ml4t · stefan-jansen · Sep 25, 2026 · Sep 25, 2026 · Sep 25, 2026 · Sep 25, 2026
diff --git a/.github/workflows/ecosystem.yml b/.github/workflows/ecosystem.yml
@@ -17,7 +17,7 @@ jobs:
   qualification:
     permissions:
       contents: read
-    uses: ml4t/ecosystem/.github/workflows/qualify-library.yml@a3080f8e09b197098c9d796240e746aca516492e # 2026-09-23 policy snapshot
+    uses: ml4t/ecosystem/.github/workflows/qualify-library.yml@066dc737a879faea51a5b8a900315441712ea89a # 2026-09-25 policy snapshot
     with:
       import-package: ml4t.data
       prerelease-exception: python-315-polars
diff --git a/.github/workflows/release.yml b/.github/workflows/release.yml
@@ -64,7 +64,7 @@ jobs:
     needs: preflight
     permissions:
       contents: read
-    uses: ml4t/ecosystem/.github/workflows/qualify-library.yml@a3080f8e09b197098c9d796240e746aca516492e # 2026-09-23 policy snapshot
+    uses: ml4t/ecosystem/.github/workflows/qualify-library.yml@066dc737a879faea51a5b8a900315441712ea89a # 2026-09-25 policy snapshot
     with:
       import-package: ml4t.data
       prerelease-exception: python-315-polars

diff --git a/docs/getting-started/provider-selection.md b/docs/getting-started/provider-selection.md
@@ -24,9 +24,9 @@ data = SyntheticProvider(seed=42).fetch_ohlcv(
 | Equity bars | Yahoo Finance, Alpaca, EODHD, Tiingo, Twelve Data, Massive, Finnhub | Some adapters require credentials, accounts, or paid history |
 | Foreign exchange | OANDA, Twelve Data, FXMacroData | OANDA and Twelve Data require credentials |
 | Futures and options | Databento, Binance, OKX | Historical or licensed data may be metered |
-| Macroeconomic series | FRED, FXMacroData | FRED requires an API key for normal use |
-| Research factors | Fama-French, AQR | Network access and source-specific terms |
-| Prediction markets | Kalshi, Polymarket | Network access and changing public endpoints |
+| [Macroeconomic series](../providers/macro.md) | FRED, FXMacroData | FRED requires an API key for normal use |
+| [Research factors](../providers/factors.md) | Fama-French, AQR | Network access and source-specific terms |
+| [Prediction markets](../providers/prediction_markets.md) | Kalshi, Polymarket | Network access and changing public endpoints |
 | Frozen equity history | Wiki Prices | Local historical dataset ending in 2018 |
 | CFTC positioning | COT | Install the `cot` extra |
 

diff --git a/docs/providers/alpaca.md b/docs/providers/alpaca.md
@@ -146,5 +146,6 @@ headers, and retries transient failures per pagination page.
 
 ## See Also
 
+- [Equity](equities.md) and [ETF](etfs.md) source references
 - [Alpaca Market Data docs](https://docs.alpaca.markets/us/docs/about-market-data-api)
 - [Provider README](index.md)
diff --git a/docs/providers/alternative_data.md b/docs/providers/alternative_data.md
diff --git a/docs/providers/aqr.md b/docs/providers/aqr.md
@@ -91,6 +91,7 @@ When using AQR data, cite the relevant papers:
 
 ## See Also
 
+- [Research factor data sources](factors.md)
 - [AQR Datasets](https://www.aqr.com/Insights/Datasets)
 - [Fama-French Provider](fama_french.md)
 - [Provider reference](index.md)
diff --git a/docs/providers/binance.md b/docs/providers/binance.md
@@ -72,6 +72,7 @@ Binance may block access from certain countries. Consider using `BinancePublicPr
 
 ## See Also
 
+- [Cryptocurrency data sources](crypto.md)
 - [Binance API](https://www.binance.com/en/binance-api)
 - [BinancePublic Provider](binance_public.md)
 - [Provider reference](index.md)
diff --git a/docs/providers/binance_public.md b/docs/providers/binance_public.md
@@ -74,6 +74,7 @@ Same as Binance:
 
 ## See Also
 
+- [Cryptocurrency data sources](crypto.md)
 - [Binance Data Portal](https://data.binance.vision)
 - [Binance Provider](binance.md)
 - [Provider reference](index.md)
diff --git a/docs/providers/coingecko.md b/docs/providers/coingecko.md
@@ -74,5 +74,6 @@ be converted into correct daily OHLC values.
 
 ## See Also
 
+- [Cryptocurrency data sources](crypto.md)
 - [CoinGecko API](https://www.coingecko.com/en/api/pricing)
 - [Provider reference](index.md)
diff --git a/docs/providers/crypto.md b/docs/providers/crypto.md
@@ -0,0 +1,30 @@
+# Cryptocurrency Data Sources
+
+Cryptocurrency data is fragmented across venues, instrument types, and chains. Preserve exchange,
+pair, quote currency, contract type, and symbol history. Aggregated prices can conceal venue
+closures, wash trading, and differences between spot, futures, perpetuals, and options.
+
+!!! note "Verified September 2026"
+    Entries were checked against the linked official product documentation. Coverage and access
+    depend on the vendor contract and can change.
+
+| Source and product | Coverage or history | Acquisition | Access | Material research caveat | ml4t-data provider |
+|--------------------|---------------------|-------------|--------|--------------------------|--------------------|
+| [Binance Spot API](https://github.com/binance/binance-spot-api-docs) | Binance spot pairs, trades, order books, and bars | REST and WebSocket APIs | Public market endpoints; credentials for private or higher-limit use | Venue availability and product access vary by jurisdiction | `BinanceProvider` |
+| [Binance Public Data](https://data.binance.vision/) | Bulk spot and futures market files | Public archive download | Free | Files are venue-specific and symbol coverage changes over time | `BinancePublicProvider` |
+| [OKX Market Data](https://www.okx.com/docs-v5/en/#rest-api-market-data) | Spot and derivatives, including perpetual funding and premium data | REST and WebSocket APIs | Public market endpoints | Instrument and jurisdiction availability change; retain instrument metadata | `OKXProvider` |
+| [CoinGecko API](https://docs.coingecko.com/) | Aggregated asset, exchange, market, and on-chain DEX data | REST API | Demo and paid API plans | Aggregated asset identity and market prices are not a substitute for venue-level execution data | `CoinGeckoProvider` |
+| [CryptoCompare API](https://developers.cryptocompare.com/documentation) | Aggregated and exchange-specific crypto market data | REST and streaming APIs | API key for supported usage | Release qualification depends on live credential validation; aggregation methodology matters | `CryptoCompareProvider` |
+| [Kaiko Market Data](https://www.kaiko.com/products/market-data) | Centralized and decentralized venues, spot and derivatives, trades and order books | API, streaming, and cloud delivery | Institutional license | Venue and instrument history depend on the contracted product | No |
+| [Tardis.dev historical data](https://docs.tardis.dev/historical-data-details/overview) | Raw and normalized messages for centralized crypto exchanges, including closed venues | API and downloadable files | Commercial plan; limited samples | Reconstruction requires exchange-specific message semantics and snapshot handling | No |
+| [CoinAPI Market Data](https://www.coinapi.io/products/market-data-api) | Multi-exchange spot and derivatives market data | REST, WebSocket, FIX, and files | API key; plan-dependent | Normalized symbols and aggregate feeds can hide exchange-specific contract details | No |
+
+On-chain metrics, developer activity, and social signals belong in the
+[alternative-data reference](alternative_data.md). This page covers tradable market data.
+
+## Related References
+
+- [Alternative data sources](alternative_data.md)
+- [Futures data sources](futures.md)
+- [Options data sources](options.md)
+- [Market data selection](market_data.md)
diff --git a/docs/providers/cryptocompare.md b/docs/providers/cryptocompare.md
@@ -68,5 +68,6 @@ Consult CryptoCompare's current terms before use. Access and limits were not ver
 
 ## See Also
 
+- [Cryptocurrency data sources](crypto.md)
 - [CryptoCompare Pricing](https://min-api.cryptocompare.com/pricing)
 - [Provider reference](index.md)
diff --git a/docs/providers/databento.md b/docs/providers/databento.md
@@ -181,6 +181,7 @@ streaming, use `provider.client` directly.
 
 ## See Also
 
+- [Equity](equities.md), [futures](futures.md), and [options](options.md) source references
 - [Databento Pricing](https://databento.com/pricing)
 - [Databento Reference](databento_reference.md) - Detailed schema guide
 - [Provider reference](index.md)
diff --git a/docs/providers/eodhd.md b/docs/providers/eodhd.md
@@ -112,5 +112,6 @@ Get your API key at [eodhd.com/register](https://eodhd.com/register).
 
 ## See Also
 
+- [Equity](equities.md), [ETF](etfs.md), and [fundamental](fundamentals.md) source references
 - [EODHD Pricing](https://eodhd.com/pricing)
 - [Provider reference](index.md)
diff --git a/docs/providers/equities.md b/docs/providers/equities.md
@@ -0,0 +1,37 @@
+# Equity Data Sources
+
+This reference compares reproducible equity price and microstructure products. For cross-sectional
+research, prefer sources that retain delisted securities and expose corporate-action inputs. For
+intraday research, distinguish consolidated US feeds from a single exchange or venue.
+
+!!! note "Verified September 2026"
+    Entries were checked against the linked official product documentation. Coverage and access
+    depend on the vendor contract and can change.
+
+| Source and product | Coverage or history | Acquisition | Access | Material research caveat | ml4t-data provider |
+|--------------------|---------------------|-------------|--------|--------------------------|--------------------|
+| [Yahoo Finance](https://finance.yahoo.com/) | Global listed equities; vendor-defined history | Web-backed API through the provider | No credential | Not a survivorship-free research panel; adjustment and retention rules are vendor controlled | `YahooFinanceProvider` |
+| [Alpaca Market Data](https://docs.alpaca.markets/docs/about-market-data-api) | US equities and ETFs from 2016 | REST and WebSocket APIs | Account; feed entitlement depends on plan | The basic feed is IEX rather than the consolidated SIP; delisted coverage is not documented | `AlpacaDataProvider` |
+| [Tiingo End-of-Day](https://www.tiingo.com/documentation/end-of-day) | US equities, funds, and ETFs with raw and adjusted prices | REST API | API key; plan-dependent limits | Not documented as a survivorship-free universe | `TiingoProvider` |
+| [EODHD](https://eodhd.com/financial-apis/api-for-historical-data-and-volumes/) | Global end-of-day and intraday securities data | REST API | API key; coverage depends on plan | Exchange, adjustment, and delisted coverage vary by market and subscription | `EODHDProvider` |
+| [Finnhub](https://finnhub.io/docs/api) | US quotes plus plan-dependent global and historical data | REST and WebSocket APIs | API key; endpoint entitlement depends on plan | Free access does not establish a complete historical universe | `FinnhubProvider` |
+| [Twelve Data](https://twelvedata.com/docs) | Multi-exchange equities and other assets | REST and WebSocket APIs | API key; plan and exchange entitlements apply | Coverage and delays differ across exchanges | `TwelveDataProvider` |
+| [Massive Stocks](https://massive.com/stocks) | US stock trades, quotes, aggregates, and reference data | REST, WebSocket, and flat files | API key; history and feed depend on plan | Confirm SIP versus exchange coverage and corporate-action treatment | `MassiveProvider` |
+| [CRSP US Stock Databases](https://www.crsp.org/research/crsp-us-stock-databases/) | Active and inactive US securities, monthly from 1925 and daily from 1962 | Institutional files, commonly through WRDS | Academic or institutional license | Identifier and delisting-return conventions require deliberate handling | No |
+| [Norgate Data](https://norgatedata.com/data-content-tables.php) | US, Australian, and Canadian stocks with plan-dependent delisted history | Local database and client integrations | Paid subscription | Daily data only; survivorship and index-membership features depend on tier | No |
+| [Sharadar Equity Prices](https://data.nasdaq.com/databases/SEP) | Active and delisted US equities from 1998 | API and bulk delivery | Paid subscription | Daily bars; adjustment fields must be selected consistently | No |
+| [NYSE TAQ](https://www.nyse.com/market-data/historical/daily-taq) | Trades and quotes reported by US exchanges | Daily institutional files | Exchange data license | Raw records need symbol-history, correction, and session processing | No |
+| [AlgoSeek US Equities](https://www.algoseek.com/products.html#us_equity) | US trades, quotes, bars, reference data, and historical constituents | Cloud delivery and bulk files | Commercial license; samples available | Product schemas and SIP-derived fields differ; confirm the exact package | No |
+| [Databento Nasdaq TotalView-ITCH](https://databento.com/datasets/XNAS.ITCH) | Nasdaq order-level data from 2018 | API and batch download | Usage-based account | Raw exchange messages are not adjusted and do not represent the full US consolidated market | `DataBentoProvider` |
+| [Nasdaq TotalView-ITCH samples](https://emi.nasdaq.com/ITCH/Nasdaq%20ITCH/) | Selected full trading days for Nasdaq-listed securities | Public binary files | Free samples | Samples are sparse and unsuitable for a broad historical panel | `ITCHSampleProvider` |
+| [LSEG Tick History](https://www.lseg.com/en/data-analytics/market-data/data-feeds/tick-history) | Cross-asset trades, quotes, and depth from global venues | API and managed file delivery | Institutional license | Venue, contributor, and field coverage depend on the licensed package | No |
+
+`DataBentoProvider` and `MassiveProvider` are multi-asset classes. The mappings above describe only
+their equity capabilities. See the [provider comparison](index.md) for package-level credentials
+and interfaces.
+
+## Related References
+
+- [ETF data sources](etfs.md)
+- [Fundamental data sources](fundamentals.md)
+- [Market data selection](market_data.md)
diff --git a/docs/providers/etfs.md b/docs/providers/etfs.md
@@ -0,0 +1,31 @@
+# ETF Data Sources
+
+ETF research often needs more than traded prices. Fund identity, liquidation history, holdings,
+classifications, fees, distributions, and net asset value can each come from a different product.
+Do not treat an active-fund screener as a historical universe.
+
+!!! note "Verified September 2026"
+    Entries were checked against the linked official product documentation. Coverage and access
+    depend on the vendor contract and can change.
+
+| Source and product | Coverage or history | Acquisition | Access | Material research caveat | ml4t-data provider |
+|--------------------|---------------------|-------------|--------|--------------------------|--------------------|
+| [Yahoo Finance](https://finance.yahoo.com/markets/etfs/) | Listed ETF prices and current descriptive data | Web-backed API through the provider | No credential | Current listings and vendor-adjusted prices do not form a survivorship-free fund panel | `YahooFinanceProvider` |
+| [Alpaca Market Data](https://docs.alpaca.markets/docs/about-market-data-api) | US-listed ETF trades, quotes, and bars from 2016 | REST and WebSocket APIs | Account; feed entitlement depends on plan | Provides market data, not a historical holdings or classification database | `AlpacaDataProvider` |
+| [Tiingo End-of-Day](https://www.tiingo.com/documentation/end-of-day) | US equity, mutual-fund, and ETF prices | REST API | API key; plan-dependent limits | Not documented as a complete dead-fund universe | `TiingoProvider` |
+| [Tiingo Fund and ETF Fees](https://www.tiingo.com/documentation/mutual-fund-and-etf-fees) | Current and historical fee records for funds and ETFs | Enterprise data delivery | Institutional agreement | Fee history is a separate product from prices and holdings | `TiingoProvider` for prices only |
+| [EODHD](https://eodhd.com/financial-apis/stock-etfs-fundamental-data-feeds) | ETF prices plus plan-dependent holdings and fund fields | REST API | API key; plan-dependent | Holdings dates, classifications, and constituent weights require point-in-time checks | `EODHDProvider` |
+| [Massive Stocks](https://massive.com/stocks) | US-listed ETF trades, quotes, aggregates, and reference data | REST, WebSocket, and flat files | API key; plan-dependent | ETF-specific holdings and net asset value are not the same as exchange market data | `MassiveProvider` |
+| [Sharadar Fund Prices](https://sharadar.com/docs/funds) | Active and delisted US-listed funds, ETFs, closed-end funds, ETNs, and ETDs from December 1997 | API and bulk delivery | Paid subscription | Daily price product; verify whether a separate dataset is needed for holdings or classifications | No |
+| [Norgate Data](https://norgatedata.com/data-package-faq.php) | US ETF and ETN history from 1993, with plan-dependent delisted coverage | Local database and client integrations | Paid subscription | Daily frequency; constituent and classification features depend on package | No |
+| [CRSP Survivor-Bias-Free US Mutual Fund Database](https://www.crsp.org/research/) | Active and inactive US open-end mutual funds with fund characteristics and returns | Institutional files, commonly through WRDS | Academic or institutional license | It is a mutual-fund research database, not a complete ETF holdings product | No |
+
+The same ticker can represent different share classes or change its fund objective. Preserve stable
+identifiers and effective dates for classifications, fees, and holdings rather than joining only on
+the current ticker.
+
+## Related References
+
+- [Equity data sources](equities.md)
+- [Fixed-income data sources](fixed_income.md)
+- [Market data selection](market_data.md)
diff --git a/docs/providers/factors.md b/docs/providers/factors.md
@@ -0,0 +1,27 @@
+# Research Factor Data Sources
+
+Factor libraries publish returns from research portfolios, not returns from investable products.
+Choose a source by universe, formation method, weighting, frequency, and revision policy. Do not
+combine similarly named factors without reconciling their construction.
+
+!!! note "Verified September 2026"
+    This category has enough distinct, maintained sources for a standalone reference. Entries were
+    checked against the linked official or author-maintained data page.
+
+| Source and product | Coverage or history | Acquisition | Access | Material research caveat | ml4t-data provider |
+|--------------------|---------------------|-------------|--------|--------------------------|--------------------|
+| [Kenneth French Data Library](https://mba.tuck.dartmouth.edu/pages/faculty/ken.french/data_library.html) | US and international factors, sorted portfolios, industries, and breakpoints; core US monthly series begin in 1926 | Downloadable ZIP, CSV, and text files | Free | Histories can be reconstructed after upstream CRSP revisions; archive vintages when reproducibility matters | `FamaFrenchProvider` |
+| [AQR Data Library](https://www.aqr.com/Insights/Datasets) | Equity and cross-asset value, momentum, quality, low-beta, trend, and long-history research portfolios | Downloadable spreadsheets | Free under dataset terms | Series are hypothetical research portfolios, not AQR product returns; methodology differs by paper | `AQRFactorProvider` |
+| [Global Factor Data](https://jkpfactors.com/data) | 153 characteristics across 93 countries and four regions, with portfolio sorts and reference files | Configurable file downloads; stock-level data through WRDS | Factor returns free under CC BY-NC 4.0; stock-level access requires WRDS | Noncommercial data license and global data-screening choices constrain reuse and comparison | No |
+
+`FamaFrenchProvider.fetch()` selects supported factor, portfolio, industry, international, or
+breakpoint datasets. `AQRFactorProvider.download()` acquires the spreadsheets and `fetch()` reads a
+supported local dataset. Neither class computes live portfolio returns.
+
+## Selection Notes
+
+- Use Fama-French for canonical academic benchmarks and portfolio sorts.
+- Use AQR for paper-specific equity and cross-asset premia such as QMJ, BAB, VME, and TSMOM.
+- Use Global Factor Data for broad characteristic definitions and consistent country coverage.
+- Store the downloaded file and retrieval date because published research histories may be
+  revised.
diff --git a/docs/providers/fama_french.md b/docs/providers/fama_french.md
@@ -110,6 +110,7 @@ ff5_mom = provider.fetch_combined(["ff5", "mom"])
 
 ## See Also
 
+- [Research factor data sources](factors.md)
 - [Ken French Data Library](https://mba.tuck.dartmouth.edu/pages/faculty/ken.french/data_library.html)
 - [AQR Provider](aqr.md)
 - [Provider reference](index.md)
diff --git a/docs/providers/finnhub.md b/docs/providers/finnhub.md
@@ -74,5 +74,6 @@ Get your API key at [finnhub.io/register](https://finnhub.io/register).
 
 ## See Also
 
+- [Equity](equities.md) and [fundamental](fundamentals.md) source references
 - [Finnhub Pricing](https://finnhub.io/pricing)
 - [Provider reference](index.md)