Skip to content

SEC EDGAR Pipeline

Quarterly fundamental data — income statements, balance sheets, and cash flow statements — from SEC EDGAR filings via the edgartools library.

What it provides

  • Income statements: revenue, gross profit, operating income, net income, EPS
  • Balance sheets: assets, liabilities, equity, debt
  • Cash flow statements: operating, investing, financing cash flows
  • Computed ratios: Piotroski F-score, ROE, ROA, current ratio, debt/equity

Prerequisites

No credentials required. SEC EDGAR is a public database.

cd ~/Documents/Q-agent/infrastructure
bash setup.sh
source .venv/bin/activate

Running the pipeline

python infrastructure/pipelines/edgar/scripts/run_pipeline.py --tickers AAPL MSFT GOOGL

To run for the full 30-stock universe:

python infrastructure/pipelines/edgar/scripts/run_pipeline.py --universe wrds

Output schema

fundamentals_annual.csv: 

ticker,period,revenue,gross_profit,operating_income,net_income,total_assets,total_debt,equity
AAPL,2023-09-30,383285000000,169148000000,114301000000,96995000000,352583000000,109280000000,62146000000

piotroski.csv: 

ticker,period,f_score,roa,delta_roa,cfo,accrual,delta_leverage,delta_liquidity,equity_offer,delta_margin,delta_turnover
AAPL,2023-09-30,7,0.283,0.021,0.312,-0.029,-0.041,0.038,0,0.031,0.044

Using in a notebook

import pandas as pd

fundamentals = pd.read_csv(
    "infrastructure/pipelines/edgar/data/fundamentals_annual.csv",
    parse_dates=["period"]
)

# Filter to recent years
recent = fundamentals[fundamentals["period"] >= "2015-01-01"]

Notes

  • EDGAR data is filed quarterly. Annual data (10-K) is the most reliable for fundamental analysis.
  • edgartools fetches directly from SEC EDGAR's XBRL data. No third-party data vendor is involved.
  • Some tickers have inconsistent XBRL tagging across years. The pipeline normalises field names where possible.