FnGuide Inc. (064850.KQ) - Deep Dive Research Report

Prepared 2026-08-14. Listing venue: KOSDAQ (Korea). Sector: Financial Services / financial data and index provision. Fiscal year-end: December. All figures in Korean won (KRW) unless stated.

A note on reporting cadence and this report's evidence base. FnGuide reports quarterly and its H1 2026 (Q2) report was due on the statutory 45-day deadline of ~14 August 2026 (today). A half-year report filed in the last few days may not yet be captured in search indexes or the bot-walled DART portal, so this report anchors verified detail to the six most recent periods I could confirm: Q1 2026 (filed ~mid-May 2026), FY2025/Q4 2025 (~March 2026), Q3 2025 (~November 2025), Q2/H1 2025 (~August 2025), Q1 2025 (~May 2025) and Q4/FY2024 (~March 2025). Unlike a US or large-cap Korean issuer, FnGuide does not publish earnings-call transcripts; its investor communication runs through DART quarterly filings, its own IR page, a Korea IR Council presentation dated 19 February 2026, and sell-side initiation (Hana Securities, 7 May 2026). Where I would normally quote a concall, I quote these sources and flag them as such.

Section 1: What the company does

FnGuide sells the raw material that Korea's investment industry runs on: cleaned, structured financial data, and the stock-market indices that ETFs are built to track.

Think of it in two halves. The first half is a data-and-research utility. A fund manager, a bank analyst, a pension-fund quant or a university finance department in Korea needs consistent, machine-readable numbers - a company's revenue going back fifteen years, its consensus earnings estimate, every brokerage report ever written on it, the daily holdings of a fund. FnGuide collects, standardises and stores that information and rents access to it through a set of desktop and Excel tools (DataGuide, QuantiWise, WiseReport) and through raw data feeds and APIs that plug straight into a client's own systems. This is the Korean-market analogue of what Bloomberg, FactSet, Refinitiv/LSEG and S&P Capital IQ do globally, but built around the specific depth of Korean-language, Korean-regulatory, Korean-market data that the global vendors cover only thinly.

The second half is the reason the stock re-rated sharply through late 2025 and into 2026: FnGuide is Korea's largest private stock-index provider. When an asset manager launches an ETF - say a dividend ETF, a covered-call income ETF, a shipbuilding-theme ETF or a semiconductor ETF - that fund needs an index to track. FnGuide designs and calculates those indices and licenses them to the ETF issuer, collecting a fee that scales with the assets under management (AUM) sitting in the funds that track them. As Korea's ETF market has exploded, and as issuers have raced to launch ever-more-specialised thematic and income products, FnGuide's index licensing revenue has compounded. By May 2026 the net assets of ETFs tracking FnGuide indices had passed KRW 65 trillion, up from KRW 50 trillion just three months earlier (News1, 2026; Herald, 2026).

The founding arc explains why the company sits where it does. FnGuide was set up in 2000 to provide financial information online. The pivotal strategic decision came in October 2018, when it merged with WISEfn (와이즈에프엔), its longest-standing domestic rival in financial data (iTooza, 2021). That merger effectively consolidated Korea's private financial-data market into a single dominant local vendor, folding two overlapping databases into one and removing the price competition that had capped both firms. FnGuide listed on the KONEX junior board in 2013 and transfer-listed to the main KOSDAQ market in 2020. It is headquartered in the Magok district of Gangseo-gu, Seoul, and is a lean operation of roughly 150 employees (stockanalysis.com).

The value proposition is different for each half. For the data business, it is trust and coverage: a manager cannot run a portfolio, market a fund or file a regulatory report on numbers they do not trust, and FnGuide's Korean-market database is deeper and better standardised than any alternative that a local institution can practically switch to. For the index business, it is a toll on a growing river: FnGuide does not take market risk on the ETFs, it simply earns a licensing fee that grows as the funds it indexes attract assets. The two halves reinforce each other, because the same data infrastructure that powers the terminals is what lets FnGuide construct and calculate indices credibly.

Section 2: Business segments

FnGuide runs three reported businesses plus a small "other" bucket. The revenue mix, as laid out in the May 2026 Hana Securities initiation, is roughly: Financial Information Services ~57%, Index ~30%, Fund Evaluation ~5%, and Other ~7% (Hana Securities, 7 May 2026).

2.1 Financial Information Services (~57% of revenue) - the cash cow

This is the original business and still the largest. It is a suite of subscription products sold to financial institutions and research bodies:

• DataGuide - an Excel-based data lookup and analysis tool. This is the workhorse for Korean buy-side and sell-side analysts: pull any listed company's financials, estimates, ratios, price and index data directly into a spreadsheet. Its stickiness comes from being embedded in analysts' daily modelling workflow.
• WiseReport - an aggregation of research from every domestic brokerage, economic research institute and partnering overseas data provider, classified and searchable in one place (FnGuide IR). The company describes its position in research distribution as dominant. Retamin is the mobile front end for this.
• QuantiWise - a quantitative analysis and back-testing platform for systematic strategies.
• Company Guide / FnGuide.com / FnSpace - the public and semi-public data portals (the comp.fnguide.com pages that show up constantly in Korean equity research are FnGuide's own product surface).
• Data Feed / API / ASP - raw structured data delivered straight into client systems, spanning the whole capital market. This is the highest-value, stickiest form of the product because it becomes plumbing inside the client's own applications.

The core capability here is not any single tool but the underlying database: two decades of Korean-market data, cleaned and standardised, deepened by the WISEfn merger. That corpus is what took years to build and would be very hard to replicate. Strategically management treats this as the stable, high-margin base that funds investment elsewhere. It grows with financial-industry headcount and data budgets - steady, not explosive (Q1 2026 disclosure noted this segment growing in the low single digits while index raced ahead).

2.2 Index business (~30% of revenue) - the growth engine

This is where the story is. FnGuide designs, calculates and licenses stock-market indices, principally as the underlying benchmarks for ETPs (ETFs and ETNs). It is the second-largest index calculator in Korea after the Korea Exchange itself, and the largest private one (Newstomato, 2026).

The economics are AUM-linked licensing: FnGuide earns a fee tied to the assets in funds tracking its indices, so revenue compounds as those ETFs grow. By May 2026, ETFs tracking FnGuide indices held roughly KRW 65 trillion in net assets across 167 products, of which 18 were "mega" ETFs above KRW 1 trillion each; FnGuide indices underlie roughly 10% of the entire Korean ETF market (News1, 2026). The growth has been driven by the three hottest corners of the Korean ETF boom: monthly-income covered-call ETFs, dividend ETFs, and thematic ETFs (semiconductors, defense, shipbuilding). A concrete example: the SOL 조선TOP3플러스 (shipbuilding Top 3 Plus) ETF, which tracks an FnGuide index, passed KRW 2 trillion in net assets in November 2025 (Daum/Newsis, 2025).

Why it exists as its own segment: it is a fundamentally different economic model (AUM royalty vs. seat-based subscription), a different customer (ETF issuers rather than analysts), and different unit economics - once an index is designed, incremental AUM carries almost no marginal cost, which is why 2025 operating profit grew far faster than revenue. This is the segment management points to for the ROE 18% / revenue-CAGR 15% ambition in the value-up plan.

2.3 Fund Evaluation (~5% of revenue)

FnGuide rates and evaluates funds, sells fund data, and provides consulting to institutional investors on fund selection and monitoring. It is small but strategically coherent: it consumes the same underlying data infrastructure and deepens FnGuide's relationships with asset managers, who are also the buyers of the index and data products. It is a cash-generative, low-growth adjacency rather than a growth bet.

2.4 Other (~7%)

This bucket includes Mydata-based customized investment-insight services and assorted consulting/integrated-research-system work. The Mydata angle - a regulated Korean open-banking/data-portability regime - lets FnGuide build personalised investment insight services on top of a user's aggregated financial data, an optionality play rather than a material earner today.

Section 3: Products and business detail

The full product catalogue maps to the segments above:

Terminal / analytical software: DataGuide (Excel add-in data terminal - the flagship analyst tool), QuantiWise (quant/back-testing), WiseReport (aggregated brokerage research distribution), Retamin (mobile research app), Company Guide (the comp.fnguide.com company-data pages ubiquitous in Korean equity research), FnSpace and FnGuide.com (data portals).

Data delivery: Data Feed, API and ASP services delivering structured capital-market data into client systems. This is the enterprise-grade, deeply embedded form of the product.

Index products: Custom and off-the-shelf equity indices for ETPs - covering broad-market, factor/smart-beta, dividend, covered-call income, and thematic (semiconductor, defense, shipbuilding, AI) strategies. FnGuide both designs bespoke indices to an issuer's specification and calculates/maintains them daily.

Fund and adjacency products: Fund ratings and fund data, institutional consulting, integrated research systems, and Mydata-based personalised insight services.

The "manufacturing" process is data engineering, not physical production. FnGuide ingests filings, price and volume data, brokerage research, corporate actions and fund holdings; cleans and standardises them into a consistent schema; stores them in a historical database going back roughly two decades; and serves them out through the tools and feeds above. The hard, defensible part is the standardisation and history - a new entrant can buy today's price feed cheaply but cannot conjure fifteen years of consistently-mapped Korean corporate financials overnight. The index business layers a rules-engine on top: an index methodology, a rebalancing calendar, and a daily calculation-and-publication pipeline that ETF issuers and the exchange rely on being correct and on time.

Geographically the business is overwhelmingly domestic - its edge is Korean-market depth, and its customers are Korean institutions. International exposure is via data-partner relationships (importing global research/data to resell into Korea) rather than exporting.

The milestone that most changed the business was the 2018 WISEfn merger, which removed FnGuide's main data rival and consolidated the local market. The milestone that is changing it now is the ETF-index flywheel: crossing KRW 50 trillion (February 2026) and KRW 65 trillion (May 2026) in tracked AUM in the space of a quarter. Management has also flagged, in its Q1 2026 disclosure, that it is building a finance-specialised large language model (LLM) and running machine-learning-based content expansion and a service renewal on top of its data corpus (Investing.com KR, 2026).

Section 4: Customers

FnGuide's customers are institutions, not retail. They fall into a few types:

• Securities firms (brokerages) - buy DataGuide and QuantiWise seats for research analysts and quants, and consume data feeds for their own platforms. The buying decision sits with research heads and IT/data procurement; the purchase criterion is coverage, accuracy and workflow fit. WiseReport also makes FnGuide the distribution rail for these firms' own research.
• Asset managers - the pivotal customer, because they buy from FnGuide twice: data/terminal products for their analysts, and index licenses for their ETFs. The index decision is made by ETF product teams; the criterion is whether FnGuide can build and reliably calculate an index that will attract AUM. Korea's largest ETF houses (the KODEX/Samsung and TIGER/Mirae Asset complexes dominate ~65%+ of the ETF market between them) are the key relationships (etfnow.co.kr).
• Banks, pension funds and insurers - data feeds and fund-evaluation/consulting.
• Research institutes and academia - DataGuide subscriptions.

Switching costs are real but different by product. For the data terminals, the lock-in is workflow and history: analysts build their models around DataGuide's Excel functions, and a firm's entire back-catalogue of research and screens is standardised to FnGuide's schema. Ripping that out means retraining staff and re-mapping data. For the index business, the switching cost is even higher and more subtle: once an ETF is launched tracking a FnGuide index and has gathered assets, the issuer cannot casually re-benchmark it without a fund reorganisation, investor communication and potential tax/tracking disruption. The installed AUM is effectively annuity revenue.

Concentration cuts two ways. On the data side, revenue is spread across many institutional subscribers - reasonably diversified. On the index side, revenue concentrates in a handful of very large ETF issuers, so the health of the KODEX and TIGER franchises matters a great deal; but because those franchises are themselves the growth engines of the Korean ETF market, the concentration reflects being plugged into the best customers rather than a fragile dependency.

Contract structure is predominantly recurring: annual data subscriptions that renew, and AUM-linked index licenses that pay as long as the ETF exists and holds assets. This gives FnGuide unusually predictable, recurring revenue for a small-cap - closer to a software/royalty model than a transactional one.

Section 5: Competitive landscape

FnGuide competes in two distinct arenas, and its position is very different in each.

In Korean financial data, its main historical rival was WISEfn, which it acquired in 2018 - so by consolidation it removed its closest local competitor and now holds a dominant position in the domestic data and research-distribution niche. The remaining competition is (a) the global vendors - Bloomberg, LSEG/Refinitiv, FactSet, S&P Capital IQ - which are stronger on global data but thinner and more expensive on Korean-specific depth, and (b) KOSCOM, the Korea Exchange's IT subsidiary, which provides market data infrastructure. FnGuide wins domestically on Korean-language coverage, price, and workflow familiarity; it loses to the globals where a client needs deep cross-border or multi-asset data. Barriers to entry in this niche are high because of the two-decade cleaned-history moat and the workflow lock-in, and they were raised further by the WISEfn deal.

In index provision, the structure is a clear #1 and #2: the Korea Exchange (KRX), which owns the flagship KOSPI/KOSDAQ benchmark franchise, sits first; FnGuide is the clear #2 overall and the #1 private index house, at roughly 10% of the ETF market by tracked assets (Newstomato, 2026). Other private index competitors exist (including global index brands MSCI and S&P Dow Jones for globally-oriented Korean ETFs), but for domestic thematic, dividend and covered-call products FnGuide is the go-to private provider. FnGuide wins here on agility - it will build a bespoke thematic index quickly to catch a product trend - and on entrenched issuer relationships. It is exposed if KRX aggressively expands its own custom-index offering, or if global index brands push harder into Korean thematic ETFs.

The market-cap column below is a peer-size reference only, for the two globally-listed comparables; it is not applied to FnGuide.

Market-cap sources: MSCI, FactSet.

The honest read: FnGuide has a genuine local moat in data (consolidated market, deep history, workflow lock-in) and a strong but more contestable position in indices (agile #1 private provider riding an ETF wave that KRX and global brands also want). It is not immune to competition, but in its home niche it is structurally advantaged.

Section 6: Industry

FnGuide's fortunes track two industry currents.

The first is the size and data intensity of Korea's investment industry - brokerages, asset managers, banks, pensions. Demand for financial data grows with financial-sector headcount, product complexity and regulatory reporting burden. This is a mature, GDP-plus grower, not a boom market, but it is stable and recurring.

The second, and the reason FnGuide is interesting right now, is the explosive growth of the Korean ETF market. Korea has become one of the most dynamic retail ETF markets in Asia, with a rapid proliferation of thematic, dividend and especially monthly-income covered-call products. As of early 2026 there were roughly 166 monthly-distribution ETFs listed in Korea, of which about 51 were covered-call strategies (etfnow.co.kr / Kodex). Every one of those funds needs an index to track, and index provision is a direct beneficiary of the launch cadence and the AUM accumulation. The Korean ETF market is dominated by two issuers - Samsung's KODEX and Mirae Asset's TIGER, together over 65% share - which means the pool of index buyers is concentrated but well-capitalised and growth-hungry.

Regulation is a tailwind on two fronts. First, Korea's "Corporate Value-up Programme," launched in 2024 to address the "Korea discount," has pushed listed companies toward higher shareholder returns and better disclosure - which increases demand for the kind of governance, valuation and returns data FnGuide sells, and has been supportive of the whole value-oriented ETF theme. Second, the Mydata regime (regulated financial-data portability) opens a lane for personalised data services.

Cyclicality: the data-subscription base is fairly defensive - institutions do not cancel core data feeds in a downturn. The index business is more market-sensitive, because AUM-linked fees fall if ETF assets shrink in a bear market or if net flows into ETFs stall. So the index growth that has driven the recent re-rating is also the part most exposed to a market correction or a slowdown in ETF launches. The AI/LLM wave is an emerging industry driver - financial institutions want machine-readable, model-ready data and finance-tuned language models, which plays to FnGuide's data corpus.

Section 7: Growth triggers

Sourced from FnGuide's DART quarterly disclosures, its 19 February 2026 Korea IR Council presentation, and the 7 May 2026 Hana Securities initiation. (FnGuide does not hold transcript-published earnings calls; these are the equivalent management-communicated forward statements.)

• Index AUM flywheel continuing - ETF net assets tracking FnGuide indices grew from ~KRW 50tn (February 2026) to ~KRW 65tn (May 2026), across 167 products with 18 above KRW 1tn each. Because index fees are AUM-linked, continued ETF asset growth flows through with minimal added cost (Q4 2025 / FY2025 disclosure, ~March 2026; News1, May 2026). Repeated across multiple periods.

• 2026 segment guidance skewed to index - the February 2026 IR set out 2026 segment revenue guidance with the index business as the primary growth driver; Hana's May 2026 note framed the same, projecting index-led double-digit top-line growth (Hana Securities, 7 May 2026; alphabiz, 2026).

• New thematic ETF index launches - FnGuide indices underpinning fast-growing theme ETFs (shipbuilding, semiconductor, defense, dividend/covered-call). The SOL 조선TOP3플러스 ETF tracking a FnGuide index passed KRW 2tn (November 2025), illustrating the pipeline of new indices converting into AUM (Q3 2025 disclosure period; Daum, Nov 2025).

• Finance-specialised LLM and AI service renewal - management stated in the Q1 2026 disclosure that it is strengthening a finance-specific LLM capability and renewing services using AI/machine-learning on top of its data corpus, positioning the data business for an AI-driven product cycle (Q1 2026 disclosure, ~May 2026; Investing.com KR).

  "현재 AI 기술을 활용한 서비스 리뉴얼 및 금융 특화 LLM 역량 강화를 진행 중" - FnGuide Q1 2026 disclosure ("currently pursuing a service renewal using AI technology and the strengthening of a finance-specialised LLM capability").

• Value-up-driven demand for governance/returns data - Korea's Value-up Programme raises institutional demand for the valuation, dividend and governance datasets FnGuide sells; FnGuide is both a supplier to and a participant in the theme (2026 value-up plan disclosure; DigitalToday, 2026).

• Mydata personalised insight services - the Mydata-based customized investment insight service is an expansion lane on top of the existing data infrastructure (Q1 2026 disclosure).

Section 8: Key risks

Corporate control dispute / governance overhang (high probability, moderate-to-high drag). This is the most company-specific risk. FnGuide has been the subject of an active management-control contest. The Hwacheon Group - through Hwacheon Machinery Works (화천기공, ~7.81%) and Hwacheon Machinery (화천기계, ~4.83%), roughly 12.6% combined - is the largest shareholder bloc, while founder and former CEO Kim Gun-ho (김군호) holds ~11.33% (Newsis, 2024). In 2024 the two sides fought through the courts over an extraordinary general meeting, board seats and, pointedly, the company's treasury shares - Hwacheon reportedly pressed for the company to hand treasury stock to the major shareholder, and the founder-linked management resisted, escalating to litigation (SmartToday, 2024). The mechanism of harm: a control fight distracts management, can trigger value-destructive uses of the balance sheet (treasury shares moved for control rather than shareholder return), and injects uncertainty into leadership continuity. The heavy insider selling by Kim Gun-ho in late 2025/2026 (Section 11) may reflect the evolving state of this contest rather than a business signal.

Index revenue is AUM-sensitive to a market/ETF correction (moderate probability, high drag on the growth narrative). The re-rating rests on ETF assets tracking FnGuide indices compounding. Because fees are AUM-linked, a bear market that shrinks ETF net assets, or a stall in ETF net inflows, would directly compress the fastest-growing, highest-incremental-margin revenue stream. The covered-call/monthly-income ETF category in particular has grown very fast and could see outflows if the yield trade sours.

Customer concentration in two ETF houses. With KODEX (Samsung) and TIGER (Mirae Asset) controlling most of the ETF market, FnGuide's index growth is tied to those franchises continuing to launch and gather assets on FnGuide indices rather than in-house or KRX/MSCI indices. A strategic shift by either issuer toward self-indexing would hurt.

Competitive encroachment on indices. KRX expanding its custom-index business, or global index brands (MSCI, S&P DJI) pushing into Korean thematic ETFs, could cap FnGuide's index share.

AI as double-edged. FnGuide is investing in a finance LLM, but generative AI also lowers the cost for others to structure and serve financial data, potentially eroding the long-run defensibility of the data-subscription moat if a well-funded entrant automates coverage.

Small-cap execution and disclosure depth. As a ~150-person KOSDAQ small cap without earnings-call transcripts, information flow is thinner and key-person dependence is higher than for a large-cap peer.

Section 9: Walk the talk

Reporting periods used, most recent first: Q1 2026 (~May 2026), FY2025/Q4 2025 (~March 2026), Q3 2025 (~November 2025), Q2/H1 2025 (~August 2025), Q1 2025 (~May 2025), Q4/FY2024 (~March 2025). The most recent verified period (Q1 2026) is within ~90 days of today; the H1 2026 report was due ~14 August 2026 and may have just been filed but is not yet captured. FnGuide does not publish earnings-call transcripts, so credibility is assessed against its DART filings, its February 2026 IR presentation and its published value-up commitments rather than against spoken guidance. This limits the granularity of promise-versus-outcome tracking, and I flag that rather than manufacture precision.

The clearest trackable commitment is the shareholder-return promise. In May 2024, alongside its first Corporate Value-up plan, FnGuide committed to a floor of KRW 180 dividend per share and a payout ratio floor of 26%, with a stated intention to step returns up over time (edaily). Against that promise, the company delivered: DPS rose from KRW 180 (FY2023) to KRW 220 (FY2024) to KRW 260 (FY2025) - three consecutive increases and each year at or above the floor it set. On this specific, datable commitment, management did what it said.

On growth, the through-line across the FY2024, H1 2025 and Q3 2025 disclosures was that the index business would carry the group. That materialised: FY2025 delivered index-driven acceleration, with operating profit growing far faster than revenue and net income up sharply year-on-year, consistent with the operating leverage management had pointed to (FY2025 disclosure, ~March 2026; alphabiz, 2026). The February 2026 IR then guided to further index-led growth for 2026, and the interim data points (tracked AUM rising from ~KRW 50tn to ~KRW 65tn between February and May 2026) validated the direction within the same year. Management's narrative and the observable AUM trajectory have moved together, which is a mark of accuracy rather than spin.

In its 2026 value-up update, management set medium-term targets of ROE ~18%, revenue CAGR ~15%, and a payout ratio floor of 25% (high-dividend status maintained) (DigitalToday, 2026). These are forward targets not yet testable, but the dividend floor is a continuation of a promise already kept, which lends them credibility.

The area where "walk the talk" cannot be cleanly assessed is governance. The board and control situation has been contested in court, and treasury-share policy has been a point of conflict between the largest shareholder and management (SmartToday, 2024). That is not a broken operating promise, but it is a reason to hold the credibility verdict slightly guarded.

Net assessment: on the operating and capital-return promises that can be tracked - rising dividends against a self-set floor, and index-led growth - management has been consistently accurate and has delivered. It reads as a management that does roughly what it says on the business, with the important caveat that the shareholder-structure dispute is an unresolved overhang that sits outside the operating track record.

Section 10: Shareholder friendliness index

Dividends. FnGuide has paid a steadily rising ordinary dividend: KRW 180 per share for FY2023, KRW 220 for FY2024, and KRW 260 for FY2025 - three straight annual increases (edaily). The reported payout ratio fell from ~61.2% (FY2024) to ~28.6% (FY2025), which is not a cut in generosity but the arithmetic of a much larger 2025 net income (up sharply on the index business): the per-share dividend rose while earnings rose faster. The dividend policy is anchored by an explicit value-up commitment - a floor of KRW 180 DPS and a 26% payout floor set in 2024, raised in ambition in the 2026 plan (payout ≥25% with "high-dividend company" status maintained, alongside ROE 18% and revenue-CAGR 15% targets) (DigitalToday, 2026).

Buybacks and dilution. In the ~90-day window MoatMap covers (since 16 May 2026), there were zero recorded share buybacks by the company. Searching the longer three-year record: FnGuide holds treasury shares (자사주), and those treasury shares have been the central object of the 2024 control dispute - the largest shareholder pressed for them to be transferred, management resisted (SmartToday, 2024). I did not find evidence of a large recent open-market buyback-and-cancellation programme in the last three years; the company's capital return over this period has run through the rising cash dividend rather than through buybacks, and the treasury stock has functioned more as a governance flashpoint than as a shareholder-return tool. Shares outstanding are approximately 11.41 million (stockanalysis.com); I found no indication of material option-driven dilution over the period, so the share count is best described as broadly stable. (Note: I could not fully verify a formal three-year buyback/cancellation history from the bot-walled DART filings within the search budget; the dividend record is well-sourced, the buyback record is stated as "not found / capital return via dividends," not as a confirmed zero-programme.)

Verdict: Returns Capital (via dividends) - a company that pays a rising, policy-backed dividend and has bound itself to a value-up floor, but whose treasury shares are entangled in a control dispute rather than being deployed for shareholder returns.

Section 11: Insider activities

Source: MoatMap's cross-market disclosure database (market KR, DART-sourced), current to 2026-08-13 23:01 UTC. Korea's DART portal is API-gated and returns blocked stubs to web search, so this block is the canonical source for recent insider dealing; I have not substituted aggregators. Note that DART discloses share counts and direction reliably but the KRW prices shown are as reported/estimated.

Recent transactions (most recent first):

(All: DART "임원·주요주주 특정증권등 소유상황보고서" / 5% Rule reports, dates as above.)

Buys - the signal. The most consistent buyer is Jeon Min-seok (전민석), who bought open-market shares on at least six occasions from January through August 2026 and was promoted from Vice President to CEO (대표이사) during that window. His 5 August 2026 purchase was his first as CEO. These are modest absolute sums (each roughly one to a few months' worth of a senior executive's pay) but the pattern is what matters: a rising executive buying repeatedly through a period of extreme share-price volatility (the stock ran from ~KRW 8,500 in late 2025 to ~KRW 29,000 by May 2026 and back to ~KRW 17,600 by August 2026) - including buying on the way up and again after the pullback - is a genuine conviction signal from the person now running the company. Box Holdings LP, a 5%+ holder, was also a steady open-market accumulator through late 2025 (three buys, Aug-Nov 2025) before reversing to sales in April-May 2026.

Sells - the why. The large sells cluster in two names. Kim Gun-ho (김군호), the founder and former CEO who holds ~11.33%, sold ~58.6k shares in December 2025 and ~151k shares in May 2026. Given his role as one pole of the 2024 control dispute, the most plausible reading is stake management tied to the governance situation and monetisation into the sharp 2026 price strength, rather than a view on operations; the specific reason is not disclosed in a filing footnote, so this is inference from context. Box Holdings LP's April-May 2026 sales (~315k shares combined) came directly into the price spike toward ~KRW 29,000 and read as profit-taking on the position it had accumulated cheaply in late 2025 - reason not disclosed, but the round-trip timing is consistent with an opportunistic financial holder. Hwacheon Machinery's January 2026 trim was tiny (6,170 shares) and immaterial to its ~12.6% bloc.

Net assessment. The activity is genuinely mixed and concentrated in a few names. On the buy side, the signal is clean and encouraging: the incoming CEO has been a persistent open-market buyer, including after a large drawdown - the clearest possible insider vote of confidence in the go-forward business. On the sell side, the large sales come from a founder-turned-financial-holder entangled in a control fight and from an opportunistic 5% holder taking profits into a spike - both more about the shareholder register and the stock's volatility than about deteriorating fundamentals. Netting it out: mildly bullish on the operating read (CEO buying), but with a governance-driven overhang as founder-linked and financial holders reduce into strength. The register is in flux, and the insider tape is best interpreted alongside the unresolved control contest rather than as a simple net-buyer/net-seller tally.

Section 12: Scenarios

Bull case. Korea's ETF market keeps compounding, and the covered-call, dividend and thematic categories that FnGuide indexes stay in favour. Tracked AUM pushes well beyond KRW 65 trillion, and because index fees are AUM-linked with almost no incremental cost, operating profit keeps outrunning revenue - the operating leverage that showed up in 2025 becomes the through-cycle norm. FnGuide converts its data corpus into a credible finance-specialised LLM and an AI-refreshed product suite, giving the mature data segment a new growth vector and defending it against commoditisation. The control dispute resolves cleanly - a stable ownership settlement removes the overhang, treasury shares get cancelled or used for genuine shareholder return, and the value-up commitments (rising dividends, ROE 18%, 15% revenue CAGR) are met. The incoming CEO who has been buying stock is proven right. FnGuide ends up recognised as the indispensable data-and-index utility of a structurally larger Korean asset-management industry.

Base case. The data business grows steadily with the financial sector, and the index business continues to grow but at a decelerating rate as the ETF market matures and the year-on-year AUM comparisons get harder. The company delivers roughly on its value-up targets: dividends keep rising modestly off the KRW 260 base, payout stays above the 25% floor, and the share count stays broadly flat. The AI/LLM initiative ships as a useful product enhancement rather than a step-change. The governance situation stays unresolved but manageable - a low-grade overhang that periodically flares in the press but does not derail operations. FnGuide remains the dominant local data vendor and the #1 private index house, a high-quality recurring-revenue business whose growth is respectable rather than explosive, with the stock's volatility driven mostly by how the market prices the index-AUM trajectory.

Bear case. A market correction or a rotation out of covered-call/income ETFs shrinks the AUM base that FnGuide's index fees are levied on, and net ETF inflows stall - the highest-margin, fastest-growing revenue line reverses, and because the market had priced in continued index acceleration, the de-rating is sharp. One of the two dominant ETF houses moves toward self-indexing or shifts new launches to KRX or MSCI indices, structurally capping FnGuide's index share. The control dispute turns destructive: a protracted board fight, treasury shares deployed for control rather than returns, management distraction and possible leadership churn erode both execution and investor trust. Generative AI lowers the barrier for a funded entrant to structure and serve Korean financial data, slowly eroding the subscription moat. In this world FnGuide's defensive data base holds up, but the growth engine and the governance both disappoint at once, and the story that drove the 2026 re-rating unwinds.