商业 · 韩国市场抛售

韩国股市暂停交易令抛售暂歇，但大幅损失仍在

韩国股市的抛售因20分钟交易暂停而中断，但损失并未因此消失。最近一次经核实的下午读数显示，KOSPI仍下跌8.13%；两个市场的收盘结果均未得到确认。

播出 文字 阅读需 8 分钟39 个来源HeyDay News Desk · About usWatch report
与 HeyDex 一起探索

通过 RSS 关注

原语言视频

文字实录 · 正在加载

译自英文原文。

新闻追踪第 3 天

1 篇报道25 家媒体39 个来源

  1. 第 1 次更新韩国股市暂停交易令抛售暂歇，但大幅损失仍在当前位置

对于眼看着韩国股票暴跌的投资者来说，此次抛售带来了两种损失：报价大幅下跌，以及一段时间内完全无法交易。2026年10月15日，两大主要股市均出现了20分钟的交易暂停。但这并未抹去损失。韩国标准时间下午1:17，基准指数KOSPI仍下跌8.13%。1 36

随着抛售加剧，韩国交易所的干预措施也逐步升级。上午短暂限制程序化卖单，随后在午间先后对KOSDAQ和KOSPI启动一级熔断机制，暂停受影响市场及相关衍生品的交易。两者的区别很重要：交易所先限制了一类订单，随后才暂停整个市场的交易。9 12

从限制订单到暂停整个市场交易

  1. 10月15日，开盘

    KOSPI开盘报5,673.01点，较前一交易日收盘已下跌5.59%。

    1
  2. 10月15日，韩国标准时间上午约9:08

    KOSPI程序化卖单暂停五分钟。随后，KOSDAQ于上午9:12暂停程序化卖单。

    212
  3. 10月15日，韩国标准时间中午约12:23

    KOSDAQ下跌8.01%后触发一级熔断机制。

    19
  4. 10月15日，韩国标准时间中午约12:38

    KOSPI在5,528.05点触发熔断机制，下跌480.76点，跌幅为8.00%。

    359
  5. 10月15日，韩国标准时间下午1:17

    KOSPI报5,515.76点，下跌8.13%——这是下午的一个盘中读数，并非收盘结果。

    36

从开盘钟声响起，跌势就已十分严重，并非到了午间触发阈值时才突然出现。走势也并非一路下行。韩联社报道称，上午11:20时KOSPI报5,604.6点，下跌6.73%；此前该指数曾出现另一个报道读数，随后部分回升，而后抛售再次加剧。这一上午的读数无法独立复核。1 3

另一个上午的快照尤其需要谨慎看待。《韩国中央日报》报道称，上午10:30时KOSPI报5,539.51点，跌幅为7.77%。按已确认的前一日收盘价6,008.81点计算，该水平意味着跌幅约为7.81%。所报指数点位与百分比无法相互吻合，两者都不应被误认为是当日最终跌幅。4 15

这些保护机制实际暂停了什么

KOSPI上午实施的限制措施称为卖方侧车机制，监测的是期货，而非现货指数。其触发条件是：交易量为前一交易日最高的KOSPI 200期货合约，跌幅至少达到5%，并持续一分钟。上午9:07:42启动五分钟暂停时，相关期货价格下跌5.27%。12

随后启动的KOSPI熔断机制则要求现货指数较前一日收盘价低至少8%，并持续一分钟。该干预措施适用于市场中的所有股票。现有材料未能确认KOSDAQ机制的完整官方数值触发条款，因此不应简单将KOSPI的详细规则套用到对应机制上。6 11

恢复交易也是一个过程，并非瞬间切换。KOSDAQ恢复交易通知规定，先进行十分钟的订单收集，随后进行单一价格竞价，再恢复连续交易。订单收集期间可以成交大宗交易。韩联社中午12:44的简讯确认，暂停交易措施已解除。KOSPI实际恢复交易的确切时间尚未确定，不过下午较晚时已有相关指数读数。10 14 36

历史对比的范围比“自7月以来首次两个市场同时暂停交易”更窄。《MoneyToday》10月15日的报道证实，上午同时启动卖方侧车机制是自7月29日以来首次，距此前已有51个交易日。这并不能证明午间范围更广的熔断机制也创下了同样纪录。37

芯片股损失叠加融资疑虑

主要芯片制造商遭受重挫。《韩国中央日报》上午10:30的快照显示，三星电子报211,000韩元，下跌9.05%；SK海力士报1,267,000韩元，下跌10.71%。上午的报道还显示，外国投资者净卖出1.92万亿韩元，约合14亿美元，净卖出势头延续至连续第八个交易时段。这些数字未经独立复核，仍属于盘中报道，而非最终回报或全天资金流数据。4 2

抛售在周四之前就已开始。《华尔街日报》报道称，KOSPI于10月13日收盘下跌约5.5%。10月14日，该指数又下跌0.44%，收于6,008.81点；KOSDAQ下跌1.57%。外国投资者在这一完整交易日净卖出1.18万亿韩元。将这一全天资金流与周四上午的数据进行比较时，必须注意两者的统计截止时间不同。31 15 16

科技股面临的担忧不容小觑，但这些因素各自对跌幅的贡献尚未得到量化。此前的报道提到融资成本上升、估值偏高，以及昂贵的人工智能基础设施能否带来持续利润的不确定性。一个具体案例是英伟达支持的数据中心运营商Firmus Grid Ltd.首次公开募股宣告失败；彭博社经《金融邮报》报道了此事。这次上市失败并未量化其对韩国股价的影响。27 29

中国存储芯片领域的竞争又增添了一项担忧。《卫报》报道，CXMT声称其新型DRAM架构取得进展，旨在提高存储密度。这是技术方面的说法，并不能确认额外商业供应已经出现。量产准备情况、产能、出货时间和客户认证均未得到证实。未来可能出现存储芯片过剩，或现有生产商丢失市场份额，但这仍是一种风险，而非新芯片已交付客户的事实。25 31

报道还提到了美国通胀和利率方面的担忧。最初报道中附带的美国通胀和股指具体数据无法独立复核。更重要的是，列出可能的压力因素，并不等于计算其因果关系：报道没有确定周四韩国市场的跌幅有多少来自利率、人工智能融资或中国竞争。1 2

网上流传的两项耸动说法仍未得到证实：市值蒸发了2030亿美元，以及KOSPI较6月峰值低了40%。前一说法需要相匹配的市值数据和货币换算依据；后一说法则需要有据可考的峰值及比较读数。10月13日一则报道称韩国市场市值已跌破5,000万亿韩元，但该报道无法提供10月15日这两项计算所需的数据。1 20

交易时段尚未结束，这限制了目前能够作出的结论。现有报道没有给出10月15日任一指数的收盘水平，因此也无法确定当日最终跌幅或收盘价连续下跌的天数。但已有证据所显示的情况已经相当重要：两个市场均触发了一级交易暂停阈值，且最近一次经核实的KOSPI读数仍显示跌幅超过8%。9 36

来源

SBS，10月15日市场暂停交易 1；《韩国时报》/韩联社，上午交易限制措施 2；韩联社，盘中跌幅 3；《韩国中央日报》，10月15日抛售 4；CNBC，CoreWeave融资担忧 5。

韩国交易所，KOSPI熔断规则 6；KOSDAQ侧车机制规则 7；KIND启动标准 8；《数字时报》，双重熔断机制 9；《首尔经济日报》/KOSCOM，KOSDAQ恢复交易通知 10；KRX，KOSDAQ交易规定 11；SIGNAL，KOSPI侧车机制通知 12。

《MoneyToday》，上午侧车机制的英文报道 13；韩联社，KOSDAQ暂停交易措施解除 14；《Digital Today》，10月14日完整交易时段 15；《亚洲商业日报》，10月14日收盘 16；KRX数据市场 17。

《朝鲜商业日报》，10月13日外国投资者抛售 18；Businesskorea，开盘市场报道 19；《首尔经济日报》，10月13日市场市值 20；《先驱商业报》，此前芯片股跌幅 21；《朝鲜商业日报》，10月12日人工智能相关担忧 22；Sisnoo，10月14日市场报道 23；Korea Stock Insight，市场概况 24。

《卫报》，CXMT及人工智能投资担忧 25；《福布斯》，芯片股下跌 26；《首尔经济日报》，人工智能融资担忧 27；彭博社经雅虎财经报道，人工智能交易担忧 28；彭博社经《金融邮报》报道，Firmus Grid及融资 29；Baystreet，人工智能市场评论 30。

《华尔街日报》，CXMT公告及亚洲股市 31；美联社，10月13日半导体抛售 32；《首尔经济日报》，10月14日市值报道 33；《华尔街日报》，备用文章网址 34。

SIGNAL，KOSPI熔断通知 35；《韩国先驱报》，下午1:17的KOSPI读数 36；《MoneyToday》，上午侧车机制的韩文报道 37；SBS Biz，上午程序化卖单限制 38；韩联社，盘中波动报道 39。

播出原貌 26 行
  1. South Korea's two main stock markets stopped trading after both indexes plunged eight percent.
  2. Each interruption lasted twenty minutes, cutting off trading as major chipmakers suffered steep losses. The test is whether those pauses stopped the slide, or simply interrupted it.
  3. The KOSDAQ triggered its first-level circuit breaker at about twelve twenty-three Thursday afternoon, after falling eight point zero one percent. The KOSPI followed about fifteen minutes later. Both halts also stopped trading in associated derivatives.
  4. The KOSPI hit its trigger at five thousand five hundred twenty-eight point zero five. That put it four hundred eighty point seventy-six points below the previous close, a fall of eight percent. But the exchange had already intervened twice that morning, using narrower restrictions.
  5. At about nine oh-eight, the exchange suspended KOSPI program sell orders for five minutes. The KOSDAQ followed at nine twelve. Those morning restrictions targeted automated selling. By twelve twenty-three and twelve thirty-eight, the interventions covered the entire affected stock markets.
  6. The KOSPI's morning restriction, called a sell-side sidecar, followed a five point twenty-seven percent fall in the relevant futures contract. Futures are contracts tied to an index's expected value. The sidecar rule watches a different instrument from the later circuit breaker.
  7. The sidecar requires a futures decline of at least five percent for one minute. The circuit breaker requires the KOSPI itself to stay at least eight percent below its previous close for one minute. Its twenty-minute halt reaches all shares in that market.
  8. The futures test uses the KOSPI 200 contract with the preceding day's highest trading volume. Those detailed rules describe the KOSPI mechanism. The historical comparison also belongs to the morning sidecars: both markets activated them together for the first time since July twenty-ninth, fifty-one trading days earlier.
  9. The KOSPI had already opened five point fifty-nine percent lower. It began at five thousand six hundred seventy-three point zero one, down three hundred thirty-five point eighty points. Selling was severe from the opening bell, well before the midday shutdowns.
  10. At ten thirty, the KOSPI was reportedly at five thousand five hundred thirty-nine point five one. By eleven twenty, it had recovered to five thousand six hundred four point six, down six point seventy-three percent. Then it fell again to the circuit-breaker threshold.
  11. The ten-thirty snapshot came with a quoted decline of seven point seventy-seven percent. That percentage doesn't match the index level. Against the previous close, the level implies about seven point eighty-one percent. The later plunge overtook that morning low.
  12. The KOSDAQ halt was lifted at about twelve forty-three. The recorded seconds differ slightly across the halt notices and activation times. Its reopening also had several stages, so lifting the interruption didn't immediately restore normal continuous trading.
  13. The KOSDAQ reopening notice allowed ten minutes to collect orders, followed by an auction setting a single price. Continuous trading came afterward. Block trades could execute from the start of order collection. The exact actual KOSPI reopening time hasn't been specified.
  14. Foreign investors had reportedly sold a net one point ninety-two trillion won by ten thirty, about one point four billion dollars. That extended their net selling into an eighth consecutive session. The figure measures money moving out through net share sales during the morning.
  15. At that same snapshot, Samsung Electronics traded at two hundred eleven thousand won, down nine point zero five percent. SK hynix stood at one million two hundred sixty-seven thousand won, down ten point seventy-one percent. Both were still trading through an unfinished session.
  16. Shareholders who kept their holdings faced sharply lower quoted values. People needing to buy or sell also faced interrupted market access. The foreign-selling total has a different meaning from the amount of stock-market value erased, and its morning cutoff matters when comparing it with earlier sessions.
  17. The KOSPI closed about five point five percent lower on October thirteenth. On October fourteenth, it lost another zero point forty-four percent, finishing at six thousand eight point eight one. The KOSDAQ fell one point fifty-seven percent that day.
  18. Foreign investors sold a net one point eighteen trillion won during the whole October fourteenth session. Samsung finished flat, while SK hynix fell. Thursday's larger morning outflow and Wednesday's completed daily total cover different periods.
  19. Firmus Grid's collapsed initial public offering gave the AI financing worries a concrete example. The Nvidia-backed data-centre operator had sought a stock-market listing. Its failure accompanied growing doubts about the cost of building AI infrastructure and whether the spending will produce lasting profits.
  20. Higher financing costs and stretched valuations added to those worries. Investors were becoming more selective about AI infrastructure deals. But the failed listing doesn't tell us how much of this Korean selloff came from financing fears, or whether particular AI projects will ultimately earn profits.
  21. CXMT, a Chinese memory-chip maker, also claimed progress on a new DRAM architecture intended to increase memory density. DRAM is a common form of computer memory. The competitive concern is that future Chinese advances could pressure established producers' market share and prices.
  22. CXMT's technology claim concerns denser memory. Commercial availability depends on production readiness and customer acceptance, and a shipment timetable hasn't been given here. A possible future memory glut remained a competitive risk, rather than a measured increase in chips already reaching customers.
  23. The international backdrop added pressure too. The cited figures put September U.S. inflation at three point seven percent year over year, above forecasts. They also put the Nasdaq's October fourteenth loss at one point eighty-eight percent, and the S and P five hundred's at one point eleven percent.
  24. AI spending, Chinese competition and interest-rate worries all accompanied the selling. Their separate contributions haven't been measured. Two dramatic online claims also remain unconfirmed: that two hundred three billion dollars was wiped out, and that the KOSPI stood forty percent below its June peak.
  25. Korean market value had fallen below five thousand trillion won on October thirteenth. That earlier valuation level can't supply a dollar-loss calculation for Thursday. Such a calculation needs matching market values, comparison times and an exchange rate.
  26. By one seventeen, the KOSPI stood at five thousand five hundred fifteen point seventy-six, down eight point thirteen percent. The pauses had interrupted trading without reversing the steep decline. That was an afternoon reading; the final closing losses for both indexes weren't yet known.

来源

39 个来源24 家媒体1 篇通稿

来源地图

追溯本篇报道的依据。图谱展示其关联；下方链接可打开本文引用的来源。

来源图谱将在您阅读时加载。

探索完整的新闻编辑室图谱 ↗

译文更新于

修订记录

  1. r1首次发布。