOSPI? Oct. 4, 2026 · Through Oct. 1, Samsung had bought 53.80m shares, 100.96% of the disclosed share count, for KRW 14.1565tn · Through Oct. 2, SK hynix had bought 18.35m shares, 76.24%; figures appear to include executions through Oct. 1, for KRW 32.0514tn · Daily purchases were cut from 600,000–650,000 shares to 500,000 on Oct. 2, with requested volume lowered to 400,000 · Oct. 2 KOSPI 7,003.74
15. Financial News · Oct. 1, 2026 article · SK hynix had bought 17.75m shares, 73.74%, through Sep. 30 · Samsung had bought 53.80m shares through Oct. 1 · Article explains that the programs are value-based
16. Sisa Journal-e (Yong Lee) · Samsung’s KRW 15tn Buyback to End on Oct. 6 Oct. 2, 2026 · Exchange execution data through Oct. 1: Samsung 53.80m shares for about KRW 14.1565tn, already above the disclosed 53,285,968-share figure · Samsung disclosure: 'The company plans to acquire treasury shares based on the stated estimated acquisition amount' · Aug. 24–Oct. 1, 27 trading days: foreign investors −KRW 7.6141tn; individuals −KRW 7.0249tn
17. Samsung Electronics IR disclosure · Expected 2026 Shareholder-Return Size and Implementation Plan Aug. 21, 2026 · 'For the third quarter of 2026, the company plans to pay approximately KRW 30tn in cash dividends, including the regular quarterly dividend' · To be finalized at an end-October board meeting · The board is scheduled to decide the remaining shareholder return at the end of January 2027 · Estimated remaining 2026 shareholder-return capacity about KRW 90–110tn, based on 50% of 2024–2026 FCF less KRW 29.3tn already implemented in 2024–2025 · The disclosure refers to what remains after the roughly KRW 30tn dividend as 'the remaining shareholder return' · Company forward-looking information
18. NewsPim · [Closing Market] KOSPI Ends Lower Near 6,803 as Investors Wait for Samsung Earnings Oct. 7, 2026 · KOSPI 6,803.90, −137.49 points or −1.98% · KOSDAQ 898.43 (−2.34%) · KOSPI including NXT: individuals +KRW 3.2616tn, foreign investors −KRW 3.1095tn, institutions −KRW 859.8bn · Samsung KRW 268,500 (−1.29%) · SK hynix KRW 1.723m (−2.82%) · USD/KRW 1,340.4
19. Alpha Economy · SK hynix Could Conduct an Additional Buyback in 4Q Oct. 7, 2026 · Cites IBK Investment & Securities analyst Woon-ho Kim · Refers to the Aug. 19 shareholder-return policy of returning at least 50% of FCF after CAPEX · The sentence 'an additional share buyback is expected to proceed during 4Q' appears as unattributed reporter narrative after the policy discussion · Not an official company announcement
20. Business Watch (Jun-hyung Ahn) · Samsung to Add KRW 15tn of Share-Based Compensation After KRW 2.6tn This Year Aug. 26, 2026 · The KRW 15tn buyback is intended to fund stock compensation including performance incentives and special management-performance bonuses · Restricted Stock Awards of 7,748,354 shares granted in January as part of excess-profit incentive compensation carry a one-year transfer restriction, two years for president-level executives · 35,292,600 PSU shares are to vest over three years from October 2028 · The size of special management-performance bonuses for the DS division will be set after business performance is finalized
21. Eugene Investment & Securities, 'Equity Strategy · Is the Flow Vacuum a False Alarm—or a Real Risk?' Oct. 6, 2026 (Jae-hwan Huh) · Since Aug. 21, other corporations had net bought more than KRW 42tn · Jan.–Sep. 2026 other-corporation net buying KRW 58.0tn versus KRW 11.4tn in 2025 · September through Sep. 23: bank deposits +KRW 43.6tn, including demand deposits +KRW 29.9tn and savings deposits +KRW 13.7tn; domestic equity funds +KRW 10.1tn; brokerage customer deposits +KRW 1.9tn; MMFs +KRW 8.8tn; bond funds +KRW 2.3tn · With the Bank of Korea policy rate at 3.0% and September inflation at 2.9%, Eugene argues the real policy rate is barely above zero · It attributes foreign selling to the speed of the rise in U.S. yields and expects pressure to ease if that rise slows · Concludes that a serious post-buyback flow vacuum is unlikely
22. Financial News · Samsung, SK hynix Buybacks Near the Finish—Can KRW 190tn of Earnings Fill the Flow Gap? Oct. 5, 2026 · September Samsung flows: individuals −KRW 8.622tn, foreign investors −KRW 4.8922tn, other corporations +KRW 10.2964tn · SK hynix: individuals −KRW 9.688tn, foreign investors −KRW 11.5844tn, other corporations +KRW 21.9808tn · FnGuide 3Q operating-profit consensus: Samsung KRW 110.2803tn, SK hynix KRW 78.0577tn · The two stocks represented about 51.2% of KOSPI market cap · Semiconductor ETF rebalancing on Oct. 8 · Samsung foreign ownership 46.41% versus 10-year average 52.92% · SK hynix earnings date not yet announced
23. Financial News · KOSPI Weakens on Concern Over Samsung and SK hynix Buyback Endings; KOSDAQ Rises Oct. 6, 2026 · KOSPI 6,941.39 (−0.89%) · Foreign investors −KRW 1.7672tn · Samsung KRW 271,500 (−1.63%) · SK hynix −3.64% · Article interprets concern over the end of buyback programs as a contributor to the decline
24. Hanyang Economy · Oct. 8, 2026 article · Daishin Securities analyst Kyoung-min Lee: sell-on-the-news flows after earnings confirmation, holiday caution, and net selling by foreign investors, institutions and financial investment firms · Article figures: foreign investors −KRW 1.9902tn, institutions −KRW 1.6658tn, individuals +KRW 2.9021tn
25. M Economy News · Oct. 6, 2026 article · Regular rebalancing across 24 domestic equity ETFs worth KRW 23.2tn, of which semiconductor-related funds were KRW 18.1tn or 78.6% · TIGER Semiconductor TOP10 had Samsung at 27.1%, above its 25% cap · Samsung Securities analyst Kyun Jeon estimated roughly KRW 200bn of Samsung selling, KRW 90bn of SK hynix buying and KRW 80bn of SK Square buying / Financial News 2026.10.06 · Holiday calendar expected to concentrate the rebalancing trades on Oct. 8
26. Korea Economic Daily (Jin-gyu Kang) · 'I Thought Samsung Would Fall Further'—Why the Stock Held Up Oct. 7, 2026 · Foreign investors sold a net KRW 24bn of Samsung on Oct. 7 · The buyer that had been taking more than KRW 500bn a day before completion was gone · Meritz Securities estimated 3Q operating profit at KRW 106.1tn, slightly above what it described as market expectations in the KRW 104tn range
27. Special Economy · Oct. 8, 2026 article, published before the results in the morning · Kiwoom Securities analyst Ji-young Han argued that investors should watch the immediate share-price reaction, foreign-flow changes and revisions to 2027 earnings estimates more closely than whether preliminary results beat consensus · Cited earnings, the holiday and options expiry as flow variables
28. Daum/Yonhap-affiliated report · Korean Stock-Market Holiday Schedule Sep. 24, 2026 · Oct. 5, Monday: substitute holiday for National Foundation Day · Oct. 9, Friday: Hangul Day market holiday
29. U.S. Bureau of Labor Statistics · Schedule of Selected Releases, October 2026 · September Employment Situation Oct. 2 at 08:30 ET · September CPI Oct. 14 at 08:30 ET · 21:30 KST
30. Bank of Korea · Monetary Policy Board Meeting Schedule · Next policy decision meeting: Thursday, Oct. 22
31. Federal Reserve Board · October 2026 Calendar · FOMC Oct. 27–28 · Press conference Oct. 28 at 14:30 ET, statement at 14:00 ET · 03:00 and 03:30 KST on Oct. 29
2026.10.08 · Snowshagal INVESTMENT NOTE NO.09snowshagal.com