Historical market data for backtesting and quantitative research.
This repository contains free samples, methodology notes, and reproducible examples for selected PowakaData research datasets.
Historical membership intervals for reconstructing the S&P 500 universe as it existed through time, with a documented point-in-time research contract.
Historical Nasdaq-100 membership and canonical daily research data for point-in-time universe construction.
Historical US dividend and stock split provider records for research workflows that need explicit event-date, adjustment, coverage, and provenance semantics.
- Public methodology & schema resource
- Python research-contract example
- Corporate actions backtesting guide
- Dataset page
- Provider-row sample: not published pending explicit redistribution/promotion authorization.
Using today's index constituents to backtest the past can introduce survivorship and universe-selection bias. A point-in-time membership contract separates historical eligibility from what is known today.
The included S&P 500 sample demonstrates interval semantics around a real membership transition and includes the contract needed to interpret interval boundaries correctly.
Historical dividends and stock splits are not automatically point-in-time. Event date, information availability, historical security identity, and price-adjustment semantics are separate questions. Applying an action to a series that already embeds the adjustment can also double count the transformation.
The Corporate Actions guide documents these issues without claiming that the historical provider-record product is point-in-time or survivorship-safe.
PowakaData also provides historical datasets across US equities, UK/LSE equities, futures, forex, crypto, commodities, indices, ETFs, and macro/economic data.
powakadata-samples/
├── README.md
├── examples/
│ ├── reconstruct_sp500_universe.py
│ ├── reconstruct_nasdaq100_universe.py
│ └── corporate_actions_research_contract.py
└── samples/
├── nasdaq100-point-in-time/
│ ├── README.md
│ ├── INTERVAL_SAMPLE.csv
│ └── EVENT_SAMPLE.csv
├── sp500-point-in-time/
│ ├── README.md
│ ├── INTERVAL_SAMPLE.csv
│ ├── EVENT_SAMPLE.csv
│ ├── INTERVAL_CONTRACT.json
│ └── METHODOLOGY.txt
└── us-corporate-actions/
├── README.md
├── SCHEMA.md
└── EXAMPLE_WORKFLOW.md
Only approved public samples are committed here. Full customer packages and unapproved provider records are not published. Where a public sample has not yet been approved, this repository links to the methodology/technical guide instead of fabricating example rows.