Free Global Ticker Database

Free stock and ETF ticker reference data with collision-safe core listings, legacy primary tickers, listing-keyed venue rows, aliases, ISIN/WKN identifiers, cross-listings, and coverage reports.

Snapshot values are generated-report backed and intentionally human-formatted with comma separators and one-decimal coverage percentages. data/reports/coverage_report.json, data/reports/source_inventory_gap.json, and data/reports/entry_quality.json are the canonical machine-readable sources for these counts. source_inventory_gap.md is authoritative for current-scope source gaps; this snapshot must not claim zero missing current-scope sources while that report lists a missing source.

Generated release assets:

Reference and audit files:

core_listings.csv is the collision-safe canonical core export:

listing_key,ticker,exchange,name,asset_type,stock_sector,etf_category,country,country_code,isin,aliases,instrument_group_key,scope_reason
NASDAQ::AAPL,AAPL,NASDAQ,Apple Inc,Stock,Information Technology,,United States,US,US0378331005,apple,US0378331005,primary_listing

tickers.csv is the legacy compatibility export:

ticker,name,exchange,asset_type,stock_sector,etf_category,country,country_code,isin,aliases
KO,The Coca-Cola Company,NYSE,Stock,Consumer Staples,,United States,US,US1912161007,coca-cola

listings.csv is the full venue export:

listing_key,ticker,exchange,name,asset_type,stock_sector,etf_category,country,country_code,isin,aliases
NASDAQ::AAPL,AAPL,NASDAQ,Apple Inc,Stock,Information Technology,,United States,US,US0378331005,apple

Important rules:

• core_listings.csv is the canonical core security export; listing_key is its collision-safe current venue/symbol key.
• tickers.csv is a compatibility export that keeps one row per globally unique ticker.
• listings.csv and listing_key are the venue-level source of truth for exchange-specific listing identity.
• ticker is globally unique only in tickers.csv; use listing_key for venue-level identity.
• Stocks use stock_sector; ETFs use etf_category.
• instrument_scopes.csv marks core, OTC extended, and secondary cross-listings.
• Core rows without ISIN are tagged as scope_reason=primary_listing_missing_isin.
• Secondary listings stay in listings.csv and cross_listings.csv; core_listings.csv keeps one primary listing row per security.
• tickers.csv.aliases is restricted to conservative natural-language aliases. ISINs, WKNs, and exchange-ticker aliases stay in data/aliases.csv, data/core_aliases.csv, and identifier exports.
• data/adanos/ticker_reference.csv is the preferred import for Adanos Sentiment API ticker detection.

JSON metadata:

{
  "_meta": {
    "version": "3.36.0",
    "built_at": "2026-08-24T17:16:02Z",
    "total_tickers": 63821
  },
  "tickers": []
}

SQLite tables: tickers, listings, aliases, cross_listings, and instrument_scopes. Additional collision-safe tables: core_listings and core_aliases.

listing_key is the collision-safe current venue/symbol key. It is not a permanent historical identifier because symbols can change or be reused. Canonical v4 separates listing lifecycles from instruments and venues, while source observations and assertion tables preserve the evidence behind accepted values. Conflicting identifiers never merge listings into one instrument; they remain quarantined assertions until reviewed evidence resolves them.

The quality contract is cumulative: merge blocks structural, identity, history, safe-merge, source-governance, and canonical-schema failures; stable additionally requires passing official-full coverage, verified contributing-source rights, complete field provenance, and MIC mappings; complete additionally requires zero metadata and official-reference gaps. A green merge check never claims that the database is already complete or legally ready for a stable data release.

Canonical implementation source is committed as ordinary reviewable files. CI rejects compressed source payloads, workflow-time patching, and self-pushing workflows. It validates the canonical CSV contract, loads the result into PostgreSQL, and verifies deterministic repeat builds.

Operational rebuilds use:

python scripts/rebuild_canonical.py

Direct execution of scripts/rebuild_dataset.py remains available only for compatibility-export validation.

• Valid ISINs are checksum-verified.
• data/reports/alias_quality.csv classifies every alias as safe, review-only, or identifier-only for mention detection.
• data/reports/adanos_detection_simulation.json measures positive alias hits and negative false-positive probes for the Sentiment API import.
• Natural-language aliases are derived from current security names on every rebuild, then normalized to API-safe aliases.
• Duplicate natural-language aliases are either assigned to a clear best owner or removed from public alias columns.
• data/reports/entry_quality.csv stores one deterministic quality row per listing_key.
• data/reports/validation_report.json is the release gate: duplicate keys, invalid ISINs, typed sector/category leakage, blank country metadata on ISIN-bearing rows, mojibake name corruption, Adanos alias findings, unexpected entry-quality warnings, stale coverage counts, stale/unclassified residual source gaps, unreviewed US-primary foreign ISINs (ticker-collision suspects, allowlisted in data/review_overrides/foreign_isin_reviewed.csv), and stale source-of-truth decisions must be clean.
• data/reports/ohlcv_plausibility.csv stores optional market-data hygiene checks; default runs are no-network and omit unchecked rows unless local OHLCV samples, --fetch-yahoo, or --include-not-checked are provided.
• Obvious common-word, wrapper, celebrity, product, junk, short, and numeric aliases are filtered.
• Rights, units, warrants, notes, preferreds, and depositary lines are filtered from the stock universe.
• Foreign OTC country metadata is corrected from valid ISIN prefixes where possible.
• Official masterfiles are kept separate from secondary sources.
• Source licenses are recorded in data/masterfiles/sources.json. Only verified_open with hashed evidence can pass official-full contracts; verified_restricted and review_required stay fail-closed. See docs/SOURCE_LICENSING.md.
• Yahoo, EODHD, XTB, and FinanceDatabase are treated as reviewed candidate sources, not as exchange authority.
• Local probe/test artifacts are ignored via output/ and test-results/. CSVs under data/ are the diffable source of truth; generated JSON, SQLite, and Parquet files are release assets.

Top exchanges by primary ticker count:

For full exchange, country, source, and verification coverage, use:

python3 scripts/build_entry_quality_report.py
python3 scripts/build_coverage_report.py
python3 scripts/build_source_inventory.py
python3 scripts/build_exchange_source_audit.py
python3 scripts/build_etf_universe_completeness.py
python3 scripts/build_completion_backlog.py
python3 scripts/build_primary_isin_completeness.py
python3 scripts/build_cfi_code_review.py
python3 scripts/build_b3_residual_isin_review.py
python3 scripts/build_b3_residual_sector_review.py
python3 scripts/build_otc_scope_review.py
python3 scripts/build_canada_residual_review.py
python3 scripts/build_canada_figi_queue.py
python3 scripts/probe_canada_figi_batch.py --batch-id canada-figi-0001 --limit 10
python3 scripts/build_alias_quality_report.py
python3 scripts/build_adanos_ticker_reference.py
python3 scripts/simulate_adanos_detection.py
python3 scripts/validate_database.py
python3 scripts/build_ohlcv_plausibility_report.py
python3 scripts/fetch_symbol_changes.py
FINANCIALDATA_API_KEY=... python3 scripts/fetch_financialdata_symbols.py

Long OHLCV fetch runs should use streaming checkpoints:

python3 scripts/build_ohlcv_plausibility_report.py --fetch-yahoo --include-not-checked --stream --resume

Quick rebuild:

python3 scripts/rebuild_dataset.py
python3 scripts/build_listing_history.py
python3 scripts/build_entry_quality_report.py
python3 scripts/build_coverage_report.py
python3 scripts/build_source_inventory.py
python3 scripts/build_exchange_source_audit.py
python3 scripts/build_etf_universe_completeness.py
python3 scripts/build_completion_backlog.py
python3 scripts/build_primary_isin_completeness.py
python3 scripts/build_cfi_code_review.py
python3 scripts/build_b3_residual_isin_review.py
python3 scripts/build_b3_residual_sector_review.py
python3 scripts/build_otc_scope_review.py
python3 scripts/build_canada_residual_review.py
python3 scripts/build_canada_figi_queue.py
python3 scripts/probe_canada_figi_batch.py --batch-id canada-figi-0001 --limit 10
python3 scripts/build_alias_quality_report.py
python3 scripts/build_adanos_ticker_reference.py
python3 scripts/simulate_adanos_detection.py
python3 scripts/validate_database.py
python3 scripts/build_ohlcv_plausibility_report.py
python3 scripts/fetch_symbol_changes.py

Planned enrichment run:

python3 scripts/run_enrichment_pipeline.py --dry-run

Use --include-secondary-network for EODHD/Yahoo candidate stages. Use --apply-reviewed-backfills only when reviewed candidates should be merged into overrides.

Main targeted backfills:

Review queue:

python3 scripts/build_entry_quality_report.py
python3 scripts/build_b3_residual_isin_review.py
python3 scripts/build_b3_residual_sector_review.py
python3 scripts/build_otc_scope_review.py
python3 scripts/build_canada_residual_review.py
python3 scripts/build_canada_figi_queue.py
python3 scripts/probe_canada_figi_batch.py --batch-id canada-figi-0001 --limit 10
python3 scripts/build_ohlcv_plausibility_report.py
python3 scripts/audit_dataset.py --write-defaults
python3 scripts/run_claude_review_queue.py --model sonnet --skip-existing
python3 scripts/build_claude_review_overrides.py --min-confidence 0.8
python3 scripts/rebuild_dataset.py

The primary-ticker universe covers 86 exchanges. Source coverage is explicit: 48 exchanges are official_full, 33 are official_partial, and 6 current-scope exchanges have an official-source candidate awaiting parser implementation; a partial listing-company page or security lookup is never presented as a complete exchange directory. Implemented primary exchange/reference inputs include Nasdaq Trader, Nasdaq Nordic, ASX, Deutsche Boerse, B3, TMX, Euronext, JPX/TSE, TWSE, TPEX, SSE/SZSE, Bursa Malaysia, BME, BMV, WSE/NewConnect, TASE, KRX, HOSE/HNX/UPCOM, CSE Sri Lanka, and SEC company tickers.

Official source candidates and reconciled source gaps are tracked in data/masterfiles/source_candidates.json and summarized by data/reports/source_inventory_gap.md. Current source coverage status: 5 missing current-scope exchanges, 5 parser todo rows, 0 real global-expansion candidates, 49 official-full exchanges, and 33 official-partial exchanges. Remaining work includes source-parser backlog plus field-completion and taxonomy coverage.

data/reports/exchange_source_audit.md is the one-row-per-exchange operational audit for product-class gaps, source freshness and availability, official denominators, recall, blocker class, and promotion readiness. data/masterfiles/exchange_scope_decisions.csv gives every partial exchange a distribution-safe public scope. Its reason_code records the blocker observed at review time; validation remains fail-closed by requiring the current audit to contain a known promotion blocker, without invalidating a safe retained scope merely because another blocker becomes higher priority. A current audit with no blocker requires explicit scope review. Commercial products are only evaluation candidates: data/masterfiles/commercial_source_options.csv does not grant redistribution rights, and its validator requires explicit contract review or prior written permission.

Secondary/reviewed enrichment inputs include EODHD, FinanceDatabase, official B3 COTAHIST files, NYSE Group Security Master sample files, TradingView free scanner metadata, XTB OMI specification data, Yahoo Finance review helpers, FinancialData.net symbol-universe matching, OpenFIGI, GLEIF, and curated production aliases from api.adanos.org.

FinancialData.net output is intentionally review-only: the international-symbols endpoint has trading_symbol and registrant_name, but no ISIN or sector. The sync writes data/financialdata/international_stock_symbols.csv, data/reports/financialdata_symbol_match.md, data/reports/financialdata_current_exchange_gaps.csv, and data/reports/financialdata_global_expansion_candidates.csv. Missing rows are split into current-exchange gaps and global expansion candidates. The follow-up scripts/build_financialdata_isin_supplements.py only writes supplemental core rows when the FinancialData discovery row matches an official active masterfile row with a valid ISIN, name gate, no existing global ticker, and no existing/selected ISIN.

• License: MIT
• Changelog: CHANGELOG.md
• M2 operations: docs/m2_operations.md
• Releases: GitHub Releases
• Contributing: CONTRIBUTING.md