Sources & Markets · data for licence
Data for licence
Live — Korea (KRX / DART) · Japan (EDINET) · Taiwan (TWSE) · United Arab Emirates (ADX and DFM). In build — China A-shares, Hong Kong, India and Saudi Arabia. Same schema, same licence, same price. A new market is not a new contract.
We build datasets from public filings — Korea, and now the UAE (ADX+DFM) — and sell what the terminals don't carry — the labour fingerprint of listed companies, the broker tape, a near-census wage panel, and disclosures ranked by likely price impact. Every figure is re-derived from the source and ships with its own correction record.
This page is grouped by how you receive it, not by subject — because the first thing worth knowing is whether you can have it right now.
Every dataset below has a free page and a free sample. The full file — every row, as CSV and Parquet, with a data dictionary — is licensed. Prices are listed. You will not be asked to request a quote.
| Licence | What you get | Price | Billing |
|---|---|---|---|
| Page & sample | The published page and a sample file. No registration. | Free | — |
| Academic | One dataset for a thesis, dissertation or published paper. Cite us and we list your work. | $79 | One-time |
| Single dataset | One dataset, complete — CSV + Parquet + data dictionary — plus every release for twelve months. | $990 ($82.50/mo) | Per year, billed annually |
| All datasets | Every dataset on this page, plus every release for twelve months. | $2,990 ($249/mo) | Per year, billed annually |
The prices above are billed once a year, with the monthly figure in brackets shown only for comparison — that annual licence has no automatic renewal; it simply ends after twelve months unless you buy again, and we will not put a card on file for it. A separate recurring monthly plan is also available — use the Annual / Monthly toggle below — which renews every month until you cancel.
Buy
Loading checkout…
Paid in US dollars, from any country. Card or PayPal — no PayPal account needed for cards. No account with us, no card kept on file, and no automatic renewal: an annual licence simply ends after twelve months unless you buy again. The download link never expires. Files arrive as one .tar.gz (7-Zip on Windows; macOS and Linux open it natively). Need an invoice for procurement? Write to admin@klifedesign.net — that route still works.
Licence terms
- Use inside your organisation, including in client deliverables and research output.
- Redistribution of the file itself, or resale as a dataset, is not included.
- Attribution: SMarkets, with the release date.
- Source data is public. We license our compilation, normalisation and coverage work — not the underlying facts, which nobody owns.
Read it here 35
Free pages that publish the counts in full — including the gaps we could not measure. Each one counts the very file we license, so the free figures and the paid figures cannot disagree.
Free page
Asia & Gulf screener — filter 2,834 companies yourself
Set your own ranges on market cap, revenue, profit, PER, PBR, ROE and debt across five markets — KOSPI, KOSDAQ, KONEX, Abu Dhabi and Dubai — and read the companies that come back, in one table, in US dollars. It is the only screener we know of that counts the companies a filter removes because the figure does not exist rather than because they failed the test: a PER screen alone removes 1,246 of 2,834 that way, including every single Gulf listing, and every other screener drops them in silence.
Free toolOpen the screener →Free page
Korean semiconductor stocks — broker target prices on one page
SK hynix, Samsung Electronics and every other Korean chip stock with broker research in our sample: number of reports and firms, range and median of each firm's latest target, and the gap to the latest close. Research published in Korean, summarised in English. The rule for what counts as semiconductor is on the page.
Free page
Korean shipbuilding stocks — broker target prices on one page
HD Hyundai Heavy Industries, Hanwha Ocean, Samsung Heavy Industries and the other Korean shipbuilders with broker research in our sample: reports, firms, range and median of each firm's latest target, and the gap to the latest close. The rule for what counts as shipbuilding is on the page.
Free page
Korean defense and aerospace stocks — broker target prices on one page
Hanwha Aerospace, Hyundai Rotem, LIG Defense & Aerospace and Korea Aerospace Industries with the other defense and aerospace names in our research sample: reports, firms, range and median of each firm's latest target, and the gap to the latest close. The rule and its named exceptions are on the page.
Free page
Korean EV battery stocks — broker target prices on one page
LG Energy Solution, Samsung SDI, POSCO Future M, L&F and the other Korean battery and battery-materials names in our research sample: reports, firms, range and median of each firm's latest target, and the gap to the latest close. The rule for what counts as battery is on the page.
Free page
Korean biotech and pharma stocks — broker target prices on one page
Samsung Biologics, Celltrion, Hanmi Pharmaceutical, Alteogen and the other Korean drug makers and biotech companies in our research sample: reports, firms, range and median of each firm's latest target, and the gap to the latest close. The rule for what counts as biotech and pharma is on the page.
Free page
Who moved a target price, and when
Every Korean broker report that raised, cut or initiated a target price — counted from the same tape we license. The window closes after 30 days at the source, so a day not collected is a day gone.
Free page
The analyst list nobody keeps
The source recomputes its Korean analyst list every day over a rolling month and keeps no history, so half the roster is gone ten days later. We photograph it daily. The field it calls a rank is a row position in an alphabetical list, and the accuracy column is blank for a third of rows — we say both rather than printing a leaderboard.
Free page
Ownership and insider filings
Large-holding reports and executive share filings as filed: who moved a stake up, who moved it down, and where shares changed hands while the percentage did not.
Free page
Convertibles, warrants and exchangeables
Korean CB, BW and EB issues as filed — private versus public placement, zero-coupon convertibles, and refix floors. Exchangeables have no refix column by design, and we say so instead of writing zero.
Free page
Workforce, tenure and the pay gap
Headcount, tenure and annual pay per person by gender for Korean listed companies — as a distribution, never as a benchmark. These are ratios, not reasons, and no company is named or ranked.
Free page
What is in the Korean financial-statement tape
Annual statements for 2,709 listed Korean companies over five years, 2021 to 2025 — 13,545 company-year rows, and the part nobody publishes: coverage is about 98 percent on KOSPI and KOSDAQ but 12 percent on KONEX. Consolidated and separate are recorded per row and never mixed. 38 percent of filers reported net income of zero or less, which is why a price-earnings screen silently drops a third of the market unless you decide what to do about it.
Free page
Korea valuation tape — PER, PBR, ROE by company
Every company listed in Korea with its price-earnings, price-book and return-on-equity figures, in English. 1,570 of 2,709 have a PER — the rest did not earn a profit or did not file, and we say which rather than dropping them. Both dates ride on every row, so you can recompute the ratio against any other trading day. No sector averages.
Free page · CSV belowRead the tape →Free page
What share of KOSPI is each company
Every KOSPI company ranked by its share of total market value — 828 companies, company-level, free and in English. The two largest are 52 percent of the market between them; the 528 smallest are 1.5 percent together. This is full market capitalisation, not the free-float index weight, and the page says so above every table.
Free page
The largest companies in Korea, ranked four ways
Market value, revenue, total assets and operating profit — four orders from one dataset, never merged into a score. Take the top ten of each and you get 21 companies, not ten; only two make all four. One firm sits 3rd on one measure and 270th on another. The lists that disagree are the point: we draw the ground and you set the needle.
Free page · CSV belowSee the four lists →Free page
Korea Index Tape — 168 KRX indices, English names and composition
Every KOSPI, KOSDAQ, KRX and thematic index KRX publishes — 168 of them, every name in English, as of the latest snapshot. Seven are computed from a formula rather than a basket, and we say which. Where the source left a placeholder instead of a real annual low, the page says not measured rather than printing a false zero.
Free page
Korea Concentration Index — market × trade
How top-heavy Korea is — Samsung’s share of the KOSPI, KOSDAQ vs KOSPI, and Korea’s biggest export and import partners, measured with one lens: whatever you rank, the top four are about half and every Gini clears 0.9. A cross of KRX and Korea Customs data no single-feed vendor builds.
Free page · CSV belowSee the index →Free page
Korea lists thousands of currency-futures rows. Almost none trade.
A listed contract and a traded contract are not the same thing. Across the sessions we hold, about one listed currency-futures contract-day in thirty printed a single trade — and of the flex rows, none did at all. The count only works because we keep the zero-volume rows that tidier feeds drop. We count the rows; we do not say anything should be delisted.
Free page
How long Korea’s listed-company CEOs have held the seat
Korean annual reports state how long each registered executive has held the position. We read that field across 2,769 listed companies, counting seats rather than people. The median is nine years — but 23% have held the seat under three, so the shorthand about decades-long owner-CEOs describes only one tail. We publish the distribution rather than an average, report the 378 unreadable filings as unmeasured, and set aside 147 rows claiming a tenure longer than the company has existed.
Free page
Korea’s monthly numbers have a moving denominator
Every monthly figure out of Korea is a count across however many days the workplace was open, and that count moves. An average February ran 18.5 days against 21.2 in July, and the annual mean fell about a tenth across fifteen years. The series also breaks twice where the survey was renumbered — and we measure each break two ways, because the stated step compares December with January and carries the season inside it. We do not splice the series, and we do not say why the count fell.
Free page
Korea’s small-business outlook has never once undershot
Korea publishes two sentiment readings for small merchants on the same 100-point scale: what the coming month is expected to look like, and what the month just lived through felt like. Across 149 months the forward reading sat above the lived one every single time, median gap 21.3 points. We publish the distance without naming its cause — a run that clean is as likely to be a property of the two instruments as of anyone’s mood.
Free page
One line item, several Korean spellings
If you read Korean filings by machine, you match line items by name — until the same item is written another way. In the sample we hold, 414 Korean line-item names collapse to 349 English terms, and 45 of those terms carry more than one Korean spelling. Stripping whitespace fixes 8 of them; the other 37 are different words for the same thing. We do not say what share of an arbitrary filing this covers, because the sweep across all listed companies has not been collected yet.
Free page
How many Korean listed companies stayed profitable
Of 2,049 companies listed in Korea that filed a net profit in all five years, 772 (37.7%) were profitable every year and 193 (9.4%) lost money every year. Counted across the whole market, not just the large caps.
Free page
Two sizes, one gold bar: how far the prices drift
The Korea Exchange gold market lists a 1kg bar and a 100g mini bar of the same 99.99% metal, settled on the same venue — and they do not close at the same price per gram. Across the sessions we hold, the mini bar closed cheaper on roughly a third of them. We count the gap; we do not say why it moves, and we do not say either size is mispriced.
Free page
Korean listings that did not trade at all
Listings on KOSPI, KOSDAQ and KONEX that recorded no trade in any session of the window — counted from the Korean government open-data price feed, with the market capitalisation they still carry between them. Some never changed price either. Whether any of them is formally suspended is not in that feed, so we do not say it is.
Free page
stk_bydd_trd and ksq_bydd_trd — every KRX Open API field explained
Every field in KRX’s two daily-trading datasets, in English, with what each one actually means — not just a column list. Built for anyone pulling the raw API and wondering what a field name like FLUC_RT or ACC_TRDVOL stands for.
Free page
Fixed vs floating-rate loans at Korean banks
Korean deposit banks’ mortgage book runs 61.7 percent fixed-rate on an outstanding-balance basis, but only 35.3 percent of mortgages newly extended that month were fixed-rate — the back book and the new flow disagree. Bank of Korea data, composition snapshot, no rate forecast.
Free page
Producer prices vs import prices since 2020
Korean producer prices are up 28.6 percent since December 2020; import prices (won basis) are up 59.2 percent over the same window, nearly double — same 2020=100 base, two different indices, no currency or tariff cause claimed.
Free page
Construction vs services production since 2023
Construction output is down 24.3 percent from its August 2023 peak; services output is up 25.2 percent over the same August-to-August window — same raw index, two industries, compared month-for-month to avoid smuggling in seasonal swings.
Free page
What Korea trades — not who it trades with
Machinery and transport equipment made up 59.0 percent of Korea’s 2025 exports but only 35.9 percent of its imports; mineral fuels run the other way, 22.2 percent of imports against 6.7 percent of exports. Classified by product (SITC), not by trading partner.
Free page
Korea’s margin debt doubled — and rotated out of KOSDAQ
Korea’s stock margin-loan balance rose from 16.8 trillion won in January 2025 to 33.4 trillion in September 2026, peaking at 38.6 trillion in June — while KOSDAQ’s share of that debt fell from 42 percent to 22 percent. Financial Investment Association data, no cause assigned.
Free page
One line of Korea’s CPI basket moved 8.6x the headline rate
Korea’s headline CPI rose 3.09 percent year-on-year in August 2026 — the mobile-phone-fee sub-index alone rose 26.74 percent. Statistics Korea data, verified directly from KOSIS.
Free page
Korea’s real wage grew 24.1 percent in a decade, then 0.2 percent since
Korea’s inflation-adjusted average monthly wage rose from 2.90 million won in 2011 to 3.60 million in 2021 — then moved just 0.2 percent over the following four years. KOSIS wage and CPI data, no cause assigned.
Free page
Korea lists bonds on three boards — government, general and small-lot yields
South Korea lists bonds on three separate boards and almost every published yield comes from one of them. The government inter-dealer board carries 8 to 12 listings; the general board 270 to 346; the small-lot board a near-fixed 40 that Korean households actually buy. Daily median yields and the spread between the boards.
Free page
Korea’s fund shelf, dated — what got registered, by year and by type
Korea’s fund registry dates every code it holds. Read by establishment year, the shelf has changed hands three times: mixed-bond funds through 2008, derivative funds from 2009 to 2023, and funds-of-funds since. 183,351 codes, every year and every type.
Download it now 16
Files you can take today. No account, no key, no waiting on us.
File
Korea Consensus Tape (daily CSV)
Every Hankyung Consensus broker report in the source’s own rolling ~31-day window — 550 reports, 57 raised/lowered target prices, 158 named analysts, 12 brokers, 116 tickers on this edition. A dated snapshot, not a live feed: a day not collected is gone for good. Not a buy/sell signal — target_change is a fact about what a broker filed.
File
Listed-company workforce table (CSV)
The company-level table the sector panel is built from — every listed company with employees, tenure, pay and market cap. Free to download.
File
Korea valuation tape (CSV)
A free sample of the 100 largest companies with PER, PBR, ROE, debt to equity, market cap, net income, equity, assets, revenue and operating income — English names and industries, the price date and the fiscal year on every row, and a reason on the rows we could not compute. Empty means not measured; it never means zero. The full tape covers all 2,709 listed companies.
File
Korea Concentration Index (daily CSV)
The concentration figures behind the index page, rebuilt every trading day, in one flat file you can cite.
File
Bond trading concentration (CSV)
How concentrated Korean bond trading is, in the same flat shape as the equity file.
File
Korea People Panel — free sample (100 companies, CSV)
The 100 companies with the largest headcount, every column — tenure, headcount and pay per person by gender, joined to KRX market cap. The full panel covers all 2,924 companies. Also available as Parquet on the product page.
File
Korea Ownership Ledger — free sample, filings (100 rows, CSV)
The 100 most recently filed large-holding (5% rule) reports, every column. The full ledger covers all 21,785 filings. Also available as Parquet on the product page.
File
Korea Ownership Ledger — free sample, executives (100 rows, CSV)
The 100 most recently filed officer/major-shareholder ownership rows, every column. The full ledger covers all 32,535 rows. Also available as Parquet on the product page.
File
Korea Mezzanine Book — free sample (100 rows, CSV)
The 100 most recent CB/BW/EB board-resolution filings, every column. The full book covers all 5,084 rows. Also available as Parquet on the product page.
File
Who are South Korea’s largest trading partners
Every partner country ranked by two-way trade, with the balance beside it — and the ranking nobody puts next to it: the largest partner and the largest surplus are different countries. China is 21 percent of all trade; the biggest surplus comes from the United States, on less trade. Customs goods basis, services excluded and said so, with the long tail of small partners counted rather than dropped.
Free pageSee the ranking →File
Korea trade dataset — every partner, every month
Exports, imports and the balance with all 243 partner countries, month by month, on a customs basis — normalised to English in one clean CSV. The raw source is Korean and awkward; this is the ready extract. Free preview; full set is a one-time buy.
$29 — one-timeSee what’s in it →File
Who holds Korean securities abroad
Every country that discloses a Korean securities position to the IMF, ranked — a floor, not a total, since non-reporting countries (China among them) are absent rather than zero. The United States, Singapore and the UK lead; financial-centre domiciles like Luxembourg and the Cayman Islands rank high for reasons that have little to do with those countries’ own savers.
Free pageSee the ranking →File
Japan-listed companies — the EDINET registry
Every company on file with Japan’s Financial Services Agency that carries a live securities code — 3,818 of them, in English, by industry, from the EDINET Code List (government open data, not the exchange’s own file). This is the roster only: financial statements, disclosures and prices are not built yet, and this page says so.
Free pageSee the registry →File
Korean material disclosures, in English
Every filing Korean listed companies made to DART that changes what a reader would conclude — supply contracts, capital raises, buybacks, control changes, CEO changes, delisting risk — tagged into event types and translated, each row linking to the original. Not the full disclosure stream: the selection rule is one rule in our code, applied the same way every day.
Free pageSee the events →File
Korean and Japanese material disclosures, side by side
The same question asked of two regulators — what must a listed company tell the market at once? — and two different answers. Korea’s material filings are mostly about money entering or leaving the company; Japan’s extraordinary reports are mostly about who controls or decides. Only three of the event types exist on both sides, and we show the counts on comparable windows rather than raw totals.
Free pageCompare the two markets →File
Has the Gulf actually diversified?
Oil and non-oil GDP for Saudi Arabia, the UAE, Qatar, Kuwait, Oman and Bahrain, as the GCC’s own statistics office splits it. The non-oil share of GDP — the number everyone quotes — rises whenever the oil price falls, and it did exactly that in 2014–2016 before falling back. We publish the real size of the non-oil economy beside it, so you can see which movements were structural and which were arithmetic.
Free pageSee both measures →
Licensed file, delivered by edition 12
Dated editions with a data dictionary and a correction record. We are building the self-serve download page now; until it opens we send editions directly, and we do not post a price for something you cannot yet fetch yourself.
Licensed file
Corporate credit data — the SMarkets Grade
A credit grade for every listed company we cover, computed from the statements the company itself filed. Six measures — interest cover, borrowings to assets, liabilities to equity, current ratio, return on assets and the Altman Z — scored against published thresholds and published weights. Every threshold, every weight and every input line item ships with the file, so you can recompute the grade yourself and disagree with it. No analyst judgement, no sector premium, no forecast. A company whose filing is missing a figure is marked NR, never given a low grade. Delivered with the rating actions the exchanges themselves disclose, so you can put our grade beside the agencies’ own.
Licensing — coming soonSee how it is computed →Licensed file
Broker target-price accuracy
Every broker report’s target price scored against the stock 12 months later. Hit rate by house — from 49% down to 13%. A multi-year archive nobody else keeps.
Licensing — coming soonSee what’s in it →Licensed file
The sector workforce panel
2,644 listed companies in 11 sectors, each carrying tenure, pay, the gender pay gap and pay-per-head — the columns the big vendors leave out. Leaders flagged. Quarterly.
Licensing — coming soonSee what’s in it →Licensed file
The National Pension wage panel
A near-census pay ladder for the Korean workforce from the national pension bill — 553,978 workplaces, 11.56m people, by industry, region and firm size, with a cap-pressure column so every figure is a floor.
Licensing — coming soonSee what’s in it →Licensed file
How companies split themselves
Half of Korea’s listed companies report themselves as a single segment — but among companies above $10bn, only 23% do. Which filings break the business out, and how finely. 737 companies from their own annual reports.
Licensing — coming soonSee what’s in it →Licensed file
Board composition by sector
Who runs Korean listed companies, by sector — the share of women officers (7.2% overall), owner-led firms, CEO tenure and boardroom age. From 35,004 officers’ own filings.
Licensing — coming soonSee what’s in it →Licensed file
Sector leaders vs the rest
Split each sector into its market-cap leaders and the rest, and our labour data separates: in all 8 classified sectors the leaders pay 1.2–1.43x more and mostly retain longer — and in heavy industry and finance, employ fewer women at the top. A size split, not a stock pick. Not investment advice.
Licensing — coming soonSee what’s in it →Licensed file
Broker candour — the vanishing Sell
Korean broker research by rating, house by house: Buy on four in five reports, Sell on almost none. Of 13 houses we track, 10 issued no Sell rating at all this year. The rating as filed — not whether it was right, and not investment advice.
Licensing — coming soonSee what’s in it →Licensed file
Analyst attention by sector
Where Korean equity research piles up and where it thins — from our own archive of single-stock broker reports. A third of listed companies drew a report this year; the busiest name drew 100, most drew none. Attention concentration, not investment advice.
Licensing — coming soonSee what’s in it →Licensed file
UAE material disclosures (ADX+DFM)
Every Abu Dhabi (ADX, 96 companies) and Dubai (DFM, 131 companies) disclosure, ranked by likely price impact against US SEC Form 8-K categories — earnings, ownership change, delisting risk, CEO change and more.
From $990/yr — buy belowJump to checkout →Licensed file
UAE board & ownership panel
Board and management rosters (ADX) plus substantial (5%+) shareholders (ADX+DFM) — the service tier, not the main UAE product.
From $990/yr — buy belowJump to checkout →Licensed file
UAE financial highlights (ADX+DFM — Revenue/Net Profit/EPS, plus ADX balance sheet)
Revenue, net profit and EPS for every ADX and DFM quarterly filing — ADX from the exchange’s own AI summary (76 operating companies; the other 20 ADX listings are ETFs and file no income statement), DFM parsed directly from the filed PDF (49 of 76 filers; the rest file scanned images we will not guess at). Total assets, liabilities and equity are included for 121 of 292 ADX filings — published only where assets = liabilities + equity reconciles against the filing.
From $990/yr — buy belowJump to checkout →
Live API 1
The same data over HTTP, for people who would rather call than download. Self-serve keys are open.
API
SMarkets Data API
Issue your own key and call the endpoints directly — research archive, concentration and trade. Free tier for trying it; paid tiers raise the ceiling.
Free tier · Pro from $99/moRead the API docs →
Built from DART annual filings, KRX close and the National Pension workplace file via data.go.kr (licence "unrestricted"). What we sell is our parse, join and correction record — not the raw filings. Sector map derived from public KSIC; no GICS data or trademark is used. Not investment advice. See our terms of use and refund policy.