Invest1 distinct publisher3 min readUpdated
The board set the ratio on book net assets and a December 17 shareholder vote. Holders end up owning two separately priced Korean tech balance sheets instead of one.
The Investor · Invest desk

Compiled by The InvestorSomething wrong?How this is made
Kakao's board resolved on the 21st to divide the company by spin-off into a new entity, Kakao AI, and a surviving entity, Kakao X [1]. What separates this from the usual AI repositioning statement is that the mechanics are already numbered: a split ratio of 0.36 to Kakao AI and 0.64 to Kakao X, determined on the book value of net assets [2], a shareholder vote on December 17, completion on January 1, 2027, and a target of January 27, 2027 for the relisting of Kakao AI and the changed listing of Kakao X [4].
Because it is a spin-off, no one is bought out. Existing shareholders receive shares in both companies in proportion to what they already hold [3]. Kakao notes the detailed schedule may still change depending on consultations with relevant institutions and procedures [5].
The ratio is the number to sit with. On book net assets, Kakao X is being allocated roughly 1.78 times what Kakao AI gets [17], and Kakao X is the side that inherits the balance sheet: techfin in Kakao Bank, Kakao Pay and Kakao Pay Securities, content in Kakao Entertainment, SM Entertainment and Kakao Piccoma, and mobility in Kakao Mobility [10]. Kakao AI takes KakaoTalk plus operating subsidiaries DKTechin and K&Works, and hands the job of supporting and managing subsidiary growth to Kakao X [8][11]. Book value is an accounting allocation, not a forecast of how the two tickers trade in late January.
Leadership tells you which entity is the continuation and which is the new mandate. Shin Jung-a, the current Kakao CEO, has been designated CEO of Kakao AI [8]. Kim Do-young, CEO of Kakao Investment and head of the group investment strategy office at the Kakao CA Council, takes Kakao X [9], which Kakao describes as a "future value investment company" tasked with growing the existing core groups while finding and funding new ones [16]. That is an investment-company remit with a portfolio of listed and quasi-listed financial assets attached.
According to Kakao, the split is meant to speed decision-making and create capital allocation by business, on the view that a single decision-making system could not execute quickly across units at different growth stages [6]. It follows about two years of winding down non-core businesses and cutting affiliate count [7].
The targets are company plans, not results. Kakao AI is aiming for more than 20 million AI daily active users by 2030 and a rise in platform dwell time of more than 50% [13], against a KakaoTalk base of roughly 50 million users [12] - so about 40% of the current user base converting into daily AI usage [18]. It also projects average annual revenue growth of around 20% through 2030 to more than 6 trillion won, with AI revenue exceeding 1 trillion won in 2030 [14][15]. That implies AI is about 17% of the 2030 revenue target [19], with the first checkpoint being a double-digit AI revenue share in 2028 [15].
Watch three things. Whether the December 17 vote passes without a ratio revision, since minority holders are being handed the split rather than paid for it [3][4]. Whether the 26 days between completion and the January 27 listing dates hold [20][5]. And whether Kakao X, freed of the AI story, discloses capital allocation like an investment company or spends against the "second Kakao Bank" brief [16].
Follow any of these and your For You feed starts watching them — no settings page required.
invest
Korean retail bought 133 billion won of gold after the rebound, not before it1 distinct publisher
invest
TheBornKorea's 37% two-day run prices 80 Canadian hotels; the MOU covers one Markham site1 distinct publisher
invest
Korea gives fractional art and royalties a listing venue, plus a capital floor1 distinct publisher
invest
A 10% premium to buy Korean stock in New York is not diversification1 distinct publisher
Ranked by verification strength, evidence, and original report placement.
Kakao (035720.KS) said it held a board meeting on the 21st and resolved to split the company through a spin-off into a new entity, Kakao AI, which will handle the AI business, and a surviving entity, Kakao X.
The split ratio was determined based on the book value of net assets, at 0.36 for Kakao AI and 0.64 for Kakao X.
Because it is a spin-off, existing shareholders will each be allocated Kakao AI and Kakao X shares according to their holdings.
Kakao plans to complete the split on January 1, 2027, following an extraordinary shareholders' meeting on December 17, and to pursue the relisting of Kakao AI and the changed listing of Kakao X on January 27, 2027.
Kakao said the detailed schedule may change depending on consultations with relevant institutions and related procedures.
According to Kakao, the spin-off aims to accelerate decision-making and establish an efficient capital allocation system by business in response to the AI era, after it determined there were limits to swiftly executing strategies for businesses at different growth stages under a single decision-making system.
Distinct publishers with included, body-backed reporting in this cluster.
en.sedaily.com
2 articles · August 20, 2026
Evidence-backed comparisons of source perspectives and observed adoption signals. Read the methodology
Which Builder, Operator, and Investor concerns the observed source mix emphasized—not a truth score.
Evidence, demonstrated adoption, hype gap, incentives, and confidence are assessed independently, each on its own current evidence. How these are measured.
Firm on mechanics, unverified on the AI thesis
The transaction facts are documented at disclosure quality and repeated consistently across both items: the 0.36/0.64 book-net-asset ratio, pro-rata share allocation, the December 17 EGM and January 1 / January 27, 2027 dates, and the two named CEOs. Everything that makes this an AI story — 20 million AI DAU, 6 trillion won of revenue, 1 trillion won of AI revenue — rests solely on Kakao's own forward statements with no baseline, no third-party analysis, and only one publisher in the cluster.
Board-approved, shareholder vote and product uptake still ahead
Adoption evidence is procedural, not usage-based. A board resolution and a scheduled shareholder vote exist, and the 12% share move shows the market has priced a reaction, but the split is not yet effective and no AI product usage is disclosed anywhere in the sources — only the pre-existing ~50 million KakaoTalk messenger base, which is the starting condition rather than evidence of agentic AI uptake.
Solid deal terms wrapped in unbacked 2030 AI numbers
The structural claims are neither overstated nor understated — ratio, dates and leadership are concrete. The overstatement sits in the AI framing: an 'AI Core Company' with a 20 million AI DAU goal and 1 trillion won of AI revenue by 2030, announced with zero current AI usage or revenue baseline, alongside a valuation passage whose SOTP figure does not reconcile with the market capitalization the same article cites. The immediate 11.76% share fall indicates the market discounted the framing rather than accepting it.
Company-authored narrative with a re-rating objective
Nearly all substantive content originates with Kakao, which has a direct interest in the split being read as value-unlocking: the sources carry its stated rationale, its 2030 targets, its 'AI Core Company' and 'future value investment company' labels, its SOTP-versus-market-cap comparison, and a scheduled analyst briefing titled 'Accelerating Growth and Enhancing Shareholder Value Through Governance Restructuring'. The cluster contains no counterparty, regulator, or independent analyst commentary to offset that, and both items come from a single financial-news publisher covering a large listed domestic issuer.
High on the transaction, low on the outcome
Confidence is asymmetric. The deal terms, calendar and leadership assignments are stated precisely and repeated without contradiction, so those can be relied on subject to the company's own schedule caveat. Beyond that, the cluster has one publisher, no independent analysis, a name-romanization discrepancy between the two items, an unreconciled valuation passage, and forward targets running to 2030 that cannot be assessed from what is supplied.