Trading TDA

Trading TDA is an early ASB Solutions Group trading application built against the TD Ameritrade API — a predecessor of my current IBKR/Alpaca automation stack. It pulls portfolio and market data through tda-api, lands it in Google BigQuery, and runs time-series forecasting over it.

The concepts

The pipeline has three stages:

• Authenticate and ingest — OAuth against the TDA API (authenticate.py), then import-portfolio.py exports portfolio data as per-ticker CSVs
• Warehouse — the CSVs load into a BigQuery trading.trading_data table with bq load --autodetect, giving SQL analytics over the whole portfolio history
• Forecast — arima.py fits ARIMA models (via statsmodels) across the portfolio, with a debug mode and per-portfolio selection, plus pandas_ta and matplotlib for indicators and charts

How to run it

Configuration lives in a local setup.py with the TDA API key, redirect URI, and a token path — the README is emphatic that the OAuth token file stays private. A startup script creates a virtualenv, sets SETUPTOOLS_USE_DISTUTILS=stdlib (needed at the time for the dependency chain), installs requirements, and assumes gcloud and bq are configured:

virtualenv my-venv && source my-venv/bin/activate
pip install -r Requirements.txt
python authenticate.py
python import-portfolio.py
python arima.py y ALL

Worth noting historically: the TD Ameritrade API this was built on has since been retired following the Schwab migration, so the repo stands as an archived snapshot of that era — the ideas (broker API → warehouse → forecast) carried forward into my later trading projects.

Language: Python Source code: github.com/ASB-Solutions-Group-Inc/trading-tda