Builds custom indexes of the top-ranked securities by market cap — across both stocks and cryptocurrencies — with configurable size and minimum weighting rules.
This project generates multiple index variants, each defined by:
- Number of constituents (top)
- Minimum weighting per constituent (weight_min)
- Sort criteria (currently market_cap)
All indexes can be composed of both stock + crypto. Both asset classes are treated equivalently during index construction. Constituents are filtered by liquidity, listing age, and asset type restrictions, then ranked by market capitalization.
For detailed construction rules, see the Index Methodology.
| Name | Max Constituents | Min Weight (%) | Latest Build |
|---|---|---|---|
| top10-min10 | 10 | 10% | top10-min10.csv |
| top20-min5 | 20 | 5% | top20-min5.csv |
| top50-min2 | 50 | 2% | top50-min2.csv |
| top100-min1 | 100 | 1% | top100-min1.csv |
| top250-min0.4 | 250 | 0.4% | top250-min0.4.csv |
| top500-min0.2 | 500 | 0.2% | top500-min0.2.csv |
- Python 3.12+
- macOS (tested) or Linux (untested)
You can run the project without API keys; it will fall back to bundled snapshot data so you can demo the tool and poke around immediately. To fetch fresh market data, add API keys:
- API key from CoinMarketCap (free)
- API key from Financial Modeling Prep (free)
- API key from Polygon.io (free)
- Clone the repo
- Create a
.envfile in the root directory:
CMC_API_TOKEN=your_coinmarketcap_api_key
FMP_API_TOKEN=your_financial_modeling_api_key
POLY_API_TOKEN=your_polygon_api_key3_(Optional)_ Modify config.yaml to customize existing and/or create new indexes.
- Run
build.shto setup the virtual environment with installed dependencies. - Run
main.pyto build the indexes defined inconfig.yaml. - Results are written to
/indexes
- Cached data is stored in the
data/directory. - If API tokens are available, the cache refreshes daily with market data.
- Without tokens, the build uses the included snapshot data.
Originally built as a quick solution for managing a custom index via Fidelity Basket Portfolios, the project now supports multiple index definitions from a single YAML configuration.
- I'm actively working this and have lots of features planned:
- Building a static website to display the index results with fancy graphs
- Research pulling in the index historical data to determine performance over time
- Hooking up GitHub actions to have this run and update itself regularly
- I'm open to adding additional features or improvements — feel free to open an issue.