On 30 June 2026, EcoPro BM, a KOSDAQ battery-materials leader, announced a ₩1.2tn rights offering, and its shares fell about 19% after hours and 8.8% the next session. In the same month Alphabet raised roughly nine times as much equity — $84.75bn — and its stock fell about 5.5%. Dilution is a negative signal in any market. The question for anyone pricing the Korea discount is why the Korean fall is so much larger. The answer lies less in the deal than in the design: a shareholder-allocation method paired with a pre-fixed issue price, meeting learned distrust of intra-group capital recycling. Korea’s 2025 duty-of-loyalty reform speaks to this, but the channels to enforce it remain narrow.
The reaction gap is about size, not direction. US seasoned offerings also draw negative announcement reactions — about −3% over the announcement window. Alphabet itself fell around 5.5% on dilution concern, even with a $10bn Berkshire anchor and a structure split across mandatory convertible preferred stock, an at-the-market program, and underwritten shares. What distinguishes Korea is depth: in Korean data, rights issues carry a mean discount near −27.5%, against −17.2% for public offers, and draw more negative reactions. EcoPro BM’s roughly 19% after-hours move sits at the far end of that pattern. The caveat is the point, not against it — markets react to dilution everywhere; the design question is why they react harder here.
Korea’s default equity raise asks existing retail holders for fresh cash. Under the Capital Markets Act, a rights offering allocates new shares to shareholders in proportion to their holdings, with unsubscribed shares passed to a public offer; the Commercial Act frames the pre-emptive right the same way. The price is then set in two stages under FSC disclosure rules — an expected price at the board resolution (₩121,200, about a 20% discount) and a final price shortly before subscription on 12 October. If the disclosed expected price anchors trading, the mechanism is self-reinforcing: as the stock falls, the final price falls with it, requiring more shares for the same proceeds. The pre-fixing rule was built to protect minorities from value-diluting underpricing; a rule meant to protect can deepen the very fall it guards against, if it anchors the market downward. US issuers largely avoid this. Rights offerings are rare in the US — the “rights-offer paradox,” driven by the adverse-selection cost of shareholder non-participation — so issuers use shelf, ATM, and marketed sales priced into the market rather than fixed in advance.
The use of proceeds is where trust breaks. About 76% of EcoPro BM’s raise — ₩915bn — funds a stake in the group’s Indonesian nickel project rather than the issuer’s own plant. In a group whose holding company and key affiliates are separately listed, that pattern reads to the market as “our cash, their project.” The counterweight is that this is not a simple expropriation story: the holding company committed to a 120% over-subscription — the mechanism permitted under Capital Markets Act §165-6(2)(2) — so the controlling holder is adding cash, not stepping aside. Korean law already frames the directors’ task here. In 2019Da280481, the Supreme Court held that a director acquiring an affiliate’s shares must concretely examine the financial burden and expected benefit on objective evidence, and that business judgment is protected only where the corporate benefit is concrete and realistically obtainable. That was a derivatives-contract case, not a rights-offering ruling — it supplies the standard, not a verdict.
The 2025 reform raises that standard. Commercial Act §382-3, enacted in 2025 and in force from March 2026, extends directors’ duty of loyalty to shareholders and adds a duty to protect all shareholders’ interests and treat them equally. Whether that duty reaches a lawful, control-participated capital raise whose proceeds flow to a group project is an open, debated question — not a prediction of liability, and no published holding has tested it on these facts. The point is narrower: a board choosing method, price, timing, and use of proceeds now does so against an explicit all-shareholder standard.
A standard without a channel is the harder problem. In the US, securities class actions are frequent and settle materially — averaging about $42.8m in the second half of 2025 — but they rest on disclosure fraud under Rule 10b-5, not on dilution itself. Korea has had a securities class-action statute since 2005, yet it has been used roughly 11 times through 2023, with certification alone often taking more than four years. The deterrent difference is procedural access, not punishment severity. A reformed duty that minority holders cannot practically enforce moves slower than the transactions it is meant to discipline.
Signals to watch. For anyone pricing Korean equity raises, the observable signals are: whether the offering is a rights issue or a marketed sale; whether the issue price is pre-fixed or set into the market; whether proceeds fund the issuer’s own operations or an affiliate stake; whether the controlling holder subscribes alongside minorities; and whether §382-3 acquires a workable enforcement channel. The reaction to watch is not the headline drop but how these five line up.