Korean Nuclear Basket: Anatomy of a Four-Stock Limit-Up

Analysis date: 2026-06-12. All prices, multiples, and investor-flow figures below were pulled from live data on this date (Yahoo Finance quotes/financials and Naver Finance investor tallies). Valuation multiples are computed from the most recently reported fiscal year (FY2025) actuals against the 2026-06-12 closing price — they are not training-data estimates.

On 2026-06-12 the Korean nuclear/power-equipment complex had a violent up-day, with four names closing locked at the +30% daily price ceiling. This piece does three things: (1) explains exactly what the price ceiling and volatility-interruption mechanics did — i.e. why a stock can print "+30% and stop," (2) separates the real catalysts from sector beta, and (3) maps the basket on two independent signals, foreign accumulation and real-earnings valuation, which turn out to intersect in only two names.

Summary

Four nuclear names locked limit-up (+30%): BHI (083650), KEPCO E&C (052690), WooriTG (032820), and KEPCO KPS (051600). The move was sector-wide (a US-nuclear tailwind overnight plus two concrete Korean order catalysts), not a single-stock story. But under the surface the basket is bifurcating: foreign capital is rotating within it, favoring specific order/catalyst names over the expensive large-cap leader, Doosan Enerbility (034020), which managed only +5% and remains under heavy foreign distribution. Cross-referencing flow against verified trailing valuations, only BHI sits cleanly at the intersection of real growing earnings + foreign accumulation + a hard catalyst. A second name, BMT (086670), triggered a fresh foreign-buying signal today from a near-zero base and is worth watching, unconfirmed.

How the +30% ceiling and the "circuit breakers" actually work

A common question: if a buyer bids the price up to the ceiling, why doesn't it just keep going — and why do all the limit-up names print exactly ~+30%? The answer is three distinct Korea Exchange (KRX) mechanisms:

1. Daily price limit band (가격제한폭): ±30%. Every KOSPI/KOSDAQ stock has a hard daily cap and floor of ±30% versus the prior session's close. A stock physically cannot trade above +30% (상한가, "limit-up") or below −30% (하한가, "limit-down") in a single session. This is why today's four ceiling names printed almost identical numbers — BHI +30.00%, KEPCO E&C +29.98%, WooriTG +29.96%, KEPCO KPS +29.61% — and why each closed at its high: once the ceiling is reached, no order can be placed higher, sellers are exhausted, and a queue of unfilled buy orders "locks" the stock at the ceiling for the rest of the day.

2. Volatility Interruption (VI, 변동성 완화장치): the per-stock circuit breaker. On the way up to the ceiling, continuous trading is repeatedly interrupted by VI — a brief (≈2-minute) single-price call-auction cooling-off triggered automatically when a stock moves too far, too fast. Static VI fires at roughly ±10% from a reference price; dynamic VI fires on a sudden instantaneous jump versus the immediately preceding trade. Each of today's limit-up names would have tripped VI several times during the session — short halts that interrupt the continuous order-book sweep, cool the move, then resume.

3. Market-wide Circuit Breakers (서킷브레이커) and Sidecar (사이드카). These are index-level, not single-stock, and are designed for crashes rather than melt-ups. A market-wide CB halts the entire market for 20 minutes if KOSPI/KOSDAQ falls ≥8% (level 1) or ≥15% (level 2), and for the day at ≥20% (level 3). A Sidecar halts program trading for 5 minutes on a sharp futures move. Neither was the operative mechanism today; the relevant tools for a single-stock melt-up are the ±30% band and VI.

The order-book mechanics, made explicit. Aggressive buyers do sweep the resting sell orders — every sell limit between the prior close and +30% gets filled on the way up (that swept liquidity is why volume on these names ran several multiples of average). The continuous sweep is punctuated by VI cooling auctions. The reason it stops at +30% is not that sellers "win" — it is the price band: orders above +30% are simply not accepted. Sellers run out below the ceiling, the buy queue stacks up at it, and the stock locks.

The catalysts vs. sector beta

Today's rally had a macro tailwind (US nuclear equities rallied overnight, lifting the whole Korean theme) layered on top of two concrete, name-specific order catalysts:

• BHI — Mitsubishi Power HRSG contract (~₩250B). BHI disclosed a ~₩250 billion contract to supply five 500MW heat-recovery steam generators (HRSG) for the Tungxiao power plant in Taiwan (deliverable by May 2031) — its first HRSG supply to Taiwan, breaking into the main-equipment market. The contract was signed last October but had been under a disclosure deferral (공시 유보) at the counterparty's request; the deferral expired today, so the market repriced it in one session. ₩250B is roughly a third of BHI's FY2025 revenue (₩774B), a large single order.
• KEPCO E&C — Czech Dukovany design scope (~₩1.25T, attribution not fully verified). The probable trigger is comprehensive-design scope tied to the Czech Dukovany 5·6 reactor program, on top of recent sell-side BUY initiations. The limit-up is verified data; the exact same-day disclosure should be confirmed on DART (electronic disclosure) before being treated as fact.

The other two ceiling names (WooriTG, KEPCO KPS) and the mid-pack rode sector beta rather than a fresh order. Tellingly, the four ceiling names closed at their highs while the rest of the complex (KEPCO, Hyosung Heavy, Woojin, Doosan Fuel Cell) faded from intraday highs of +8–10% to closes of +3–5% — money concentrated into the catalyst names and distributed the generic beta.

Basket scoreboard (2026-06-12 close)

Investor flow: foreign accumulation is the leading signal

The 20-session foreign flow (Naver Finance tallies) front-ran which names exploded. Foreign capital was building the three biggest catalyst names through the May–June correction, while distributing the leader:

Two distinct facts: (1) foreign accumulation during the decline was predictive of today's winners, but (2) on the breakout day itself, foreign largely took profit into the spike on three of four ceiling names — only BHI had foreign as the actual marginal breakout buyer. Doosan's lag is fully explained by flow: it is the only basket name foreign distributed heavily, and they sold it the whole way down and today. It has no foreign-accumulation base under it, which is why it cannot lead even on a four-limit-up day.

Valuation: trailing P/E from FY2025 actuals (not stale estimates)

Computed as 2026-06-12 close ÷ FY2025 diluted EPS, with forward P/E from the quote where available. The headline multiples are misleading for several names because of one-off items — flagged below.

Three findings that reframe the move:

1. Two of the four limit-ups have poor or fake trailing earnings. WooriTG — the biggest foreign-accumulation name and a +30% lock — ran an operating loss in FY2025; its tiny headline net profit is non-operating/minority noise, so its P/E is meaningless. It is a pure thematic story stock. KEPCO E&C's 67x is flattered by ~₩68B of unusual items; its core operating income halved YoY, so on normalized earnings it trades near ~179x, not 67x.
2. BHI is the cleanest limit-up by a wide margin — the only +30% name that also shows real, fast-growing, operating-driven earnings (net income tripled to ₩65.2B on ₩75.5B operating income) at a moderate ~38x trailing / ~43x normalized.
3. The cheapest real-earnings names did NOT lock limit-up. Iljin Power (11x), Bosung Power (20x), and BMT (~14x normalized) — all with clean, growing operating earnings — moved only +4% to +9%. The tape rewarded catalyst + story over cheap + profitable. That is a momentum-driven session, not a value-driven one.

Does the foreign signal show up in the cheap names? Mostly no.

Checking the flow under the three cheapest real-earnings names tests whether value and the foreign signal coincide. They largely do not:

• Iljin Power is the important negative: the best fundamental name has foreign ownership collapsing (8.4% → 5.9%) with selling into today's bounce. Cheap and getting cheaper with foreign exiting is the classic value-trap shape — not a buy on flow despite the valuation.
• Bosung Power is a high-turnover retail name with no coherent smart-money trend. Neutral.
• BMT is the one new setup the day produced: cheapest normalized earnings in the basket, real fast-growing earnings (NI ₩4.1B→19.0B), and foreign initiating today (+65K, Goldman and JP Morgan both top buyers) from a near-empty ~3% base with room to build. It resembles a "day 1" version of BHI's earlier setup — but one session off a flat base is a flag, not a trend, and needs 2–3 sessions of confirmation.

The value-vs-momentum map

The two independent signals — foreign accumulation and cheap real earnings — almost never coincide in this basket. They intersect in exactly two names: BHI (confirmed weeks ago, paid off today) and BMT (triggered today, unconfirmed). Everything else is either a story stock the earnings don't support, or a cheap name foreign money won't touch.