Samsung Electronics opens Korean earnings season every quarter with a pre-market flash known as jamjeong silsaek (provisional results), and it contains exactly two numbers: consolidated revenue and consolidated operating profit. There is no net income, no segment breakdown, no guidance, and no conference call — and the full filing does not arrive for weeks. Because Samsung is historically the largest single weight on the KOSPI and the first major Korean company to report each quarter, the entire market trades the gap between that one operating-profit figure and consensus within minutes of the 9:00 a.m. open.
October 2026 showed why the mechanics matter more than the headline. Samsung’s third-quarter flash window — the first days of the month after quarter-end — coincided with heavy foreign selling: the KOSPI closed at 6,625.93 on 2026-10-08, and foreign investors net sold roughly 2.07 trillion won of KOSPI shares that session. A flash window that overlaps with heavy foreign selling, whatever the headline number eventually shows, is exactly the kind of tape this guide is built to explain.
What the Flash Contains — and What It Deliberately Omits
The jamjeong silsaek is a voluntary regulatory disclosure (gongsi, the Korean term for an exchange filing) released through the Korea Exchange disclosure system before the market opens. It is deliberately sparse.
| Item | Preliminary flash | Final results (weeks later) |
|---|---|---|
| Consolidated revenue | Yes | Yes |
| Consolidated operating profit | Yes | Yes |
| Net income | No | Yes |
| Segment detail (semiconductors, mobile, display) | No | Yes |
| Conference call / management commentary | No | Yes |
The practical consequence: the first hour of trading is spent reverse-engineering the mix. Analysts and desks publish rapid flash notes estimating how much of the group operating profit came from the semiconductor division versus mobile, display, and consumer electronics. The stock therefore often reprices twice — once on the headline surprise at the open, and again mid-morning as credible mix estimates circulate. In unusual quarters Samsung has occasionally attached a brief explanatory note to the flash, which is itself a signal: the company only explains when it judges the headline would otherwise be misread.
Why the Flash Leads the Final Filing by Weeks
The flash typically lands within roughly the first five to ten calendar days after quarter-end, with the exact date usually signaled a day or two in advance. The final results — with net income, the segment profit-and-loss statement, and the earnings call — follow roughly three to four weeks later, filed through DART, Korea’s electronic disclosure system. The gap exists because the flash is compiled from a near-complete internal closing of the books, while the final filing requires full consolidation and review.
Two actionable points follow. First, treat the flash as the market-moving event and the final as the detail event: price discovery on the quarter happens on flash day, and the final filing mostly moves the stock only when segment detail or net income diverges from what flash-day mix estimates assumed. Second, headline revisions between flash and final are usually modest, so a trader who waits for the “real” number is waiting for a number the market has already traded.
How the Market Reads the Operating-Profit Number
The reference point is consensus — the average of brokerage estimates compiled by local data vendors. But the reaction function has three layers, and beginners usually only see the first.
- The surprise itself. Small beats or misses within a few percent of consensus are often absorbed quietly. The large single-day moves tend to come from double-digit-percent surprises, or from quarters where informal “whisper” expectations had drifted far from the published consensus.
- The composition. Memory semiconductors are a deeply cyclical business, and group operating profit can be flattered or depressed by items the flash does not disclose: inventory valuation reversals when memory prices recover, one-off provisions, or performance-bonus accruals. A beat the market suspects was driven by an inventory write-back reversal is treated very differently from a beat driven by shipments and pricing — even though the flash shows the identical number.
- Positioning. When the stock has rallied hard into the print, the flash resolves uncertainty, and resolved uncertainty is itself a trigger for position unwinds. This is the structural reason “record profit, stock down” is a recurring Korean-market pattern rather than a paradox: the flash is often the scheduled exit for momentum money, especially when it coincides with options expiry or index rebalancing.
The Same-Morning Read-Through: SK Hynix, the Supply Chain, and the KOSPI
SK Hynix does not publish a preliminary flash; it reports full results with a call several weeks after Samsung. Samsung’s flash is therefore traded as the first hard data point on memory pricing and demand for the whole quarter, and SK Hynix frequently moves as much as, or more than, Samsung at the open. The read-through cascades down the supply chain — semiconductor equipment, materials, and substrate names — and into the index itself: Samsung alone has historically accounted for a large share of KOSPI capitalization, and together with SK Hynix the pair can represent roughly a quarter or more of the index in strong memory cycles. A big Samsung surprise is, mechanically, a KOSPI event.
The practical routine for the first thirty minutes of a flash-day session is to watch three tapes at once: Samsung itself, SK Hynix as the leveraged read-through, and the foreign-investor net-flow tape, because foreign money (oegugin flows in local market shorthand) is the marginal buyer or seller of Korean large-cap semiconductors.
Worked Example: Reading Foreign Flows Around a Flash Window
Here is an actual sequence of daily foreign net flows on the KOSPI through early October 2026 — the window in which Samsung’s third-quarter flash would ordinarily fall — in units of eok won (one eok won = KRW 100 million, so 10,000 eok won = 1 trillion won):
| Session (2026) | Foreign net flow (eok won) |
|---|---|
| Sep 22 | +494 |
| Sep 23 | -4,942 |
| Sep 28 | -32,682 |
| Sep 29 | -29,922 |
| Sep 30 | -20,646 |
| Oct 1 | -5,558 |
| Oct 2 | -1,080 |
| Oct 6 | -17,575 |
| Oct 7 | -26,188 |
| Oct 8 | -20,704 |
Read it step by step. First, convert the most recent figure: -20,704 eok won on October 8 is roughly 2.07 trillion won of net selling — large, but not an outlier in this table. Second, look left: foreigners had been net sellers for eight consecutive completed sessions before that, including two sessions in late September that were each heavier than October 8 itself. Third, draw the correct conclusion: selling on any single day in this window was the continuation of an established distribution trend, not a verdict delivered by one catalyst alone. Attributing the entire -20,704 to “earnings disappointment” would misread a session that sat squarely inside an established selling trend. Fourth, check the absorption capacity on the other side: investor deposits (tujaja yetakgeum, idle cash sitting in brokerage accounts) stood at roughly 100.4 trillion won and margin lending (sinyong yungja, broker loans for stock purchases) at roughly 33.5 trillion won as of 2026-10-07 — context for whether domestic retail can keep absorbing foreign supply.
Common Mistakes
- Trading the flash as if it were final. Net income and segment detail arrive only with the full filing; a flash-day thesis built on an assumed division mix can be wrong even when the headline was right.
- Attributing all flash-day flow to earnings. Flash dates can coincide with options expiry, ETF rebalancing, and pre-existing flow trends, as the worked example shows.
- Comparing the flash to year-ago numbers without checking one-offs. Memory-cycle inventory effects and provisions distort simple year-on-year operating-profit comparisons.
- Assuming no call means no information. The speed and direction of the sell-side mix estimates — and whether Samsung attached an explanatory note — are information in themselves.
FAQ
When exactly is the Samsung flash released?
Historically within roughly the first five to ten calendar days after quarter-end, before the 9:00 a.m. Seoul open, with the date typically signaled through the disclosure calendar a day or two ahead. Because it precedes the opening bell, the first traded price already embeds the surprise.
Do other Korean companies publish preliminary results?
Most do not. A handful of other large caps, such as LG Energy Solution, also release provisional figures early in the month, which is why the first week after quarter-end functions as Korea’s unofficial earnings-season opening. SK Hynix skips the flash format and reports full results with a call later in the month.
How different can the final numbers be from the flash?
Headline revenue and operating profit revisions are typically modest, since the flash is compiled from a near-complete internal closing. The genuinely new information in the final filing is net income, the segment breakdown, and management commentary — which is where flash-day mix assumptions get confirmed or corrected.
Where can I track foreign flows and market data on a flash day?
Daily investor-type net flows are published on the KRX data portal and mirrored in accessible form on Naver Finance; investor deposits and margin-lending balances are published in KOFIA’s statistics service; macro series are available from the Bank of Korea’s ECOS database.
Sources
- KRX data portal (data.krx.co.kr) — daily investor-type trading flows and index data
- Naver Finance (finance.naver.com) — investor trend pages and company disclosure summaries
- KOFIA statistics (freesis.kofia.or.kr) — investor deposits and margin-lending balances
- Bank of Korea ECOS (ecos.bok.or.kr) — macroeconomic and financial statistics
Disclaimer: This is an information and analysis publication, not investment advice. See our Methodology for data sources, standards, and our corrections policy.