Aiden Donghoon Lee
Chadwick International School, Incheon, Korea.
DOI: 10.4236/ojbm.2026.146164   PDF    HTML   XML   1 Downloads   45 Views   Citations

Abstract

Shareholder activism is often viewed as a mechanism for correcting governance failures and creating shareholder value. Yet much of the evidence supporting this view comes from markets with dispersed ownership, where activists primarily discipline managers. Less is known about whether activism creates lasting value in concentrated ownership systems, where conflicts arise between controlling and minority shareholders. Korea provides a timely setting. Since 2022, activist funds and retail investor coalitions have driven a rapid expansion of campaigns against family-controlled corporate groups, a wave widely described as a democratization of corporate governance and a remedy for the persistent “Korea Discount”. The distributional consequences of that wave have received little empirical scrutiny. This study examines twenty-one verifiable activist campaigns against Korean listed firms between 2022 and 2025, asking when value is created and whether it persists beyond the market’s initial response. Campaign announcements generate a significant positive market reaction, with a mean cumulative abnormal return of 6.96 percent over the [−5, +5] window. That premium, however, is front-loaded. Among campaigns with resolved governance outcomes and sufficient follow-up data, there is no statistically reliable evidence of sustained positive abnormal returns twelve months later, and the pattern is asymmetric; successful campaigns fail to show lasting gains, while the two campaigns that failed outright suffer sharply negative long-horizon returns. The findings distinguish the market’s expectation of governance improvement from evidence of lasting value creation. They also carry a distributional implication for a movement celebrated as empowering minority shareholders. Whether activism creates value and who captures that value are therefore different questions, and in concentrated ownership markets the second may mark the structural limit of what activism alone delivers.

Keywords

Shareholder Activism, Corporate Governance, Event Study, Korea Discount, Minority Shareholders

Share and Cite:

1. Introduction

Between 2022 and 2025, shareholder activism in Korea moved from a marginal practice to a central fact of corporate governance. Industry tallies recorded 77 Korean listed companies publicly subjected to activist campaigns in 2023, up from 49 a year earlier, one of the largest totals of any market in the world (Diligent Market Intelligence, 2024). The campaigns were waged not only by domestic specialist funds but by retail shareholders organized through online platforms, and much public commentary described the development in expansive terms: a “game-changing transformation” of Korean governance (Chung, 2023), a democratization of corporate power in a market long defined by controlling families, and a possible remedy for the persistent valuation gap known as the “Korea Discount”. The corporate sector answered with an opposing narrative, warning that Korean companies were becoming exposed targets for funds seeking short-term gains (Federation of Korean Industries, 2024). The two narratives disagree about who benefits from activism, but they share an assumption, namely, that the effects of these campaigns on shareholder value are large and identifiable. Neither narrative has yet been tested against campaign-level return evidence from the wave itself.

Shareholder activism is commonly evaluated by asking whether an intervention creates value for the target firm. In the United States, where ownership is relatively dispersed, event studies have repeatedly found positive stock-price reactions when activist positions become public, and influential work has found little evidence that these gains are subsequently reversed (Brav et al., 2008; Bebchuk et al., 2015). But the question of whether activism creates value can obscure a second issue. Shareholders do not all encounter an activist campaign at the same point in time. Activists generally establish positions before a campaign becomes public; existing shareholders participate in the announcement repricing; investors who enter later buy after some expected governance benefit has already been capitalized; and continuing shareholders benefit only if anticipated reform translates into durable corporate value. The timing of value creation therefore matters for understanding who is positioned to benefit from activism.

That question has particular force in Korea because the constituency in whose name the wave has been celebrated is largely retail. The number of individual investors holding Korean listed shares rose from roughly 6.1 million in 2019 to more than 14 million by 2022, meaning that a substantial share of the population now holds a direct stake in listed companies (Korea Securities Depository, 2023; Lee, 2020; Nam, 2022). Many joined activist campaigns as participants, delegating proxies through online platforms and voting at contested meetings; many more encountered activism as investors deciding whether to buy targeted stocks after campaigns became public. If the value associated with activism is priced mainly at the moment of announcement, it accrues to whoever already holds the stock in that window, and a movement joined after the announcement may deliver less to its late-arriving constituency than the celebratory narrative implies. Whether that is so is not a question about anyone’s intentions. It is an empirical question about when abnormal returns arise, and it is the question this paper addresses.

Korea also provides an institutionally distinctive setting in which to ask it. Many large Korean firms are organized within chaebol groups, family-controlled business networks in which founding families can exercise control through pyramidal ownership, affiliate stakes, and other control-enhancing structures even when their direct cash-flow ownership is comparatively modest (Claessens et al., 2000; Ducret & Isakov, 2020). The resulting governance problem is not only the familiar conflict between managers and dispersed shareholders. It also involves conflicts between controlling and minority shareholders over capital allocation, related-party transactions, succession, and the distribution of corporate value. These features are central to debates over the Korea Discount, and they mean that evidence on activism from dispersed-ownership markets cannot simply be assumed to transfer.

The best-known campaign of the wave illustrates both its promise and its interpretive difficulty. Align Partners, holding only a small stake in SM Entertainment, successfully challenged a long-standing production arrangement with a company owned by SM’s founder and helped secure governance changes that were widely treated as a breakthrough for domestic activism (Cha, 2022; Wang & Chan, 2025). The case also shows why a simple statement that “shareholders gained” can conceal different economic experiences across a campaign: Align was positioned before the campaign became a major public event, later investors encountered a stock price that already reflected changing governance expectations, and SM subsequently became the subject of a control contest involving HYBE and Kakao whose competing tender offers make the stock’s dramatic price path unattributable to activism alone (CNBC, 2023a, 2023b). Section 2.3 discusses the episode in detail. Here the case raises the question the rest of the paper pursues, namely, how the value created by activism is distributed among shareholders.

This study asks when the value associated with shareholder activism is created, whether it persists as campaigns move toward observable governance outcomes, and what that timing implies about which shareholders are positioned to benefit. It constructs a campaign-level sample of twenty-one verifiable firm-events between 2022 and 2025 using DART regulatory filings, public campaign materials, activist-fund disclosures, and systematic searches of Korean financial media. The empirical design separates three stages of engagement: the public announcement of activist involvement, the later governance milestone at which an observable outcome is confirmed, and the long-horizon post-announcement period.

The evidence is asymmetric in strength. The short-window announcement effect is clear: mean cumulative abnormal return (CAR) is +6.96 percent over [−5, +5] and +3.63 percent over [−1, +1]. The long-horizon evidence is much weaker. Among campaigns that reached a governance milestone and have sufficient subsequent data, twelve-month CARs are statistically indistinguishable from zero. The paired announcement-versus-milestone comparison runs in the expected direction but is not statistically significant. Controlling-shareholder concentration is unrelated to long-horizon CAR in the full sample and becomes negatively associated only in the smaller subsample for which long-horizon returns are observable, where the evidence is suggestive rather than conclusive. These results do not identify activist-versus-retail wealth transfers, because investor-level trades are not observed. They do, however, establish a temporal distinction: the announcement premium belongs to shareholders already holding the stock over the event window, while investors who enter after that repricing face a different return opportunity.

The paper makes three contributions. First, it provides campaign-level abnormal-return evidence for Korea’s post-2022 activist wave, a period that existing research has so far approached primarily through institutional and legal analysis, policy and industry reporting, and case studies rather than through a campaign-level event study (Section 3.2). Second, it separates announcement value from durable value and connects that distinction to the temporal distribution of gains across shareholder positions. Third, it places ownership concentration at the center of the analysis without treating the current small sample as capable of delivering a definitive structural test. In Korea, activist campaign announcements produce a clear positive market reaction, but whether the associated gains last, and who is positioned to capture them, remain open questions.

2. Institutional Background

2.1. Concentrated Ownership and Minority-Shareholder Governance

Korean corporate governance is shaped by a separation between cash-flow ownership and effective control. Research on East Asian corporations documents the widespread use of pyramidal structures and affiliated holdings that allow controlling shareholders to exercise voting power exceeding their direct economic stake (Claessens et al., 2000). In Korea, these structures are especially salient within chaebol groups. They can stabilize control and support long-term coordination across affiliates, but they also create the possibility that decisions benefiting the controlling family will not maximize proportional value for minority shareholders (La Porta et al., 1999; Ducret & Isakov, 2020).

This institutional setting changes the governance problem that activism confronts. In a dispersed-ownership firm, an activist typically seeks to overcome shareholder passivity and discipline management. In a concentrated-ownership firm, mobilizing outside shareholders may still leave the activist facing a durable controlling blockholder. Governance gains therefore depend not only on persuading other investors but also on the legal rules, voting structure, and coalition arithmetic that determine whether a minority challenge can alter corporate decisions. Korean legal scholarship has similarly emphasized the tension between the company-centered formulation of directors’ duties and the proportional interests of shareholders, particularly in conflicts involving controlling shareholders (Lee, 2023a).

The “Korea Discount” is often discussed as the market consequence of these broader governance frictions, although no single mechanism can explain it fully. In a cross-country analysis, the Korea Capital Market Institute attributes the discount most strongly to weak shareholder payouts and to low profitability and growth, with governance quality, accounting opacity, and a small institutional-investor base playing significant but secondary roles (Kim & Kang, 2023; see also Kim & Lee, 2025). Ducret and Isakov (2020) connect chaebol structure to valuation discounts, while Na and Jung (2025) identify inheritance taxation as one mechanism that can shape controlling-shareholder incentives around firm value. In this paper, such mechanisms are treated as components of a broader concentrated-ownership environment rather than as a single structural cause of low valuation. The empirical question is whether that environment also conditions what happens after activist intervention.

2.2. The Post-2022 Activist Expansion

Shareholder activism is not new to Korea. Earlier episodes included the NGO-led minority-shareholder movement after the Asian financial crisis, foreign activist and control contests in the 2000s, and National Pension Service voting activism (Choi & Cho, 2003; Kim et al., 2009; Ko & Kim, 2020). These episodes, however, differ from the post-2022 wave in both the identity of the activists and the surrounding institutions. Domestic specialist funds now conduct public campaigns using shareholder letters, online media, proxy solicitation, and coordinated voting, while individual shareholders have become a much larger presence in the market (Chung, 2023; Kamalnath, 2025).

Several institutional changes lowered the cost of such campaigns. The Korean Stewardship Code increased expectations that institutional investors explain voting decisions and engage with governance issues, and the National Pension Service became an important reference point in close votes (Kang & Chun, 2022; Kim & Chang, 2018). Electronic voting and online proxy-delegation tools reduced the logistical costs of collecting minority-shareholder support (Sherratt, 2024). At the same time, the expansion of retail participation (Figure 1, panel (a)) created a larger constituency receptive to arguments about capital allocation, shareholder returns, and the Korea Discount.

Figure 1. Growth in the retail-investor base and shareholder proposals around the emergence of Korea’s post-2022 activist wave. Panel (a) reports individual investors holding shares of December fiscal-year-end listed companies; panel (b) reports listed companies with shareholder proposals on March AGM agendas. Sources: Korea Securities Depository (2023), Lee (2020), Nam (2022), and Kim (2023).

The scale of the wave, and the language surrounding it, are both part of the institutional record. As demonstrated in Figure 1, Panel (b), The number of December year-end listed companies facing shareholder proposals at March annual general meetings rose from 28 in 2022 to 44 in 2023, an increase of 57 percent (Kim, 2023), and industry tallies count 77 Korean companies publicly subjected to activist demands in 2023, among the highest totals of any market (Diligent Market Intelligence, 2024; Cha & Ha, 2023). Commentary has ranged from descriptions of a “game-changing transformation” of Korean governance (Chung, 2023) to corporate-sector warnings that Korean firms have become disproportionately exposed targets (Federation of Korean Industries, 2024). These competing characterizations are not evidence about returns. They do, however, establish that the question of who benefits from Korean activism is contested in public discourse, which is precisely why it calls for direct investigation of campaign-level returns.

The wave also developed alongside significant corporate-law reform. The Commercial Act amendments enacted across 2025 and early 2026 expanded directors’ duties toward shareholders, strengthened cumulative voting requirements for large listed companies, and addressed treasury-share cancellation (Kim & Chang, 2025, 2026; Lin, 2025). Most campaigns in this study began before the full effect of those reforms could be observed. Their passage is therefore better treated as part of the changing institutional backdrop than as an explanation for the announcement effects estimated below.

2.3. The SM Entertainment Episode

The campaign that came to define the wave began in early 2022, when Align Partners, a domestic fund holding roughly one percent of SM Entertainment, publicly challenged the company’s long-standing practice of paying substantial producing fees to Like Planning, a private company owned by SM’s founder and chief producer (Cha, 2022; Wang & Chan, 2025). The demand was a routine related-party-transaction complaint. Its resonance was much larger: it framed a prominent company’s governance as a case of value flowing from all shareholders to a controlling insider, and it did so in public, through shareholder letters and media engagement, at a moment when SM’s retail shareholder base was large and attentive. Over the following year the campaign achieved most of what it sought. SM ended the Like Planning arrangement, and by the March 2023 general meeting the board had been reconstituted, an outcome widely treated as the first decisive domestic-activist victory of the wave (Wang & Chan, 2025).

What happened next is the reason SM cannot serve as a clean measure of what activism alone produces. The governance contest escalated into a control contest. SM’s management agreed to issue new shares and convertible bonds to Kakao; the founder obtained a court injunction blocking the issuance, in a decision that prompted significant Korean legal debate about the proportionate interests of shareholders (Lee, 2023a); HYBE purchased the founder’s stake and launched a tender offer; and Kakao responded with a competing tender offer before HYBE withdrew (CNBC, 2023a, 2023b). SM’s share price rose dramatically over this period, but the increase reflected a bidding war for control at least as much as any repricing of governance reform.

Figure 2. Reported SM Entertainment share-price levels at key campaign events, February 2022 to May 2023. Points reproduce levels reported in contemporaneous coverage of the campaign and the 2023 control contest (Cha, 2022; CNBC, 2023a, 2023b; Park & Seo, 2023; Peoples, 2023); the connecting line is schematic rather than a daily price series, and the event axis is not to scale.

The aftermath turned the distributional question into public controversy. Within weeks, Korean media reported that Align’s founder had sold SM shares at approximately 110,000 won, shortly after publicly suggesting that governance improvement could carry the price to 300,000 won (Cha, 2023a; Lee, 2023b; Lee & Kim, 2023). Align responded that the shares sold belonged to the founder’s holding company, not to the fund, whose stake in SM was unchanged at 268,500 shares, or 1.12 percent, and that the sale covered the holding company’s debt repayment and operating costs (Align Partners Capital Management, 2023). This paper takes no position on these explanations; the reception, however, merits discussion. Despite the rebuttals, the episode was widely received as exposing a gap between an activist fund’s public advocacy and its private economics, and many of the retail shareholders who had rallied behind the campaign were left disappointed (Cha, 2023b; Lee, 2023b). By May 2023 the shares had fallen to roughly 100,000 won, about 38 percent below the March peak (Figure 2), and outlets that had covered the campaign sympathetically began cautioning retail investors against buying on the strength of activist campaigns alone (Cha, 2023b; Park & Seo, 2023). The episode shows that the question this paper studies, when activist-associated value is priced and who is positioned to capture it, was being asked in real time by the market itself.

3. Literature Review and Hypothesis Development

3.1. Announcement Effects and Long-Run Performance in the Activism Literature

The international literature provides a clear benchmark for the first stage of an activist campaign. Brav et al. (2008) document positive abnormal returns of roughly seven percent around U.S. activist disclosures, and Klein and Zur (2009) find comparable positive announcement effects for campaigns run by hedge funds and other entrepreneurial activists. Becht et al. (2017) extend the evidence internationally and find substantial announcement effects across regions, with returns concentrated in engagements that achieve concrete outcomes. Activist reputation, campaign type, and the probability of a concrete outcome all affect the magnitude of the market response (Krishnan et al., 2016; Becht et al., 2017). A positive announcement CAR is therefore usually interpreted as the market capitalizing expected improvements in governance, operations, payout policy, or control. Surveying three decades of this literature, Denes et al. (2017) conclude that activism has become more value-increasing over time, and that its measured effects depend heavily on the form of the intervention: interventions in which the activist accumulates a significant ownership block are associated with improvements in share value and firm operations, while activism without block formation shows insignificant or very small effects.

The harder question is whether that initial premium persists. Bebchuk et al. (2015) find no evidence of long-term value destruction following U.S. hedge-fund activism, while de Haan et al. (2019) show that long-run estimates are sensitive to weighting and sample composition. Greenwood and Schor (2009) further show that a substantial portion of U.S. activist gains is concentrated in targets that are subsequently acquired. These findings make it inappropriate to assume that a short-window activism premium will have the same persistence in a market where ownership is more concentrated and takeover channels are more constrained.

3.2. Shareholder Activism Research in Korea

Korean shareholder activism did not begin in 2022. A first wave followed the Asian financial crisis, when the NGO People’s Solidarity for Participatory Democracy (PSPD) used minority-shareholder rights to challenge governance at major chaebol firms. Choi and Cho (2003) analyze these campaigns and the market’s reaction to them, establishing both the feasibility of minority-shareholder challenge in Korea and its dependence on a small number of committed organizers rather than on broad investor participation. A second wave consisted of foreign funds and control contests in the 2000s, including the widely discussed confrontations at SK and KT&G. Kim et al. (2009) study outside blockholders that switched their disclosed purpose from passive to active and find positive announcement returns. A third, institutional wave followed the introduction of the Stewardship Code and the National Pension Service’s adoption of it: Kim et al. (2020) document changes in institutional voting behavior, Ko and Kim (2020) show that market reactions to NPS voting intentions depend on media dissemination, and Kang (2022) offers a more skeptical assessment of NPS activism as an instrument exposed to political influence.

Each of these waves differs from the post-2022 wave in ways that limit the transferability of its evidence. The PSPD movement was civic rather than financial, and its campaigns did not aim primarily at shareholder returns. The foreign-fund episodes involved outside acquirers of large blocks confronting individual chaebol groups, and they predate the growth of retail participation. NPS activism operates through the voting of a single large quasi-public institution rather than through public campaigns by small funds. The current wave’s defining features, domestic specialist funds with small stakes, mass retail participation, online proxy platforms, and demands centered on payout policy and treasury shares, have no close analogue in the earlier periods (Chung, 2023; Kamalnath, 2025). Evidence from these earlier periods therefore says little about the return dynamics of the current wave.

Research that addresses the post-2022 wave directly falls into four strands, each with characteristic strengths and limits. The first is institutional and legal analysis. Kang and Chun (2022) trace how the Stewardship Code reshaped the incentives of Korean institutional investors and created preconditions for public campaigns; Lee (2023a) uses the SM litigation to analyze whether directors owe duties to the proportionate interests of shareholders; and Kamalnath (2025) examines how social media is changing the mechanics of shareholder coordination. This strand explains why campaigns became possible and how they are fought, but it offers little systematic evidence on what they have been worth to shareholders.

The second strand is policy and industry reporting. Hong (2023), writing for the Korea Capital Market Institute, documents the growth in activist funds and draws implications for market policy, while Diligent Market Intelligence (2024) and the Federation of Korean Industries (2024) provide the campaign tallies cited above. These sources are the best available record of the wave’s scale, but they are descriptive, aggregate rather than campaign-level, and in some cases produced by parties with positions in the policy debate.

The third strand is systematic empirical work on adjacent outcomes. Kang et al. (2025) study the determinants of shareholder proposals across Korean listed companies, the study closest in method to this one; its outcome variable, however, is the incidence of proposals rather than their price effects, and many prominent campaigns proceed through letters and negotiated settlements without a formal proposal.

The fourth strand is the case study. Wang and Chan (2025) reconstruct the Align Partners campaign at SM Entertainment in detail, offering a depth no event study can match, but a single case, particularly one entangled with a takeover battle, supports no inference about the distribution of outcomes across campaigns.

To the best of the author’s knowledge, no strand provides campaign-level announcement and long-horizon return estimates for the post-2022 wave, and none engages directly the question of when activist-associated value is priced and who is positioned to receive it. That is the gap this study addresses.

3.3. The Distributional Question

Most event studies summarize activist value creation at the target-firm level: if the stock earns a positive abnormal return around disclosure, “shareholders” are said to gain. That aggregation is analytically useful but can hide differences in how gains arise and who is positioned to receive them.

If activist campaigns reliably produced durable value, the timing of entry would carry little weight. Late entrants would pay more, but they would hold claims on a lastingly better firm, and the difference would be ordinary market mechanics. Entry timing becomes a distributional question only when the value priced at announcement is not reliably realized afterward, because what the late entrant bought is then, in part, an expectation that did not hold; whether that is the case in Korea is precisely what the persistence tests below examine. The question is sharpened by how the wave was built. Activists actively recruited retail shareholders into these campaigns, through proxy-delegation drives, electronic voting, and sustained media engagement, so the investors most exposed to unrealized expectations are the ones the campaigns themselves mobilized.

Campaigns are waged in the name of minority shareholders as a group, but the economic experience of that group is not uniform. An early shareholder, a post-announcement buyer, and a shareholder who remains invested through a governance milestone occupy different positions in the campaign lifecycle even if all three are “minority shareholders”. The present data cannot identify those investors by class, and the paper therefore does not claim that activists systematically win while retail investors systematically lose. Instead, it asks a narrower distributional question: at what stage is activist-associated value priced, and what does that timing imply about which shareholder positions are exposed to the upside and downside of the campaign?

3.4. Hypotheses

The institutional argument developed above generates a sequence of testable predictions tracing the lifecycle of activist-generated value.

H1. Activist campaign announcements generate significantly positive cumulative abnormal returns, reflecting the market’s initial pricing of expected governance improvement.

H2. Among campaigns that achieve a governance milestone, long-horizon (twelve-month) post-announcement CARs are not significantly positive; the campaign generates no detectable abnormal performance beyond the repricing that occurs at announcement.

H3. Long-horizon CAR following an activist campaign is negatively associated with controlling-shareholder ownership concentration: greater concentration predicts weaker post-announcement performance.

H4. The CAR observed around the governance milestone is smaller than the CAR generated at campaign announcement, indicating that the market prices the arrival of the activist more than the realization of governance reform.

H5 (exploratory). The durability of governance milestones, that is, the persistence of their effect over time, varies across campaigns; this is examined qualitatively through case evidence rather than formal testing.

Confirmation of H1 alongside H2 would establish the paper’s central temporal claim that value is priced at announcement but is not retained over the longer horizon. H3 links this dynamic to the structural features of concentrated ownership, and H4 provides a diagnostic distinguishing expectation-driven pricing from reform-driven pricing.

4. Data and Sample Construction

4.1. Sample Construction

Activist campaigns were identified through three complementary sources. The first was DART (Data Analysis, Retrieval and Transfer System) regulatory filings: major-holding disclosures triggered when an investor acquires 5 percent or more of a listed firm, and shareholder-proposal filings submitted ahead of annual general meetings. The second was public campaign material, including shareholder letters, press releases, and public statements, identified through a systematic search of Korean financial media (The Bell, Maeil Business Newspaper, the Korea Economic Daily, and the Seoul Economic Daily) and activist-fund websites. The third was the author’s database of corporate-governance conflict cases, compiled during a research internship at Action for Corporate Transparency (ACT), which provided an initial sample frame.

Initial screening produced thirty-seven candidate campaigns. Each candidate was then verified individually against primary disclosures and contemporaneous media, yielding a final analysis sample of twenty-one firm-events. Fourteen candidates were excluded, and the grounds for exclusion fall into three categories. Seven were determined not to be activism at all but rather private-equity or buyout transactions; these were concentrated among KCGI campaigns, reflecting the firm’s documented 2025 pivot away from public activism toward principal investment and acquisition, such that its later positions represent friendly or control-oriented deals rather than governance challenges. Six were excluded because no public campaign could be verified: although a 5 percent holding was disclosed, no public demand, shareholder proposal, or governance outcome was observable, a pattern consistent with quiet accumulation rather than activism (most of these involved Truston Asset Management). One was excluded because the campaign began in 2018, outside the sample window (Hanjin KAL).

This screening-then-verification procedure is itself methodologically consequential. Identifying campaigns by the reputation of well-known activist funds generates systematic false positives, most visibly in the case of KCGI, and case-level verification is required to remove them. The Appendix records the primary-source basis for each campaign in the analysis sample and for each candidate campaign excluded during screening.

A further construction decision concerns the Samsung C&T campaign, in which a coalition of five funds targeted the same firm at the same time. Because the unit of analysis is the firm-event, and because all coalition members share an identical price series and identical event dates, treating them as separate observations would triple-count a single event and bias the cross-sectional statistics. The three disclosed filings were therefore consolidated into a single firm-event. Figure 3 summarizes the full construction of the sample, from the thirty-seven candidates through the exclusions and the coalition consolidation to the twenty-one analysis events.

Figure 3. Sample construction. Thirty-seven candidate campaigns were screened against primary disclosures; exclusions (private-equity/buyout transactions, unverifiable public campaigns, one pre-2022 campaign) and the consolidation of the three Samsung C&T coalition filings into a single firm-event yield the twenty-one-event analysis sample.

For each campaign, three dates were recorded where available: the announcement date (the earliest public disclosure of activist engagement), the milestone date (the date on which the most significant governance outcome, typically an AGM vote or board resolution, was confirmed), and the exit date (the date of any public disclosure of a position reduction). Exit dates are observable only in a minority of cases, because most activist funds in the sample held sub-5 percent positions that carry no mandatory disclosure obligation. For this reason, the design measures long-horizon performance from the announcement date rather than from the exit date; Section 7.1 discusses this substitution.

4.2. Outcome Classification

Each campaign was classified according to what the target firm actually changed rather than the activist’s rate of demand fulfillment, which is unobservable and introduces endogeneity. Conditioning the classification on post-announcement stock performance, or on the activist’s inferred private expectations, would contaminate the very dependent variable the study seeks to explain; the classification therefore relies only on documented corporate action.

Campaigns were coded as Successful when the primary demand was met, irrespective of whether the realized figure matched the activist’s headline ask (nine cases); as Partially Successful when a multi-item demand was only partly met (nine cases); as Unsuccessful when all core demands were rejected or withdrawn (two cases); and as Unresolved in the single case whose governing AGM outcome had not been confirmed at the time of analysis (A+ Asset). Where a governance outcome was achieved and subsequently neutralized (for example, an audit-committee seat won under the 3 percent rule and later lost through affiliate stake-splitting) the campaign is classified by its status at the milestone, with the question of subsequent durability treated separately under H5. This separation prevents the outcome variable from absorbing the durability dynamic it is meant to be distinguished from.

4.3. Sample Eligibility for Long-Horizon Tests

The long-horizon tests (H2 and H4) require that a campaign’s governance milestone be observable and that the twelve-month (approximately 250 trading day) window following that milestone fall within the available price history, which extends to mid-2026. Eligibility is therefore determined by the milestone date, while the long-horizon CAR itself is accumulated from the announcement date. The two dates differ, so a campaign can have ample post-announcement price history and still be ineligible because its milestone falls beyond the sample period; Tovis, announced in February 2022 with a milestone in March 2026, is the clearest such case. A three-level eligibility flag was therefore constructed. Fourteen firm-events are coded yes, with a comfortable margin of post-event data (milestones dated 2024 or earlier); one is coded marginal, where the window reaches the very edge of the available data (Coway, with a 37-day margin); and six are coded no, because their milestones fall in 2026 and the post-event window extends beyond the data, or because the outcome is unresolved. H1, which concerns the announcement window, uses all twenty-one firm-events. Of the fourteen eligible events, the twelve that achieved a governance milestone form the core long-horizon sample, rising to thirteen when the marginal case is added in a robustness specification; the two unsuccessful campaigns, though within the data window, are analyzed separately in the outcome analysis of Section 6.5. The paired H4 test uses the thirteen firms for which both the announcement and milestone windows are observable. This rule makes the inclusion of a firm in any given test explicit and non-arbitrary rather than a matter of researcher discretion. Table 1 lists the twenty-one campaigns with their announcement and milestone dates, outcome codes, and eligibility flags.

Table 1. Twenty-one activist campaigns in the analysis sample, ordered by announcement date.

#	Activist fund	Target firm	Announced	Milestone	Outcome	Elig.
1	Truston Asset Management	Taekwang Industrial	2022-02-09	2024-03-29	Partially successful	Yes
2	Cha Partners	Tovis	2022-02-24	2026-03-26	Successful	No
3	Align Partners	SM Entertainment	2022-02-28	2023-02-03	Successful	Yes
4	Cha Partners	Sajo Oyang	2022-03-03	2022-03-24	Partially successful	Yes
5	Anda Asset Management	SK Chemicals	2022-03-15	2022-09-28	Partially successful	Yes
6	Align Partners	JB Financial Group	2022-05-20	2024-03-28	Successful	Yes
7	Truston Asset Management	Korea Alcohol Industrial	2022-09-21	2023-03-28	Successful	Yes
8	Truston Asset Management	BYC	2022-12-20	2023-03-24	Partially successful	Yes
9	Truston Asset Management	LF Corp	2022-12-22	2024-03-20	Partially successful	Yes
10	Cha Partners	Namyang Dairy	2023-03-09	2023-03-31	Partially successful	Yes
11	Anda Asset Management	KT&G	2023-03-14	2023-03-28	Unsuccessful	Yes
12	Five-fund coalition	Samsung C&T	2024-02-27	2024-03-15	Unsuccessful	Yes
13	Cha Partners	Kumho Petrochemical	2024-03-06	2024-03-22	Partially successful	Yes
14	Dalton Investments	Kolmar Holdings	2024-11-06	2025-03-31	Successful	Yes
15	Align Partners	Doosan Enerbility	2024-11-27	2024-12-10	Successful	Yes
16	Align Partners	Coway	2025-02-14	2025-05-08	Partially successful	Marginal
17	Align Partners	Gabia	2025-03-27	2026-03-26	Successful	No
18	Align Partners	STIC Investments	2025-03-28	2026-01-20	Successful	No
19	Align Partners	Dentium	2025-03-31	2026-03-31	Partially successful	No
20	Align Partners	SoluM	2025-09-04	2026-03-16	Successful	No
21	Align Partners	A+ Asset Advisor	2025-11-18	-	Unresolved	No

Notes: Announcement dates are the earliest public disclosure of activist engagement; milestone dates are the date on which the most significant governance outcome was confirmed (Section 4.1). Outcome coding follows Section 4.2: nine campaigns are coded Successful, nine Partially successful, two Unsuccessful, and one Unresolved. Elig. reports the long-horizon eligibility flag of Section 4.3 (fourteen Yes, one Marginal, six No); the twelve Yes-flagged campaigns coded Successful or Partially successful form the core long-horizon sample, and the Marginal case (Coway) is added in the robustness specification. Entry 12 consolidates three separately filed 5 percent disclosures, by City of London Investment Management, Whitebox Advisors, and Anda Asset Management, into a single firm-event representing a five-fund coalition. Exit dates are observable for only three campaigns and are therefore not tabulated (Section 7.1).

4.4. Stock Price and Ownership Data

Daily adjusted closing prices for all twenty-one firms and for the two market benchmarks (the KOSPI (^KS11) and KOSDAQ (^KQ11) indices) were retrieved from Yahoo Finance over a window of approximately eighteen months on either side of each announcement. Adjusted closing prices were used throughout, incorporating split and dividend adjustments as required by the market model.

Controlling-shareholder ownership was obtained from DART’s largest-shareholder disclosures, taken from the most recent periodic report preceding each announcement, so that the measured stake reflects ownership at the onset of the campaign. The combined stake of the largest shareholder and related parties was recovered for all twenty-one firms, ranging from 7.44 percent (KT&G, a dispersed-ownership firm) to 62.68 percent (BYC), with a mean of 32.41 percent. The breadth of this distribution is important for H3, since testing whether ownership concentration predicts weaker long-horizon performance requires meaningful variation in the explanatory variable. Two control variables were collected: foreign ownership, matched to the announcement date (sourced from Naver Finance, as the KRX data service was inaccessible and the DART API does not report aggregate foreign ownership), and National Pension Service (NPS) ownership, available only where it exceeds the 5 percent disclosure threshold and therefore present for only two firms. Market capitalization was computed as the announcement-date closing price multiplied by shares outstanding from DART.

5. Methodology

5.1. The Market Model

The study employs the event-study methodology of Fama, Fisher, Jensen, and Roll (1969), as formalized by MacKinlay (1997). Expected returns are estimated using the market model,

$R_{it}={\text{α}}_i+{\text{β}}_i{R}_{mt}+{\text{ε}}_{it}$ (1)

where $R_{it}$ is the log return on firm i on day t and $R_{mt}$ is the log return on the relevant market index. Parameters are estimated by OLS over the estimation window [−250, −30], leaving a thirty-day buffer before the event to avoid contamination from anticipatory price movement. Crucially, each firm is benchmarked against the index of its own exchange (KOSPI-listed firms against the KOSPI index and KOSDAQ-listed firms against the KOSDAQ index) because benchmarking a KOSDAQ stock against the KOSPI would misstate its expected return and distort the resulting abnormal returns. The estimated parameters fall within the normal range for individual Korean equities (betas between 0.11 and 1.23; R² between 0.02 and 0.30).

5.2. Abnormal Returns and Event Windows

The abnormal return is the residual from the market model, and the cumulative abnormal return is its sum over the event window,

${\text{AR}}_{it}=R_{it}-\left({}_{\widehat{\text{α}}}+{}_{\widehat{\text{β}}}{R}_{mt}\right),CA{R}_i\left[t_1,t_2\right]=\sum {\text{AR}}_{it}\left(t=t_1,\cdots ,t_2\right)$ (2)

Short windows ([−1, +1], [−5, +5], and [−10, +10]) capture the immediate reaction used to test H1 and H4. Long windows ([0, +30], [0, +60], [0, +120], and [0, +250]) trace the persistence of activist-generated value for H2.

For the long windows, the CAR is computed in two versions: the full market model (α + β) and a market-adjusted return that excludes the estimated alpha and retains only the beta term. The reason is that an estimated daily alpha on the order of 0.35 percent compounds to a large drift over 250 trading days, so a long-horizon CAR computed with the full model is sensitive to the alpha assumption. Reporting both versions allows the robustness of any long-horizon conclusion to be assessed against that assumption rather than presumed.

5.3. Statistical Tests

For H1 and H2, the null hypothesis is that the mean CAR equals zero. Given the small sample, each parametric one-sample t-test is accompanied by a non-parametric Wilcoxon signed-rank test, and the proportion of positive CARs is reported alongside the mean, since that proportion is less sensitive to outliers than the mean itself.

Because the sample is small, a power analysis is reported for H2: the minimum detectable effect size is computed for the eligible sample so that a null result can be correctly interpreted as either evidence of no effect or, more conservatively, as a limitation of statistical power. This distinction is treated as central rather than incidental to the interpretation of H2.

H4 is tested as a paired comparison between the announcement-window and milestone-window CARs for the same firms, again using both a paired t-test and a Wilcoxon signed-rank test.

H3 is tested by cross-sectional OLS with HC3 heteroskedasticity-robust standard errors,

${\text{CAR}}_i[0,+250]=\text{α}+{\text{β}}_1\cdot {\text{Controlling}}_i+{\text{β}}_2\cdot {\text{Foreign}}_i+{\text{β}}_3\cdot \ln \left({\text{MktCap}}_i\right)+{\text{ε}}_i$ (3)

where the coefficient of interest is ${\text{β}}_1$ , predicted to be negative under H3. A specification adding a campaign-outcome term is estimated as an auxiliary model. Because the National Pension Service variable is observed for only two firms, it is excluded from the primary specification and entered only as a robustness dummy, since including it as a continuous regressor would collapse the model through listwise deletion.

5.4. Robustness Checks

The regression is specified on two samples. H3 asks whether an activist-generated gain dissipates as ownership concentration rises, which presupposes that the campaign has run far enough for a governance outcome to be observable within the return window. The primary estimation sample is therefore the fifteen firm-events that satisfy the long-horizon eligibility rule of Section 4.3, comprising the fourteen events flagged Yes and the one flagged Marginal. The six excluded events are those whose governance milestones fall in 2026, beyond the available price history, or whose outcome remains unresolved. The pooled specification on all twenty-one events is reported alongside the primary estimate.

Several robustness exercises are built into the design. Long-horizon results are reported under both the full and market-adjusted models. H2 is estimated across all four long windows so that the short-to-long trajectory of the CAR can be examined rather than inferred from a single endpoint. H3 is estimated on both samples defined above, and the eligible-sample result is compared with the pooled estimate. Finally, the outcome variable is coded both as a binary indicator (Successful versus other) and as an ordinal scale, and the two codings are compared.

6. Results

6.1. Announcement Effects (H1)

The announcement of an activist campaign produces a significant positive market reaction, consistent with H1. Over the [−5, +5] window, the mean CAR across the twenty-one firm-events is +6.96 percent, significant under both the parametric t-test (p = 0.022) and the Wilcoxon signed-rank test (p = 0.026), with 66.7 percent of events (fourteen of the twenty-one) exhibiting positive CARs. Figure 4, panel (a), summarizes the announcement-window estimates.

The economic magnitude of this effect warrants emphasis alongside its statistical significance. An excess return of nearly 7 percent, realized within an eleven-trading-day window, represents a substantial repricing of the target firm around the news that an activist has arrived. Announcement and long-horizon windows were screened against contemporaneous Korean financial media and DART filings for firm-specific events other than the campaign itself. A concurrent event was identified for eleven of the twenty-one firm-events and is recorded case by case in Table A3 of the Appendix; the most consequential are SM Entertainment’s control contest, Kumho Petrochemical’s family ownership dispute, Doosan Enerbility’s group restructuring, the contested control transfer at Namyang Dairy, and a second activist fund campaigning simultaneously at KT&G. The estimates are therefore best read as market reactions associated with activist engagement rather than as effects caused by it alone. At that point nothing at the company has actually changed: no demand has been met, no board seat won, no governance reform enacted. The market is capitalizing the expected value of future improvement into the share price at the moment of mobilization, and the premium is priced on anticipation, not realization.

The narrower [−1, +1] window yields a mean CAR of +3.63 percent, marginally outside the 5 percent threshold on the parametric test (p = 0.0504) but significant on the non-parametric test (Wilcoxon p = 0.042), indicating that roughly half of the total announcement effect is already reflected in the price within the immediate three-day window around disclosure. The effect is therefore robust to the choice of window and to the parametric assumption, and its magnitude is closely comparable to the announcement effects documented for U.S. hedge fund activism (approximately 7 percent; Brav et al., 2008) and for Asian engagements internationally (6.4 percent; Becht et al., 2017), a notable convergence given the fundamentally different ownership environment. This establishes the first stage of the value lifecycle traced by the hypotheses: the market prices the arrival of an activist as good news, and it does so immediately, substantially, and in advance of any realized reform.

6.2. Long-Horizon Drift (H2)

The central test concerns what happens after the announcement window closes. Because the [0, +250] window accumulates abnormal return from the announcement date forward, it measures whether a campaign delivers performance beyond the repricing already capitalized at announcement. To address whether that initial repricing was itself retained, a cumulative measure spanning the pre-announcement date through the annual horizon is also reported. Among the eligible campaigns that achieved a governance milestone, the mean [0, +250] CAR is statistically indistinguishable from zero in every specification. In the core sample (N = 12), the full-model mean is −2.4 percent (p = 0.72); in the robust sample (N = 13), it is −6.7 percent (p = 0.38). Extending the accumulation back to five trading days before the announcement changes almost nothing: the mean CAR [−5, +250] is −2.4 percent in the core sample (95 percent confidence interval [−16.0, +11.1]; t-test p = 0.70; Wilcoxon p = 0.73) and −6.9 percent in the robust sample (95 percent confidence interval [−22.5, +8.8]; p = 0.36). The reason is that these campaigns display almost no pre-announcement drift: the mean CAR [−5, −1] is −0.05 percent in the core sample and −0.13 percent in the robust sample. The announcement premium is therefore not visible at the annual horizon whether or not the announcement window itself is included in the accumulation. The market-adjusted versions, which exclude the estimated alpha, are also insignificant but positive in sign (+6.6 percent core and +3.2 percent robust over [−5, +250], both p ≥ 0.54), so the direction of the long-horizon point estimate depends on the alpha assumption even though the absence of a detectable effect does not. No specification yields a p-value below 0.33.

This null result, however, must be interpreted through the lens of statistical power rather than read at face value as evidence of no effect. Statistical power is the probability that a test will detect a genuine effect if one exists; with very few observations, even a real effect can easily go undetected. The power analysis makes the constraint concrete. With twelve eligible events, the minimum effect detectable at conventional significance is a CAR of approximately 19.7 percentage points, corresponding to Cohen’s d of 0.889, an effect nearly a full standard deviation of the observed CAR distribution; with thirteen events, the threshold rises to approximately 22.3 percentage points. In plain terms, the test could register significance only if the average successful campaign caused its target to underperform the market by roughly one-fifth of firm value over the following year, an implausibly extreme effect that no theory of activism, including this paper’s, predicts. Against the effect sizes actually observed, the achieved power is 0.06 in the core sample and 0.14 in the robust sample: even if the true long-horizon effect were exactly as negative as the point estimates suggest, a sample of this size would fail to detect it between 86 and 94 percent of the time. The correct conclusion is therefore neither that the long-horizon effect is zero nor that the framework is unsupported, but that no effect short of an unrealistically massive one could have cleared the significance bar at this sample size. What the data do establish is narrower and, for that reason, more defensible: there is no evidence that successful campaigns generate sustained positive long-horizon abnormal returns. The announcement premium documented under H1 is not detectably retained twelve months on.

Examining the CAR across successive windows reinforces this reading while cautioning against over-interpretation. The trajectory is non-monotonic; it does not move in one direction: mildly negative at [0, +30] (−2.3 percent), modestly positive at [0, +60] and [0, +120] (+1.5 to +2.5 percent), and negative again at [0, +250], where the robust sample drifts to −6.7 percent. The day-level mean CAR path, plotted in Figure 4, panel (b), rises to a local high near +2.9 percent in the days immediately following announcement, recovers to a mid-horizon peak of roughly +6.5 percent around day 139, and then drifts downward. In the robust sample this drift turns negative beyond approximately day 170 and reaches about −7 percent by the annual horizon, whereas the core sample shows a brief return to positive territory around days 207 to 221 before ending near −2.4 percent. The proportion of events with positive CARs falls from about 50 percent in the short run to 38.5 percent (roughly two in five) at the annual horizon. Taken together, the pattern is suggestive of long-horizon drift away from the announcement premium, and possibly toward partial reversal; but, given the power limitation, it is reported as a directional tendency rather than a confirmed result.

Figure 4. Cumulative abnormal returns from the event-study estimates. Panel (a): mean announcement-window CARs across all twenty-one firm-events, with parametric (t) and Wilcoxon (W) p-values. Panel (b): day-level mean CAR path for milestone-achieving campaigns, full market model, core (N = 12) and robust (N = 13) samples, cumulated from five trading days before the announcement so that the end point is the CAR[−5, +250] reported in Section 6.2; the series is smoothed with a five-day centered moving average for readability, and markers and end-point labels are the unsmoothed estimates. No long-horizon window differs significantly from zero (all p ≥ 0.33; minimum detectable effect ≈ 19.7 percentage points).

6.3. Ownership Concentration and Long-Horizon Performance (H3)

H3 holds that concentrated ownership limits what an activist campaign can deliver: the more tightly a firm is controlled by its dominant shareholder, the weaker its abnormal performance over the year following the campaign. In regression terms, this predicts a negative coefficient on the controlling stake. Because Equation (3) regresses the post-announcement CAR on ownership without conditioning on each firm’s announcement CAR, the coefficient identifies an association with subsequent performance; it does not identify whether ownership concentration predicts dissipation of the initial announcement effect.

From Figure 5’s estimation of coefficients of the fifteen firm-events that satisfy the long-horizon eligibility rule of Section 4.3, the controlling-stake coefficient is negative, in the direction H3 predicts (−0.017 in the market-adjusted specification, p = 0.077), and foreign ownership enters with a similar negative sign. The controlling-stake coefficient is marginal rather than conventionally significant, and the sub-sample is small. Foreign ownership is the one long-horizon coefficient that clears conventional significance in either specification (p = 0.014 in the full model, p = 0.054 market-adjusted), though at fifteen observations, and with a control variable that was not the object of a stated hypothesis, this is best read as a pattern worth testing on a larger sample rather than as a finding of the present design. In the pooled sample of all twenty-one firm-events, which adds the six events whose milestones fall beyond the available price history or whose outcome is unresolved, the coefficient on controlling ownership is essentially zero and statistically insignificant. Two of the six added events have post-announcement windows truncated by the end of the price history, at 185 and 138 trading days; excluding them leaves the pooled coefficient negative but still indistinguishable from zero. The contrast is consistent with the association being diluted once firms in which it could not operate are pooled in, but it is equally consistent with the sub-sample result being noise; at fifteen observations the two readings cannot be separated. H3 is therefore neither confirmed nor cleanly rejected; it survives as a conditional tendency in the eligible sample.

Figure 5. H3 coefficient estimates: controlling stake and foreign ownership in the [0, +250] CAR regression, eligible sample (N = 15), full-model and market-adjusted specifications, with 95 percent confidence intervals based on HC3 heteroskedasticity-robust standard errors. All point estimates are negative, in the direction H3 predicts. The controlling-stake coefficients are marginally significant at best; foreign ownership is significant in the full-model specification and marginal in the market-adjusted one.

6.4. Expectation versus Realization (H4)

H4 asks whether the market reacts more strongly to the arrival of the activist than to the actual delivery of reform; that is, whether the milestone event, the moment a governance demand is confirmed as won at an AGM vote or board resolution, moves the price less than the original announcement did. The test is a paired comparison, measuring both reactions within the same firms so that differences between firms cancel out. The result runs in the predicted direction but is not significant: across the thirteen firms with both events eligible, the mean announcement-window CAR ([−5, +5]) is +1.02 percent, against −0.50 percent at the milestone, a difference of +1.52 percentage points (paired t-test p = 0.64; Wilcoxon p = 0.54). Figure 6, panel (a), displays the comparison. The sign is consistent with the prediction (more value is priced when the activist arrives than when reform is confirmed), but the small sample again precludes a confident claim. The paired-sample announcement CAR (+1.02 percent) is also much smaller than the full-sample announcement effect (+6.96 percent), indicating that the largest announcement reactions occur disproportionately among campaigns without an observable milestone window; the paired comparison should therefore be read as evidence about milestone pricing, not as a restatement of H1.

6.5. An Asymmetry by Outcome

A supplementary regression of long-horizon CAR on campaign outcome yields a result that complicates any simple statement that activism either creates or destroys value, and is reported here in full. The coefficient on campaign success is consistently positive across all specifications. Successful campaigns are associated with relatively higher (that is, less negative) long-horizon returns than failed ones. This does not contradict H2. H2 concerns the absolute level of long-horizon CARs among successful campaigns, which is statistically indistinguishable from zero; the regression concerns the relative ranking, in which success ranks above failure even though both sit at or below zero in absolute terms.

The economic content of the asymmetry emerges from the two ends of the outcome distribution (Figure 6, panel (b)). Campaigns that achieved a governance milestone average −2.4 percent over the twelve-month horizon, a figure statistically indistinguishable from zero (p = 0.72): the governance premium priced at announcement is, on average, neither extended nor fully surrendered. Failed campaigns, by contrast, exhibit pronounced long-horizon losses (KT&G at −13 percent and Samsung C&T at −27 percent), magnitudes that exceed the announcement premium itself. Both failure cases carry concurrent events documented in Table A3: KT&G faced a second activist fund contesting the same meeting, and the Samsung C&T event consolidates a five-fund coalition. The magnitudes should therefore be read as the market’s response to a failed governance challenge at those firms rather than to any single fund’s campaign. The market, in other words, does not merely give back the expectation of reform when a campaign fails; it imposes an additional penalty, plausibly reflecting what the failure reveals: that the controlling shareholder is more entrenched than previously believed, and that any future governance challenge is correspondingly less likely to succeed. The accurate characterization is therefore not that success destroys long-horizon value, but that failure destroys value while success merely fails to create it.

Viewed from the position of an investor who buys at announcement-inflated prices, the payoff structure is starkly asymmetric. If the campaign succeeds, the realistic gain over the following year is approximately zero; if it fails, the loss is roughly one-eighth to one-quarter of invested value after adjusting for the market. The post-announcement entrant thus holds a position with the return profile of a written option: the upside is capped near zero while the downside is deep. In plainer terms, this investor is exposed to most of the risk while positioned to capture almost none of the reward, and takes that position at a price already inflated by the very expectation whose failure defines the loss.

Figure 6. The asymmetry of activist-generated returns. Panel (a) Paired announcement- versus milestone-window CARs for the thirteen firms with both events eligible (Section 6.4). Panel (b) Long-horizon [0, +250] CARs by campaign outcome. Milestone-achieving campaigns (N = 12) average −2.4 percent, statistically indistinguishable from zero, while the two campaigns that failed outright show losses of −13 and −27 percent (Section 6.5).

6.6. Summary

The results form a coherent if statistically modest picture. The announcement effect is real and robust (H1). No abnormal performance beyond the announcement repricing is detectable over the twelve-month horizon (H2). Ownership concentration is negatively associated with long-horizon performance within the eligible sample, and only marginally (H3). The market prices the activist’s arrival somewhat more heavily than the realization of reform, in the predicted direction but without significance (H4). And failed campaigns, rather than successful ones, account for the deepest long-horizon losses. H5, the durability of governance milestones, could be examined only through cases: the milestone-neutralization example recorded in Section 4.2 shows that a formally won outcome can later be reversed, but the sample contains too few such sequences for systematic analysis. The only result that clears conventional statistical significance is H1; the remainder are best read as directionally consistent with the paper’s temporal account of activist value while falling short of confirmation under the power available in a sample of this size.

Read together, these results support a temporal account of how activist-associated value is created and distributed. The sequence runs as follows. The market prices the arrival of the activist immediately and substantially (H1), and directionally, though without statistical significance, prices arrival more heavily than the realization of reform (H4). The premium is expectation-driven, capitalized before any governance change occurs. What follows depends on whether the expectation is validated. When the campaign succeeds, the premium erodes toward zero over the annual horizon (H2); when it fails, the expectation is unwound violently, with losses far exceeding the original premium (Section 6.5). The distributional implication follows from the timing structure of the campaign. The investors positioned to capture the announcement premium are those already holding shares before mobilization becomes public. The investors exposed to the subsequent drift, and to the failure tail, are those who enter during or after the public phase, at prices already inflated by the expectation the campaign itself generated. In Korea’s retail-heavy market, that later-entering population is disproportionately composed of individual investors. This distributional reading is an inference from the timing of returns rather than a direct observation of investor-level trading, a limitation Section 7 makes explicit; but on that reading, the asymmetry of Section 6.5 is not an anomaly in this account; it is its sharpest expression. The campaign converts a legitimate governance grievance into an expectation premium; the premium accrues to those who held their positions before the grievance was publicized; and the risk of the expectation failing is borne disproportionately by those who entered during the campaign’s public phase.

7. Limitations and Future Research

7.1. Limitations

Sample size and statistical power. The most consequential limitation is the size of the sample. Twenty-one firm-events, reduced to twelve or thirteen for the long-horizon tests, place severe constraints on statistical power: as Section 6.2 reports, the long-horizon tests could detect only effects on the order of twenty percentage points, far larger than any plausible true effect, so the central hypothesis (H2) and the cross-sectional hypothesis (H3) cannot be confirmed even where point estimates run in the predicted direction. This smallness is substantive rather than a sampling artifact: the resulting constraint on power is the unavoidable price of studying a phenomenon at the moment of its emergence rather than a generation afterward. A census of twenty-one campaigns cannot support the inferential machinery of a thousand-firm panel, but no larger sample of Korea’s post-2022 activist wave exists to be drawn, and the transparent hypothesis structure, auditable exclusion records, and explicit power analysis of the present design establish the baseline against which the maturing campaigns of 2025 and 2026 can be re-examined with substantially greater power. Event studies of comparably scaled samples exist for activism in other emerging markets, but conclusions about long-horizon dynamics here must be read as directional rather than definitive.

Unobservability of the exit phase. The distributional account is sharpest at exit: if value departs when the fund departs, exit would be the natural event to study. In practice, exit dates proved largely unobservable. Most funds in the sample held sub-5 percent positions carrying no mandatory disclosure obligation, so the moment of exit is recorded in only a minority of cases, often through media inference, and any exit-based subsample would carry selection bias if funds that disclose exits differ systematically from those that do not. The study therefore substitutes a long-horizon, announcement-anchored measure of non-persistence for a direct exit-event study. The substitution is defensible (value not retained at twelve months cannot have been durably retained through any interior exit date) but it is a proxy, and the cleanest test of the mechanism remains out of reach with public data.

Aggregate returns rather than investor-level flows. The event study measures abnormal returns to the firm, not returns to investor classes. The distributional interpretation (activists capture the premium, retail shareholders bear the subsequent drift) is inferred from the timing of CARs relative to the phases of engagement, not observed in trading records; the asymmetry-by-outcome finding of Section 6.5 is likewise an inference from firm-level returns sorted by outcome. Direct confirmation would require investor-class trading data; the Korea Exchange’s daily investor-type trading statistics are a natural source for such a test, which Section 7.2 develops as future work. Stated precisely, the analysis identifies the return opportunity facing shareholders positioned before the announcement and those entering after it. It does not identify which investor class actually realized each return, and no claim about losses borne by any particular group of investors is tested here.

Event-date ambiguity. Activist campaigns rarely present a single clean announcement date. Information leaks through rumor, informal engagement, and media speculation before a formal filing, which can attenuate measured announcement CARs; the extended [−10, +10] window partially captures anticipatory movement but cannot eliminate the problem. Milestone dating is similarly interpretive, particularly in cases of a first-AGM defeat followed by a second-AGM breakthrough, where the choice of meeting is consequential for H2 and H4. These judgments were made transparently and recorded, but they remain judgments.

Selection into targeting. Activists do not choose targets randomly; selection correlates with undervaluation, governance quality, and ownership structure, which themselves predict returns. The market model removes market-wide movement but not sector shocks or target-specific mean reversion, both of which can influence the post-announcement path. Two additional designs would separate these channels: a matched-peer comparison, in which each target is paired with non-targeted firms on the same exchange with similar market capitalization and industry, and a Korea-specific factor model estimated over the same windows.

Ambiguity in classifying campaign outcomes. A final limitation concerns the coding of campaign success itself. Activist demands are frequently presented as a bargaining posture. That is, an activist coalition often advances maximal or deliberately inflated proposals, knowing the firm will not concede them in full. This situation creates a genuine interpretive problem for the researcher, because the firm’s response is rarely a clean acceptance or rejection but a matter of degree, and evaluating that degree invites subjectivity. To contain this limitation, the study anchors its classification to the primary demand rather than to the proportion of concessions obtained. A campaign was coded as successful when the primary demand is met, irrespective of whether the degree of the response fell short of the demand. For instance, if the activist requested a dividend of $2 per share, and the firm responded by issuing a dividend of $1.50 per share, the campaign was coded as successful. A campaign was coded as partially successful when a multi-item demand was only partly satisfied. Indeed, this rule narrows the researcher’s interpretative discretion involved but does not eliminate it. Identifying which item in a bundled platform constitutes the primary demand, and deciding when a concession is substantive rather than cosmetic, still require human judgment, and these decisions are reflected directly into the outcome-conditioned results of H2 and the asymmetry-by-outcome analysis of Section 6.5, where the distinction between success and partial success does real interpretive work.

7.2. Future Research

This study identifies several directions for future investigation. First, the Commercial Act amendment sequence (passage in March 2025, veto in April 2025, three amendments passed between July 2025 and February 2026) provides a natural experiment for testing whether legislative governance reform creates value in concentrated ownership systems. An event study of market reactions to each legislative event, with cross-sectional variation in firm-level ownership concentration, would complement the activist-level analysis in this paper.

Second, the announcement-versus-persistence pattern documented here could be tested comparatively across other concentrated ownership markets, including Japan (post-Stewardship Code) and Southeast Asian markets. Cross-country variation in ownership concentration, legal protection, and activist activity would allow for more generalizable conclusions about the structural limits of activism.

Third, the distributional analysis could be refined with investor-class trading data. The Korea Exchange publishes daily net purchases by investor type (individual, institutional, and foreign) for each listed stock; matching those flows to campaign announcement windows would test directly whether retail net buying rises after disclosure while previously positioned investor classes reduce exposure, converting this paper’s timing-based inference into direct observation.

Fourth, the interaction between activism and legislative reform warrants further study. Do activist campaigns that coincide with or follow legislative reform produce more durable value gains? Or does legislative reform substitute for activism, reducing the marginal impact of private enforcement? These questions connect the empirical findings of this paper to the broader institutional design challenge of minority shareholder protection in concentrated ownership systems.

7.3. Conclusion

This paper set out to ask not whether shareholder activism creates value in Korea, but for whom that value is ultimately realized. The 2022 turn toward activism has been widely celebrated as a democratization of corporate governance, and the SM Entertainment campaign has become its paradigmatic success story. Against that backdrop, the paper assembled the full population of verifiable public activist campaigns against Korean listed firms between 2022 and 2025, twenty-one firm-events in total, and traced cumulative abnormal returns across the announcement, milestone, and post-campaign phases of engagement.

The evidence supports a more qualified reading of the activist wave than its public reception suggests. Campaign announcements generate a robust positive market reaction, with a mean cumulative abnormal return of roughly 7 percent over the announcement window, confirming that the market prices the arrival of an activist as good news. But that initial premium does not appear to translate into durable value for the shareholders who remain invested; long-horizon returns drift back toward, or below, pre-campaign levels, and while the small sample precludes statistical confirmation of this drift, the point estimates run consistently in the predicted direction. The results qualify both narratives described in the introduction. The market reaction to activist arrival is positive and immediate, but the gain is concentrated in the announcement period and accrues to shareholders who already hold the stock; investors who enter afterward obtain little additional return when campaigns succeed and incur substantial losses when they fail.

These patterns give empirical content to a distributional reading of the wave: campaigns mobilize dispersed shareholders around legitimate governance grievances and the resulting price appreciation is concentrated in the announcement window itself, which is available primarily to holders already positioned at that date; the subsequent drift and the failure tail fall in the period that follows it. The concrete form of that asymmetry deserves plain statement. The roughly 7 percent activism premium is realized within eleven trading days of disclosure, so it falls in a window available primarily to holders already positioned at that date; on this sample’s evidence, an investor who buys after a campaign becomes public earns additional abnormal returns near zero over the following year when the campaign succeeds, and suffers losses of 13 and 27 percent in the two observed failures. For the minority shareholders in whose name the wave has been celebrated, the voice gained through activism has therefore not yet been matched by value retained: participation in activism’s politics has, so far, run ahead of participation in its returns.

The paper’s contribution is therefore twofold. Substantively, it provides campaign-level announcement and long-horizon return estimates for Korea’s post-2022 activist wave, a period that earlier work has approached through case studies, determinants analyses, and institutional accounts, and it documents the structural limits of shareholder activism in a concentrated ownership system, where long-horizon performance following activist campaigns is not distinguishable from zero and is, in the eligible sample, negatively associated with controlling-shareholder concentration. Conceptually, it reframes the evaluative question surrounding activism from value creation to value distribution, and supplies a transparent, auditable design, comprising explicit hypotheses, documented exclusion records, and a candid power analysis, against which Korea’s maturing campaigns and the effects of the 2025 Commercial Act amendments can be re-examined as the phenomenon grows. Whether the activist wave of 2022 ultimately widened corporate governance or merely repriced it will be answered not by the celebration of individual victories, but by whether the value they create is retained by the shareholders in whose name it is won.

Acknowledgements

I sincerely thank Professor Jungsuk Han, Professor Joon Mahn Lee, and Dr. Tae Jun Yoon for taking the time to review an earlier version of this manuscript and for providing thoughtful and constructive feedback. Any errors that remain are my own.

Appendix

The campaign-level dataset for this study was constructed by the author from public sources. Campaign identification, announcement dates, and milestone dates rest on DART regulatory filings; the receipt number of the filing used to date each announcement is reported in Table A1, and every filing can be retrieved from the Financial Supervisory Service’s DART system at https://dart.fss.or.kr. Contemporaneous Korean financial media were used to verify campaign content and governance outcomes, and are cited in the reference list. Daily adjusted closing prices for the twenty-one sample firms and for the KOSPI (^KS11) and KOSDAQ (^KQ11) indices were retrieved from Yahoo Finance. Controlling-shareholder ownership was taken from DART periodic reports and foreign ownership from Naver Finance, as described in Section 4.4.

The Appendix records the basis for each campaign in the analysis sample and for each campaign excluded during screening, so that the construction described in Section 4.1 can be audited case by case. Table A1 presents the identification and dating of the twenty-one analysis events; Table A2 presents the governance outcome recorded at each milestone together with the outcome and eligibility codes used in the tests; Table A3 records the concurrent-events screen described in Section 6.1; and Table A4 and Table A5 summarizes the exclusion record referred to in Section 4.1.

Table A1. Identification and primary sources for the twenty-one analysis events.

#	Activist fund	Target firm	Ticker	Exchange	DART corp code	Announced	DART receipt number
1	Truston Asset Management	Taekwang Industrial	003240	KOSPI	00153393	2022-02-09	20220209000348
2	Cha Partners	Tovis	051360	KOSDAQ	00390365	2022-02-24	20220224004323
3	Align Partners	SM Entertainment	041510	KOSDAQ	00260930	2022-02-28	20220228004871
4	Cha Partners	Sajo Oyang	006090	KOSPI	00141875	2022-03-03	20220303000133
5	Anda Asset Management	SK Chemicals	285130	KOSPI	01267170	2022-03-15	20220314001232
6	Align Partners	JB Financial Group	175330	KOSPI	00980122	2022-05-20	20220520000635
7	Truston Asset Management	Korea Alcohol Industrial	017890	KOSDAQ	00158565	2022-09-21	20220921000248
8	Truston Asset Management	BYC	001460	KOSPI	00122579	2022-12-20	20221220000006
9	Truston Asset Management	LF Corp	093050	KOSPI	00593032	2022-12-22	20221222000494
10	Cha Partners	Namyang Dairy	003920	KOSPI	00107598	2023-03-09	20230309000870
11	Anda Asset Management	KT&G	033780	KOSPI	00244455	2023-03-14	20230314000256
12	Five-fund coalition	Samsung C&T	028260	KOSPI	00149655	2024-02-27	20240227006954; 20240227007425; 20240228008282
13	Cha Partners	Kumho Petrochemical	011780	KOSPI	00106368	2024-03-06	20240306000712
14	Dalton Investments	Kolmar Holdings	024720	KOSPI	00203290	2024-11-06	20241106000144
15	Align Partners	Doosan Enerbility	034020	KOSPI	00159616	2024-11-27	20241127000847
16	Align Partners	Coway	021240	KOSPI	00170558	2025-02-14	20250214002689
17	Align Partners	Gabia	079940	KOSDAQ	00506294	2025-03-27	20250327000826
18	Align Partners	STIC Investments	026890	KOSPI	00117498	2025-03-28	20250328002106
19	Align Partners	Dentium	145720	KOSPI	00526599	2025-03-31	20250331002895
20	Align Partners	SoluM	248070	KOSPI	01159233	2025-09-04	20250904000382
21	Align Partners	A+ Asset Advisor	244920	KOSPI	00684802	2025-11-18	20251118000006

Notes. Announcement dates are the earliest public disclosure of activist engagement (Section 4.1). Each receipt number identifies the DART filing used to date the announcement and can be retrieved at https://dart.fss.or.kr. Entry 12 consolidates the three separately filed 5 percent disclosures of the Samsung C&T coalition, by City of London Investment Management, Whitebox Advisors, and Anda Asset Management, into a single firm-event; all three receipt numbers are listed and the earliest filing date is used. For entry 5 the disclosure was filed on 2022-03-14 and the announcement is dated 2022-03-15, the following trading day.

Table A2. Governance milestones, outcomes, and long-horizon eligibility.

#	Target firm	Milestone	Documented governance outcome at the milestone	Outcome	Elig.
1	Taekwang Industrial	2024-03-29	All three fund-nominated directors elected at the FY2023 annual general meeting; the company publicly accepted the shareholder proposal.	Partially successful	Yes
2	Tovis	2026-03-26	Fund-affiliated outside director confirmed at the 27th annual general meeting, following cumulative treasury-share cancellation and an announced spin-off.	Successful	No
3	SM Entertainment	2023-02-03	“SM 3.0” announcement confirming acceptance of the primary demand, the termination of the Like Planning production arrangement.	Successful	Yes
4	Sajo Oyang	2022-03-24	Minority-nominated audit-committee member elected under the 3 percent rule; the other four proposals were defeated.	Partially successful	Yes
5	SK Chemicals	2022-09-28	Board resolved to cancel 389,489 common shares; the in-kind dividend of the SK Bioscience stake was refused.	Partially successful	Yes
6	JB Financial Group	2024-03-28	Two shareholder-backed outside directors elected at the 11th annual general meeting against the board’s recommendation.	Successful	Yes
7	Korea Alcohol Industrial	2023-03-28	Fund nominee elected audit-committee member and outside director at the 39th annual general meeting.	Successful	Yes
8	BYC	2023-03-24	All four shareholder proposals defeated by the controlling family at the annual general meeting.	Partially successful	Yes
9	LF Corp	2024-03-20	Board resolved a treasury-share buyback of up to KRW 15 billion a year for three years. Attribution to the fund is inferred from disclosed holdings rather than a public campaign.	Partially successful	Yes
10	Namyang Dairy	2023-03-31	Minority-nominated audit-committee member elected under the 3 percent rule; dividend, par-value split and treasury proposals defeated.	Partially successful	Yes
11	KT&G	2023-03-28	Board slate and the board’s KRW 5,000 dividend carried; the activist dividend proposal drew roughly 1.5 percent and the ginseng spin-off was not put through.	Unsuccessful	Yes
12	Samsung C&T	2024-03-15	Coalition dividend proposal drew roughly 23 percent against the board’s 77 percent; the buyback proposal was rejected.	Unsuccessful	Yes
13	Kumho Petrochemical	2024-03-22	All three shareholder proposals defeated at the 47th annual general meeting; the company separately announced its own policy to cancel half of its treasury shares.	Partially successful	Yes
14	Kolmar Holdings	2025-03-31	Fund director-appointment proposal carried at the 35th annual general meeting and the nominee joined the board.	Successful	Yes
15	Doosan Enerbility	2024-12-10	Split-merger transferring the Doosan Bobcat stake withdrawn and the extraordinary general meeting of 12 December cancelled.	Successful	Yes
16	Coway	2025-05-08	Treasury-share buyback trust resolved, followed by a 40 percent payout policy and an increase in the outside-director ratio.	Partially successful	Marginal
17	Gabia	2026-03-26	Both fund director nominees elected at the 27th annual general meeting and the compensation-disclosure proposal carried.	Successful	No
18	STIC Investments	2026-01-20	Corporate value-up disclosure committing to cancellation of the treasury position and a management succession plan.	Successful	No
19	Dentium	2026-03-31	Treasury stake of 22.09 percent cancelled in full; the fund’s audit-committee nominee narrowly lost at the annual general meeting.	Partially successful	No
20	SoluM	2026-03-16	Negotiated governance settlement: proposals withdrawn in exchange for a majority-independent board and independent committees.	Successful	No
21	A+ Asset Advisor	—	Articles amendment to expand separately elected audit-committee members filed for the 2026 annual general meeting; the vote was not confirmed at the time of analysis.	Unresolved	No

Notes. Milestone dates are the date on which the most significant governance outcome was confirmed (Section 4.1). Outcome coding follows Section 4.2: nine campaigns are coded Successful, nine Partially successful, two Unsuccessful, and one Unresolved. Elig. reports the long-horizon eligibility flag of Section 4.3: fourteen Yes, one Marginal, six No. The twelve Yes-flagged campaigns coded Successful or Partially successful form the core long-horizon sample; the Marginal case is added in the robustness specification; the fifteen Yes and Marginal events form the estimation sample for H3. Entry 9 is retained on the evidence of disclosed holdings rather than a public campaign, and the attribution of the buyback resolution to the fund is inferred rather than confirmed. Entry 21 had no confirmed annual-meeting outcome at the time of analysis and is coded Unresolved, which places it outside all long-horizon tests.

Table A3. Concurrent-events screen.

#	Target firm	Window	Concurrent firm-specific event identified
1	Taekwang Industrial	—	None identified. The proxy contest and public exchanges are the campaign itself rather than a separate event.
2	Tovis	—	None identified. The 2025-2026 spin-off falls well outside the annual horizon measured from the 2022 announcement.
3	SM Entertainment	[0, +250]	Control contest. HYBE purchased the founder’s stake and launched a tender offer, Kakao responded with a competing offer, and a court enjoined the share and convertible-bond issuance to Kakao. The entire contest falls inside the annual horizon.
4	Sajo Oyang	—	None identified inside the windows. Six group affiliates subsequently acquired roughly 3 percent each to neutralize the 3 percent rule, a response that developed over the following three years.
5	SK Chemicals	—	None identified inside the windows. The 2021 physical spin-off and listing of SK Bioscience precedes the announcement and is the premise of the campaign rather than a concurrent event.
6	JB Financial Group	—	None identified.
7	Korea Alcohol Industrial	[0, +250]	Subsidiary listing. The KOSDAQ listing of Purit and the resulting differential dividend to non-controlling shareholders, disclosed 2023-08-18, fall inside the annual horizon.
8	BYC	—	None identified. The related-party allegations are the substance of the campaign rather than a separate event.
9	LF Corp	—	None identified. Attribution of the buyback resolution to the fund is itself inferred rather than confirmed (Table A2).
10	Namyang Dairy	Both	Contested control transfer. The disputed sale of the controlling family’s 53.08 percent stake to Hahn & Co was live at the announcement and was resolved by the Supreme Court in January 2024, inside the annual horizon.
11	KT&G	Both	Second activist and a company return program. Flashlight Capital Partners campaigned at the same annual general meeting with its own competing proposals, and the company announced a three-year shareholder-return plan of roughly KRW 900 billion during 2023.
12	Samsung C & T	—	None separate from the event itself. The announcement consolidates three coalition filings made on 2024-02-27 and 2024-02-28 (Table A1).
13	Kumho Petrochemical	Both	Family control dispute. The fund acted as proxy for the individual largest shareholder, a member of the controlling family in a long-running dispute with the chairman, so the campaign is not separable from that contest.
14	Kolmar Holdings	Both	Founding-family control dispute and associated hostile-acquisition tensions ran alongside the campaign.
15	Doosan Enerbility	Both	Group restructuring. The split-merger transferring the Doosan Bobcat stake, announced 2024-07-21, was opposed by proxy advisers, the National Pension Service and the financial regulator, and was withdrawn on 2024-12-10.
16	Coway	—	None identified. The decline in payout ratio after the 2020 change of controlling shareholder is background to the campaign rather than a concurrent event.
17	Gabia	[0, +250]	Activist’s own tender offer. The fund ran a tender offer from 2025-11-25 to 2025-12-14, formed a voting coalition raising its holding to about 14.3 percent, and obtained a court injunction in February 2026, all inside the annual horizon.
18	STIC Investments	[0, +250]	Parallel activist holders and a countervailing company action. A bloc of roughly 26.2 percent formed around the fund together with two other holders, and the company resolved a treasury disposal on 2025-08-22 that ran counter to the cancellation demand.
19	Dentium	[0, +250]	Retirement of a treasury block. The company cancelled its treasury position of about 22.09 percent of shares in two tranches in February and March 2026, which mechanically changes the share count.
20	SoluM	—	None identified inside the windows. The KRW 120 billion third-party-allotment preferred issue of June 2025 precedes the announcement and is the flashpoint of the campaign. The long-horizon window is truncated at 185 trading days by the end of the price history.
21	A+ Asset Advisor	[−5, +5]	Activist’s own tender offer, launched at the announcement and completed in December 2025, with the offer price raised from KRW 8,000 to KRW 9,000. The long-horizon window is truncated at 138 trading days by the end of the price history.

Notes. Each announcement and long-horizon window was screened against contemporaneous Korean financial media and DART filings for firm-specific events other than the activist campaign itself. Window reports where an identified event falls relative to the announcement date: the announcement window [−5, +5], the annual horizon [0, +250], or both. A concurrent event was identified for eleven of the twenty-one firm-events. The estimates are therefore best read as market reactions associated with activist engagement rather than as effects caused by it alone, and the long-horizon results in particular are not attributable to the campaign in isolation for the firms flagged here. Events that precede the announcement, such as the SK Bioscience spin-off and the SoluM preferred issue, are recorded because they motivate the campaign, but they cannot contaminate a window that begins after them. Two firms have long-horizon windows truncated by the end of the available price history, as noted in the table; both are coded No under the eligibility rule of Section 4.3 and therefore enter no long-horizon test.

Table A4. Candidate campaigns excluded during verification.

#	Activist fund	Candidate firm	Ticker	Exchange	First filing	Exclusion ground
1	KCGI	Hanjin KAL	180640	KOSPI	2022-03-03	Outside sample window
2	KCGI	LIG Acuver	073490	KOSDAQ	2022-04-01	Private equity or buyout
3	Truston Asset Management	Gamsung Corporation	036620	KOSDAQ	2022-05-10	No verifiable public campaign
4	KCGI	LIG Nex1	079550	KOSPI	2022-10-12	Private equity or buyout
5	Truston Asset Management	Asems	136410	KOSDAQ	2023-01-30	No verifiable public campaign
6	Truston Asset Management	Cosmecca Korea	241710	KOSDAQ	2023-07-13	No verifiable public campaign
7	Truston Asset Management	C & C International	352480	KOSDAQ	2023-11-16	No verifiable public campaign
8	KCGI	Nextin	348210	KOSDAQ	2024-06-21	Private equity or buyout
9	Truston Asset Management	STX Engine	077970	KOSPI	2024-08-22	No verifiable public campaign
10	KCGI	Hanyang Securities	001750	KOSPI	2024-09-19	Private equity or buyout
11	KCGI	Wins	136540	KOSDAQ	2024-10-07	Private equity or buyout
12	KCGI	Sandoll	419120	KOSDAQ	2025-08-29	Private equity or buyout
13	Truston Asset Management	SMCG	460870	KOSDAQ	2025-08-29	No verifiable public campaign
14	KCGI	BitPlanet	049470	KOSDAQ	2025-09-12	Private equity or buyout

Notes. Initial screening produced thirty-seven candidate campaigns. Fourteen were excluded on the grounds recorded here: seven as private-equity or buyout transactions, six because no public campaign could be verified, and one because the campaign began before the sample window. The three Samsung C&T coalition filings were then consolidated into a single firm-event, yielding the twenty-one-event analysis sample. Table A5 gives the documented basis for each exclusion.

Table A5. Documented basis for each exclusion.

#	Candidate firm	Basis for exclusion
1	Hanjin KAL	Campaign initiated in 2018, before the sample window; the 2022 filing is a trailing disclosure. The bulk of the stake was sold on 2022-03-28.
2	LIG Acuver	Part of the friendly KCGI-LIG new-growth private-equity partnership built on a 2021 exchangeable bond. No demand, opposition, or shareholder proposal on record.
3	Gamsung Corporation	Sub-5 percent position with no verifiable shareholder letter, stated demand, or annual-meeting proposal.
4	LIG Nex1	The fund subscribed the entire KRW 100 billion exchangeable bond issued by the LIG holding company in 2021, a cooperative financing arrangement rather than a governance challenge.
5	Asems	Engagement position with no documented demand, letter, or proposal; attribution of the company’s treasury action to the fund is unconfirmed.
6	Cosmecca Korea	Disclosed holding above 5 percent with no verifiable public demand and no governance outcome.
7	C&C International	Sub-5 percent position with no verifiable public demand and no governance outcome.
8	Nextin	Agreed friendly control acquisition with a profit-sharing arrangement alongside the incumbent chief executive; the share-transfer contract collapsed on 2024-08-25.
9	STX Engine	Disclosed holding with no public demand or proposal; the position is consistent with an event-driven bet on the controlling shareholder’s announced exit.
10	Hanyang Securities	The fund was selected as preferred bidder and acquired a controlling 29.59 percent stake. It is the acquirer, not an outside challenger to management.
11	Wins	The entire 15.64 percent holding was sold back to the controlling shareholder within weeks of disclosure; no sustained campaign was conducted.
12	Sandoll	Cooperative purchase of 19.2 percent from the largest shareholder as a succession and inheritance-funding solution following the founder’s death.
13	SMCG	Position of 5.29 percent reduced to 2.31 percent within roughly two months; no public demand and no governance outcome.
14	BitPlanet	Principal investment carrying a share-pledge financing arrangement; no demand, proposal, or governance outcome on record.

Notes. The concentration of private-equity and buyout exclusions in a single fund reflects that fund’s documented 2025 pivot away from public activism toward principal investment and acquisition, discussed in Section 4.1. The concentration of unverifiable campaigns in a second fund reflects that fund’s practice of building sub-5 percent positions without a public demand, a pattern consistent with quiet accumulation rather than activism.

Conflicts of Interest

The author declares no conflicts of interest regarding the publication of this paper.

References

[1]	Align Partners Capital Management (2023). 에스엠 주식대여 및 얼라인홀딩스 보유 지분 매각 보도 관련 얼라인파트너스 입장 [Align Partners’ Position on Media Reports Regarding SM Share Lending and the Align Holdings Stake Sale] [Press Release]. https://alignpartnerscap.com/insight/press_view.html?no=35
[2]	Bebchuk, L. A., Brav, A., & Jiang, W. (2015). The Long-Term Effects of Hedge Fund Activism. Columbia Law Review, 115, 1085-1156.
[3]	Becht, M., Franks, J., Grant, J., & Wagner, H. F. (2017). Returns to Hedge Fund Activism: An International Study. The Review of Financial Studies, 30, 2933-2971. [Google Scholar] [CrossRef]
[4]	Brav, A., Jiang, W., Partnoy, F., & Thomas, R. (2008). Hedge Fund Activism, Corporate Governance, and Firm Performance. The Journal of Finance, 63, 1729-1775. [Google Scholar] [CrossRef]
[5]	Cha, J. H. (2022). K-Pop Giant SM up against Align Partner’s Shareholder Activism. The Korea Economic Daily Global Edition. https://www.kedglobal.com/corporate-governance/newsView/ked202210040013
[6]	Cha, J. H. (2023a). Korean Activist Fund Mired in Controversy over SM Stock Deals. The Korea Economic Daily Global Edition. https://www.kedglobal.com/shareholder-activism/newsView/ked202305020023
[7]	Cha, J. H. (2023b). Align Partners’ Shareholder Activism? Think Twice. The Korea Economic Daily Global Edition. https://www.kedglobal.com/the-ked-view/newsView/ked202305110012
[8]	Cha, J. H., & Ha, J. E. (2023). South Korean Companies Rattled by Activist Shareholders. The Korea Economic Daily Global Edition. https://www.kedglobal.com/shareholder-activism/newsView/ked202302060023
[9]	Choi, W. Y., & Cho, S. H. (2003). Shareholder Activism in Korea: An Analysis of PSPD’s Activities. Pacific-Basin Finance Journal, 11, 349-363. [Google Scholar] [CrossRef]
[10]	Chung, J. H. (2023). Korea’s Shareholder Activism—A Game-Changing Transformation Since 2022. ECGI Blog. https://www.ecgi.global/publications/blog/koreas-shareholder-activism-a-game-changing-transformation-since-2022
[11]	Claessens, S., Djankov, S., & Lang, L. H. P. (2000). The Separation of Ownership and Control in East Asian Corporations. Journal of Financial Economics, 58, 81-112. [Google Scholar] [CrossRef]
[12]	CNBC (2023a). K-Pop Stocks Soar after Agency behind BTS Announced It Would Become Its Rival’s Top Shareholder. https://www.cnbc.com/2023/02/10/k-pop-stocks-hybe-sm-entertainment.html
[13]	CNBC (2023b). K-Pop Giant Hybe Drops Bid for SM Entertainment, Ending Takeover Battle. https://www.cnbc.com/2023/03/13/k-pop-giant-hybe-drops-bid-for-sm-entertainment-ending-takeover-battle.html
[14]	de Haan, E., Larcker, D., & McClure, C. (2019). Long-Term Economic Consequences of Hedge Fund Activist Interventions. Review of Accounting Studies, 24, 536-569. [Google Scholar] [CrossRef]
[15]	Denes, M. R., Karpoff, J. M., & McWilliams, V. B. (2017). Thirty Years of Shareholder Activism: A Survey of Empirical Research. Journal of Corporate Finance, 44, 405-424. [Google Scholar] [CrossRef]
[16]	Diligent Market Intelligence (2024). Shareholder Activism Annual Review 2024. Diligent Corporation. https://www.diligent.com/resources/research/shareholder-activism-2024
[17]	Ducret, R., & Isakov, D. (2020). The Korea Discount and Chaebols. Pacific-Basin Finance Journal, 63, Article 101396. [Google Scholar] [CrossRef]
[18]	Fama, E. F., Fisher, L., Jensen, M. C., & Roll, R. (1969). The Adjustment of Stock Prices to New Information. International Economic Review, 10, 1-21. [Google Scholar] [CrossRef]
[19]	Federation of Korean Industries. (2024). 주주행동주의 부상과 과제 [The Rise of Shareholder Activism and Its Challenges]. https://www.fki.or.kr/kor/news/statement_detail.do?bbs_id=00035476&category=ST
[20]	Greenwood, R., & Schor, M. (2009). Investor Activism and Takeovers. Journal of Financial Economics, 92, 362-375. [Google Scholar] [CrossRef]
[21]	Hong, J. (2023). 최근 행동주의펀드 현황 및 국내 시사점 [Recent Developments in Activist Funds and Domestic Implications]. 자본시장포커스 [Capital Market Focus]. Korea Capital Market Institute. https://www.kcmi.re.kr/publications/pub_detail_view?syear=2023&zcd=002001016&zno=1717&cno=6102
[22]	Kamalnath, A. (2025). Finfluencers and Other Tech Disruptions to Corporate Law—Insights from South Korea and India. The Chinese Journal of Comparative Law, 13, cxaf002. [Google Scholar] [CrossRef]
[23]	Kang, S. Y. (2022). (Some) Problems of Shareholder Activism and ESG: The Case of the National Pension Service in Korea. SSRN Electronic Journal. [Google Scholar] [CrossRef]
[24]	Kang, S. Y., & Chun, K.-H. (2022). Korea’s Stewardship Code and the Rise of Shareholder Activism. In D. Katelouzou, & D. W. Puchniak (Eds.), Global Shareholder Stewardship (pp. 239-260). Cambridge University Press.
[25]	Kang, S., Kim, S., & Lee, Y. (2025). Determinants of Shareholder Proposals in Korean Listed Companies. Korean Journal of Financial Studies, 54, 171-209. [Google Scholar] [CrossRef]
[26]	Kim & Chang (2018). National Pension Service Adopts the Stewardship Code [Legal Update]. https://www.kimchang.com/en/insights/detail.kc?sch_section=4&idx=18871
[27]	Kim & Chang (2025). Promulgation of September 9, 2025 Amendment to Korean Commercial Code and Additional Legal Updates [Legal Update]. https://www.kimchang.com/en/insights/detail.kc?sch_section=4&idx=33096
[28]	Kim & Chang (2026). Third Amendment to Korean Commercial Code on Mandatory Treasury Share Cancellation [Legal Update]. The Legal 500. https://www.legal500.com/developments/thought-leadership/third-amendment-to-korean-commercial-code-on-mandatory-treasury-share-cancellation/
[29]	Kim, J. S., & Kang, S. H. (2023). 코리아 디스카운트 원인 분석 [An Analysis of the Causes of the Korea Discount] (Issue Report 23-05). Korea Capital Market Institute. https://www.kcmi.re.kr/report/report_view?report_no=1522
[30]	Kim, M. K., & Lee, S. H. (2025). 주식시장 할인율 국제 비교와 코리아 프리미엄 과제 [An International Comparison of Stock Market Discounts and the Korea Premium Agenda] (Issue Report 25-14). Korea Capital Market Institute. https://www.kcmi.re.kr/report/report_view?report_no=2170
[31]	Kim, S. M., Park, K. S., & Jung, C. S. (2020). The Effects of Stewardship Code on the Exercise of Voting Rights by Institutional Investors at Shareholders’ Meetings. Korean Journal of Financial Studies, 49, 515-543. [Google Scholar] [CrossRef]
[32]	Kim, W., Kim, W., & Kwon, K. S. (2009). Value of Outside Blockholder Activism: Evidence from the Switchers. Journal of Corporate Finance, 15, 505-522. [Google Scholar] [CrossRef]
[33]	Kim, Y. H. (2023). 2023년 정기주총 리뷰, 3대 키워드는…주주제안·소유분산·E&S [2023 Annual General Meeting Review: Shareholder Proposals, Dispersed Ownership, and E&S as the Three Keywords]. 녹색경제신문 [Green Economy Newspaper]. https://www.greened.kr/news/articleView.html?idxno=302600
[34]	Klein, A., & Zur, E. (2009). Entrepreneurial Shareholder Activism: Hedge Funds and Other Private Investors. The Journal of Finance, 64, 187-229. [Google Scholar] [CrossRef]
[35]	Ko, K. R., & Kim, W. (2020). The Effect of Pension Fund Activism on the Stock Market and the Role of Media: Evidence from Korea. Korean Journal of Financial Studies, 49, 489-513. [Google Scholar] [CrossRef]
[36]	Korea Securities Depository (2023). 2022년 12월결산 상장법인 주식 소유자 현황 [Stock Ownership Status of December Fiscal-Year-End Listed Companies, 2022] [Press Release]. https://www.ksd.or.kr/ko/about-ksd/ksd-news/press-release
[37]	Krishnan, C. N. V., Partnoy, F., & Thomas, R. S. (2016). The Second Wave of Hedge Fund Activism: The Importance of Reputation, Clout, and Expertise. Journal of Corporate Finance, 40, 296-314. [Google Scholar] [CrossRef]
[38]	La Porta, R., Lopez‐de‐Silanes, F., & Shleifer, A. (1999). Corporate Ownership around the World. The Journal of Finance, 54, 471-517. [Google Scholar] [CrossRef]
[39]	Lee, S. C. (2020). 지난해 12월결산 상장법인 주식 소유자 619만명…전년비 10.3%↑ [Stock Owners of December Fiscal-Year-End Listed Companies Reach 6.19 Million Last Year, up 10.3% Year-on-Year]. E-Today. https://www.etoday.co.kr/news/view/1867842
[40]	Lee, S. H. (2023a). 행동주의 펀드와 주주의 비례적 이익—에스엠 사례(서울동부지법 2023.3.3. 선고 2023카합10034 결정)를 소재로 [Activist Funds and the Proportionate Interests of Shareholders: Based on the SM Entertainment case (Seoul Eastern District Court, Decision 2023 Kahap10034, March 3, 2023)]. Korean Journal of Securities Law, 24, 1-46.
[41]	Lee, S.-M. (2023b). SM, 지배구조 개선되면 30만원 간다더니…11만원에 판 얼라인 대표 [Said SM Would Reach 300,000 Won If Governance Improved, Then Sold at 110,000: The Align CEO]. Money Today. https://www.mt.co.kr/stock/2023/05/02/2023050217251641115
[42]	Lee, Y., & Kim, S. (2023). Ex-Goldman Banker’s Activist Fund Sparks Battle over K-Pop Giant. Bloomberg. https://finance.yahoo.com/news/sm-activist-shareholder-pushes-hybe-235327288.html
[43]	Lin, S. (2025). Korea Moves Forward on Governance Reform. Asian Corporate Governance Association. https://www.acga-asia.org/blog-detail.php?id=110
[44]	MacKinlay, A. C. (1997). Event Studies in Economics and Finance. Journal of Economic Literature, 35, 13-39.
[45]	Na, H. J., & Jung, T. T. (2025). Passing the Shareholders’ Tax Burden to the Corporation: Evidence from Inheritance and Gift Tax. Corporate Governance: An International Review, 33, 1567-1597. [Google Scholar] [CrossRef]
[46]	Nam, K. D. (2022). [Graphic News] Number of Stock Investors Jumps over 50% in 2021 on Market Boom. The Korea Herald. https://www.koreaherald.com/article/2829234
[47]	Park, S. H., & Seo, J. W. (2023). 7만원→16만원 ‘쑥’ 에스엠 주가…2월 16·28일, 무슨 일 있었나 [SM Shares Shot from 70,000 to 160,000 Won: What Happened on February 16 and 28?]. Money Today. https://www.mt.co.kr/stock/2023/10/20/2023102009015928420
[48]	Peoples, G. (2023). Kakao Launches Tender Offer to Buy 35% of SM Entertainment. Billboard. https://www.billboard.com/pro/kakao-tender-offer-sm-entertainment/
[49]	Sherratt, R. (2024). In-Depth: Q&A with Bside Korea. Diligent Market Intelligence. https://www.diligent.com/resources/blog/in-depth-q-and-a-with-bside-korea
[50]	Wang, C. C. Y., & Chan, B. (2025). Align Partners and SM Entertainment: Korean Shareholder Activism Meets K-Pop (A) (HBS Case No. 125-065). Harvard Business School. https://www.hbs.edu/faculty/Pages/item.aspx?num=66883