中文
One Print, Three Prices: How Broadcom's Miss Became a Circuit-Breaker in Seoul
Jun 5, 2026 阅读中文版

One Print, Three Prices: How Broadcom's Miss Became a Circuit-Breaker in Seoul

Yesterday's pulse covered the print itself: U.S. markets priced Broadcom as a single stock down about 13%, with the sector still rising. This one follows the news as it traveled west overnight. The same report became a KOSPI down 4.33% with an intraday circuit-breaker in Seoul, SK Hynix down 9.92%; in Tokyo it was a roughly 2.6% two-day pullback from a record high. The difference isn't in the news — it's in how high each market had been bid up and how its ownership was structured. The more vertical the run and the more foreign money was already heading for the exit, the lower the tolerance for anything short of perfect. This piece makes no call on direction; it just lays out the three prices.

Tags
semiconductorscross-marketcontagionkoreajapanrisk-off
Tickers
AVGOSK HynixSamsung^KS11^N225Tokyo Electron

One print, three time zones

Yesterday’s pulse took apart Broadcom’s reaction: every hard number looked great — revenue of $22.2bn, a double EPS beat, AI semiconductor revenue around $10.8bn, up roughly 143% year over year — except Q3 AI chip guidance of about $16bn fell short of the roughly $17.2bn the market wanted, and that already-lofty full-year $100bn target wasn’t raised (sources: CNBC, Yahoo Finance, June 3–4, 2026). The U.S. answer on June 4: AVGO closed -13.0% at $416.83, shedding about $285bn in market value (sources: Yahoo Finance daily data and media, June 4, 2026).

But U.S. markets cut only that one stock. The same day, NVDA was up +2.5%, SMH about -1%, SPY +0.45%; over the trailing five sessions (May 28 → June 4), SOXX was +7.7%, Micron +10.9%, AMD +7.0%, well ahead of SPY’s +0.4% (source: Yahoo Finance daily data, May 28 – June 4, 2026). The U.S. treated Broadcom as a single-stock event, and the sector kept climbing.

Then the news traveled west overnight. By the June 5 Asian session, the same event got re-priced — and every market printed a different price.

Seoul: amplified into a circuit-breaker

Korea priced it as a systemic event. At the June 5 close, the KOSPI was -4.33% at 8,265.62, having tripped its circuit-breaker intraday at one point near -6.9% (sources: KRX market data, plus Nikkei Asia, CNBC, June 5, 2026). The two memory giants took the worst of it: SK Hynix -9.92% to ₩2,070,000, Samsung Electronics -6.40% to ₩329,000 (source: KRX market data, June 5, 2026).

On the day, it looks like a crash. Stretch it out and you get a different picture: Samsung is still +9.9% over the trailing five sessions — it had a single day near +10% on June 1; Hynix is -9.6% over five days. The hardest down-day landed right after a steep climb.

The ownership side matters more. Foreigners had already been net sellers for five straight sessions: about 24.86m shares of Samsung, about 1.90m of Hynix, with foreign ownership ratios slipping on both (source: KRX/Naver Finance investor-trading data, foreign data through June 4, 2026). Broadcom’s guidance wasn’t the cause of Seoul’s drop so much as the exit that already-departing money had been waiting for. The print just gave the outflow something to call itself.

Tokyo: a pullback from a record high

Tokyo printed yet another price — not a circuit-breaker but a high-level pullback. The Nikkei 225 closed June 5 at 66,588, -1.31%, which on top of June 4’s -1.36% leaves it about 2.6% off the June 3 record high of 68,402 over two days (sources: Trading Economics, Yahoo Finance daily data, June 3–5, 2026). The index was mild; the chip-equipment names inside it were not: Tokyo Electron fell close to 6% on June 5 (snapshots range -5.6% to -6.9%), Advantest about -4.4% to -5.0%, Renesas -5.1% (sources: Trading Economics, TradersUnion, June 5, 2026). SoftBank, an AI-exposure asset rather than a chipmaker, had already plunged -11.3% on June 4 and slipped another -1% on June 5 (sources: Yahoo Finance daily data, gurufocus, June 4–5, 2026).

The pattern is plain to see: Tokyo Electron was +21.7% over the prior five sessions (May 28 → June 4; source: Yahoo Finance daily data). The hardest fall belonged to what had just risen fastest. “The higher the bid, the louder the drop” reads even more directly in Tokyo than in Seoul.

As for Europe, through June 4, ASML in Amsterdam was still +8.8% over five days, with no sector-wide selling (source: Yahoo Finance daily data, through June 4, 2026). Europe’s June 5 tape wasn’t available for this piece, so we leave it.

The other side of the trade

The bears: Broadcom’s flat guidance pricked a “priced-for-perfection” setup and is an early warning that AI capex is topping — on this read, Asia’s violent reaction isn’t an overshoot but a market pricing the risk the U.S. didn’t dare to (sources: Motley Fool, 24/7 Wall St, June 4–5, 2026). The weak spot: extrapolating a sector top from one guidance figure is single-point storytelling, and nobody can cleanly separate how much of Korea’s drop was pre-existing foreign exit versus fundamentals.

The bulls: this is an expectations reset, not a fundamental break. After the print, several houses kept Buy ratings and raised targets — KeyBanc to $575, Jefferies $550, Deutsche Bank $515, Morgan Stanley $502 (sources: Seeking Alpha, Yahoo Finance, June 4–5, 2026, each a house view, not this publication’s). The weak spot is just as clear: it assumes good fundamentals pull price back quickly, but if Asia’s money wants to cut exposure rather than re-rate, the timetable isn’t set by fundamentals.

This piece settles neither. The next signpost is already on its way — Micron reports soon (source: 24/7 Wall St, June 5, 2026), another memory-heavy, deeply Asia-exposed name.

Closing

No forecast — just the structure as it stands. One print, three time zones, three prices. The U.S. cut a single stock and stopped there; Seoul blew it up into a circuit-breaker; Tokyo sat in between, easing off a record high. What actually got priced was never Broadcom — it was how vertical each market’s run had been and how fast the foreigners were already leaving. Broadcom merely handed them the exit.

Yesterday asked how long the flag-bearer could carry it. Today shows the same flag getting re-weighed by local crowding as it crosses the Pacific.

What to watch

The following is an observation frame, not a trading signal.

  • Whether Korea’s foreign flow reverses: June 5’s drop sits on top of five straight days of net selling. Watch whether foreigners keep leaving or buy back — that says more than the index level about whether this is structural de-risking or an emotional overshoot.
  • “Height of the prior run” as a proxy for transmission strength: Tokyo Electron fell near 6% after a +21.7% five-day run. Watch whether, normalized, the hardest falls are broadly the fastest prior risers — a falsifiable observation.
  • Micron as the cross-market mirror: another memory-heavy, deeply Asia-exposed name. Watch whether it repeats the “U.S. single stock, Asia whole sector” split.
  • Europe and the U.S. June 5 catch-up: this piece didn’t capture Europe’s or the U.S.’s June 5 close. Watch whether they track Seoul or track June 4’s U.S. tape.

Sources: Broadcom Q2 FY2026 results (fiscal quarter ended May 3, 2026, reported after the close June 3, 2026), as reported by CNBC and Yahoo Finance (June 3–4, 2026); U.S. equity prices from Yahoo Finance daily data (May 28 – June 4, 2026); Korea data from KRX and Naver Finance plus Nikkei Asia and CNBC (June 5, 2026), foreign-flow data through June 4, 2026; Japan data from Trading Economics, TradersUnion, Yahoo Finance and gurufocus (May 28 – June 5, 2026); Europe from Yahoo Finance daily data (through June 4, 2026); broker price targets are house views, not this publication’s. All moves are labeled with base and end dates; Asia figures are June 5 closes, while Europe’s and the U.S.’s June 5 closes were not available at the time of writing. This piece is a personal observation and does not constitute investment advice; no price targets are set.

This content represents independent research and personal opinion for informational purposes only. Nothing herein constitutes investment advice or a recommendation to buy or sell any security. Past performance is not indicative of future results.