Historical market data for backtesting and quantitative research.
PowakaData provides inspectable historical datasets for quantitative researchers, systematic traders, developers, and anyone who wants to verify the data behind a backtest.
This repository contains free data samples and reproducible examples for selected PowakaData datasets.
Backtests are only as reliable as the data and assumptions behind them.
PowakaData focuses on research-ready datasets that can be inspected directly rather than hidden behind a proprietary API.
- Inspectable files
- Free samples before purchase
- Historical equities, futures, forex, crypto, commodities and more
- Point-in-time S&P 500 and Nasdaq-100 membership data
- Data designed for backtesting and quantitative research
- One-time purchase — no subscription required
A common source of bias in historical equity research is using today's index constituents to test strategies in the past.
For example, a 2018 S&P 500 backtest built from today's membership can implicitly use information that was not available in 2018.
Point-in-time membership data allows the research universe to be reconstructed using the constituents applicable to each historical period.
This repository includes examples showing how to:
- load point-in-time membership data;
- reconstruct a historical index universe;
- avoid using future membership information;
- join membership data to historical market data;
- build reproducible research inputs.
PowakaData currently provides historical datasets across:
- US equities
- UK / LSE equities
- S&P 500 point-in-time membership
- Nasdaq-100 point-in-time membership
- Futures
- Forex
- Crypto
- Commodities
- Indices
- ETFs
- Macro and economic data
Free samples are available so researchers can inspect file structure, columns, timestamps, and data format before purchasing a complete dataset.
Website: https://powakadata.com/
Explore a real pre-purchase sample of the PowakaData S&P 500 Point-in-Time Historical Membership Dataset.
- Inspect the S&P 500 Point-in-Time sample
- View the interval sample
- View constituent events
- Read the interval contract
- Read the methodology
- Run the Python reconstruction example
On 2026-08-18, the sample contains a real S&P 500 membership transition:
- AVB is no longer active.
- RDDT becomes active.
An EXACT effective_to boundary is exclusive. Treating every interval end as inclusive would incorrectly keep AVB in the historical universe.
The included Python example implements the interval contract and verifies this boundary automatically.
Browse the full PowakaData catalog →
powakadata-samples/
├── README.md
├── examples/
│ └── reconstruct_sp500_universe.py
└── samples/
└── sp500-point-in-time/
├── README.md
├── INTERVAL_SAMPLE.csv
├── EVENT_SAMPLE.csv
├── INTERVAL_CONTRACT.json
└── METHODOLOGY.txt