| --- |
| pretty_name: Stocks Yearly Risk Factors |
| language: |
| - en |
| license: other |
| task_categories: |
| - text-classification |
| - time-series-forecasting |
| tags: |
| - finance |
| - quantitative-trading |
| - backtesting |
| - stocks |
| - equities |
| - sec-filings |
| - text |
| - yearly |
| - 10-k |
| - risk |
| size_categories: |
| - 100K<n<1M |
| extra_gated_prompt: "This dataset is free to browse and gated for download. Approval is tied to a Papers With Backtest subscription, which also covers the other datasets in this organisation and the strategy catalogue at https://paperswithbacktest.com. Plans and what each one includes: https://paperswithbacktest.com/pricing" |
| --- |
| |
| # Stocks Yearly Risk Factors |
|
|
| The risk factors companies disclose in their annual reports since 2015, each classified in a three-level taxonomy, with the passage that supports the label. |
|
|
| <!-- MEASURED:START - rewritten by dataset_cards/refresh.py --> |
| |
| Shape measured at the next refresh of `dataset_cards/facts.json`. Refreshed monthly. |
|
|
| <!-- MEASURED:END --> |
|
|
| ## Why It Matters |
|
|
| Item 1A of the 10-K is where a company says what it fears. The papers on disclosed risk -- the count and novelty of risk factors as a predictor of volatility, the cross-section of firms exposed to one named risk (a supplier, a regulator, a currency, a climate), the text-based peer groups -- had no classified source. Here each disclosed risk is one row with its category, so a paper selects the firms citing a given risk in a given year without reading the documents. |
|
|
| Rows carry the filer's CIK and a `symbol` resolved from it through the tickers reference and the EDGAR index, one ticker per filer. Registrants with no listed equity at all -- private companies with public debt, trusts, subsidiaries -- and delisted names whose CIK neither reference knows are **not** in this copy: no strategy can hold them. Measured on the source, that is the larger part of the filers in the early years and a shrinking one since, so the coverage is of the listed market, not of EDGAR. `date` is the filing date, the day the text became public. |
|
|
| ## Load It |
|
|
| ```bash |
| pip install --upgrade pwb-toolbox |
| ``` |
|
|
| ```python |
| from pwb_toolbox import datasets as pwb_ds |
| |
| df = pwb_ds.load_dataset("Stocks-Yearly-RiskFactors", symbols=["AAPL"]) |
| print(df.iloc[-1, :]) |
| ``` |
|
|
| ## Columns |
|
|
| | Column Name | Description | |
| | :--- | :--- | |
| | **symbol** | Ticker resolved from the filer's CIK. | |
| | **date** | Filing date on EDGAR. | |
| | **cik** | Filer's SEC Central Index Key. | |
| | **primary_category** / **secondary_category** / **tertiary_category** | The taxonomy, from the broad family to the specific risk. | |
| | **supporting_text** | The passage of item 1A the classification rests on. | |
| | **filing_url** | The document on sec.gov. | |
| |
| ## Access |
| |
| Browsing the card and the schema is open to anyone. Downloading the files needs an approved request, tied to a subscription: [what each plan includes](https://paperswithbacktest.com/pricing). The same subscription covers every dataset in this organisation and the strategy catalogue. |
| |
| ## Elsewhere |
| |
| - [The strategy catalogue](https://paperswithbacktest.com/strategies) |
| - [`pwb-toolbox`](https://github.com/paperswithbacktest/pwb-toolbox), the loader used in the snippet above |
| - [Every dataset in this organisation](https://huggingface.co/datasets/paperswithbacktest) |
| |
