BIS Full Data Sets Bulk Downloads
Datadory delivers diversified capital markets data covering the BIS Full Data Sets Bulk Downloads: all 27 statistical topics the Bank for International Settlements compiles - locational and consolidated banking, debt securities, credit to the non-financial sector, credit-to-GDP gaps, debt service ratios, global liquidity indicators, exchange-traded and OTC derivatives, the Triennial Survey, property and consumer prices, bilateral and effective exchange rates, central bank balance sheets and eight payments-system tables - each topic as one complete-history table in four parallel layouts (column CSV, flat CSV, SDMX compact, SDMX generic), spanning 50-plus economies with series reaching back as far as 1946. Get a sample of this dataset and inspect the field structure before you commit.
What is the BIS Full Data Sets Bulk Downloads?
The whole BIS statistical shelf in one consistent shape. The Bank for International Settlements publishes 27 statistical topics - locational banking statistics (WS_LBS_D_PUB), consolidated banking statistics (WS_CBS_PUB), debt securities (WS_NA_SEC_DSS), international debt securities (WS_DEBT_SEC2_PUB), credit to the non-financial sector (WS_TC), credit-to-GDP gaps (WS_CREDIT_GAP), debt service ratios (WS_DSR), global liquidity indicators (WS_GLI), exchange-traded derivatives (WS_XTD_DERIV), OTC derivatives (WS_OTC_DERIV2), the Triennial Survey (WS_DER_OTC_TOV), residential and commercial property prices, consumer prices, bilateral and effective exchange rates, central bank total assets and policy rates, and eight CPMI payments-system tables.
The discipline is what makes it valuable: every topic ships as a single full-topic table carrying the complete historical series, not a rolling window, and every topic comes in four parallel layouts - a column-layout CSV (one row per series, one column per period), a flat-layout CSV (one row per observation), and SDMX 2.1 messages in compact and generic flavours. Twenty-seven topics times four layouts is 108 archives sharing one design grammar: learn the shape once, and cross-market leverage, property cycles, FX regimes and payment flows all open with the same key.
Get a sample of this dataset and we will return rows shaped exactly like the dictionary below, for the topics and economies you name.
What do BIS Full Data Sets rows look like?
Two layouts, one truth. The flat layout spends rows on observations; the column layout spends columns on time. Both carry the same dimension key:
# Flat layout -- one row per fully-keyed observation (shape preview)
FLOW_ID : WS_NA_SEC_DSS # topic key; e.g. debt securities statistics
FREQ : <frequency code>
REF_AREA : <reporting economy code>
INSTR_ASSET : <instrument / asset code>
MATURITY : <maturity band code>
UNIT_MEASURE : <unit code>
UNIT_MULT : <scale multiplier>
TIME_PERIOD : <observation period>
OBS_VALUE : <observed statistic>
OBS_STATUS : <observation status flag>
OBS_CONF_STATUS : <confidentiality flag>
# Column layout -- one row per series, history running rightward
<code+label pairs for each dimension> ... then one column per period
periods : 1946-Q1 ... latest # earliest start varies by topicSlots marked <...> are populated once we cut your sample; the column names and the dual-layout design are the archive set's own, not a Datadory invention. In the debt-securities topic alone the column layout runs to roughly 469 columns - the dimension key, then four decades of quarterly vintages marching off to the right. The same rows land identically whether they arrive as an API response, flat files, or a table in your warehouse.
What fields does the BIS Full Data Sets dictionary define?
Eleven fields anchor the core dictionary, and they repeat across every topic - which is precisely why the archive set compounds in value. Each definition below describes what the field holds in the published tables; the example column shows its shape in a sample cut.
Additional fields on request: the per-topic dimension axes beyond the core key - counterparty sector, currency denomination and valuation codes that differ between banking, securities and derivatives topics - plus the column sets behind the eight CPMI payments tables and the Triennial Survey turnover breakdowns. Topic-specific columns are confirmed against live records when we prepare your sample.
Where does coverage run, and at what grain?
Three chips summarize the footprint:
- Geography: global by construction. Securities and banking statistics cover 50-plus economies; exchange-rate and price series reach broad country sets; banking statistics pair reporters against counterparties in well over 100 countries.
- Time frame: complete history per topic from each series' own start - as early as 1946 - with publication stamps across the set running from April through August 2026. No rolling window, no vintage-stitching: one pull returns the whole timeline.
- Granularity: topic-level tables holding series keyed by their full SDMX dimensions - area, sector, instrument, maturity, currency, valuation - down to individual observation rows in the flat layout.
Set against the wider Datadory catalog - where the average quality score across all 1,744 datasets is 7.81 - this slice scores 9/10, carried by the breadth of topics behind one shared schema and the completeness of the histories.
How is BIS Full Data Sets delivered through Datadory?
API, files, or your warehouse. Daily, weekly, or hourly.
You choose the channel and the cadence; the field dictionary above travels unchanged across all three. Teams that think in panels take the column layout into a warehouse and pivot locally; teams that think in events take flat rows where the dimension key becomes the join spine. Want two topics daily and twenty-five weekly? That is a settings conversation, not a re-integration project.
Who builds on BIS Full Data Sets data?
Ranked by how directly the archive set answers their day job:
- Investors & quants. Cross-border credit, leverage and derivatives positioning across 50-plus economies is the raw material for macro factor models and tail-risk research - complete histories mean backtests need no splicing.
- Market researchers & consultants. Property price cycles, consumer prices and credit-to-GDP gaps across dozens of countries quantify financial-depth narratives with official figures rather than patched-together national sources.
- Competitive intelligence & product teams. Payments-system tables and banking structure reveal where clearing, settlement and correspondent activity concentrate.
- Data scientists. One dimension grammar across 27 economic domains is unusually clean training material for nowcasting and panel models.
- Developers & builders. Four parallel layouts per topic means one ingestion module serves every topic you ever add.
- Journalists & academics. Citation-grade international statistics on banking, debt and payments, with the depth to check any headline against the full series.
For contrast inside the family: the dedicated BIS locational banking statistics slice goes deeper on one banking panel, while the debt securities comparison weighs single-topic depth against this whole-shelf breadth.
Which notes pair with this dataset?
Notes that pair well with this page:
- Diversified capital markets data hub - the full pooled view of the industry, from exchange statistics to fixed-income transparency.
- Debt securities statistics vs the full bulk shelf - single-topic depth against twenty-seven-topic breadth, scored honestly.
- Persona pages - what investor-quant, market-research and developer teams each do with this slice, industry by industry.
BIS Full Data Sets field dictionary: core fields, types, definitions and example shapes
| Field | Type | Definition | Example |
|---|---|---|---|
| Column-layout table | text | Full-topic table with one row per series: each SDMX dimension appears twice (code and label side by side), followed by one column per time period. | ~469 columns in the debt-securities topic |
| Flat-layout table | text | One row per observation carrying TIME_PERIOD, OBS_VALUE and status flags alongside the dimension codes. | One row per observation |
| SDMX 2.1 message | xml | Structured XML rendering of the same series in compact (codes only) or generic (with labels) form for SDMX-aware pipelines. | Compact or generic variant |
| FLOW_ID | string | Topic identifier selecting which of the 27 statistical shelves a table belongs to, e.g. WS_LBS_D_PUB for locational banking statistics or WS_NA_SEC_DSS for debt securities. | WS_NA_SEC_DSS |
| FREQ | enum | Observation frequency code for the series; topics span annual, quarterly and monthly rhythms depending on what they measure. | Frequency code |
| REF_AREA | string | Reporting economy or aggregate the series describes; individual economy codes sit beside all-reporting aggregates. | Economy or aggregate code |
| INSTR_ASSET | enum | Instrument or asset axis separating loans and deposits from debt securities, derivatives and related categories within a topic. | Instrument code |
| MATURITY | enum | Maturity band code where a topic splits positions by original or remaining maturity. | Maturity band code |
| UNIT_MEASURE | string | Unit the observation is expressed in - currency amounts, indices, percentages or counts, per the topic's measurement conventions. | Unit code |
| UNIT_MULT | integer | Scale multiplier applied to observations so magnitudes stay machine-comparable (millions versus units, for instance). | Scale multiplier |
| TIME_PERIOD | date | Observation period. Period columns in the column layout reach back to 1946-Q1 for the longest-running topics; the flat layout carries one period per row. | 1946-Q1 onward |
| OBS_VALUE | number | The observed statistic itself, scaled by UNIT_MULT where applicable - the number every chart ultimately plots. | Scaled numeric observation |
| OBS_STATUS | string | Observation status flag qualifying how a value was compiled or estimated. | Status flag |
| OBS_CONF_STATUS | string | Confidentiality flag marking observations subject to disclosure restrictions. | Confidentiality flag |
Questions buyers ask
Which BIS topics are included in the Full Data Sets archive?
All 27 published statistical topics: locational and consolidated banking statistics, debt securities, international debt securities, credit to the non-financial sector, credit-to-GDP gaps, debt service ratios, global liquidity indicators, exchange-traded and OTC derivatives, the Triennial Survey, residential and commercial property prices, consumer prices, bilateral and effective exchange rates, central bank total assets and policy rates, and eight CPMI payments-system tables. One consistent schema spans the lot.
How far back does the history go?
To the start of each series. Archives hold complete historical runs rather than rolling windows, and the longest-running topics reach back to 1946-Q1. A single pull returns the entire timeline, so backtests and decade-spanning charts need no vintage-stitching.
Column layout or flat layout - which should I choose?
Both ship for every topic, so this is a pipeline taste, not a trade-off. The column layout puts one row per series with history running rightward - compact for analysts pivoting locally. The flat layout puts one row per observation with TIME_PERIOD and OBS_VALUE as explicit columns - friendlier to SQL joins and event-style processing.
When would I need the SDMX variants instead of CSV?
When your stack already speaks SDMX. Each topic renders as a 2.1 XML message in compact form (codes only, smallest payload) or generic form (labels included, human-readable). Everything the CSV layouts carry is present; only the envelope changes. Teams without an SDMX heritage can ignore both and lose nothing.
Does one archive really cover a whole topic?
Yes - each topic arrives as a single full-topic table containing every published series in that shelf, in each of the four parallel layouts. There is no per-country or per-quarter fragmentation to reassemble, which is what makes cross-topic joins a one-afternoon job rather than a quarter-long project.
Can I get a sample cut to specific topics and economies?
That is what the sample is for. Name the topics - banking claims, credit gaps, property prices, whatever your model eats - the economies and the period ranges, and Datadory returns rows matching the field dictionary above, with the topic-specific dimension columns confirmed against live records before anything recurring switches on.
See the rows before you pay anything.
Name this dataset and we send real records from it — scoped to the fields you asked for.