Automated Trading with a Compound Interest Strategy
⚠️ DISCLAIMER I am a programmer, not a licensed financial advisor. Always consult a certified professional for investment decisions. Do not rely solely on anonymous sources or this software—conduct your own research.
- 🛠️ This project is under active development and may change frequently.
⚠️ Use at your own risk.- 🚨 Always audit the output.
CoinBot is a modular framework for automated trading using averaging strategies:
- Cost Averaging — Buy fixed amounts at regular intervals.
- Dynamic Cost Averaging — Buy/sell/hold based on conditions around a fixed target.
- Value Averaging — Buy/sell/hold based on a moving dynamic target.
A future version will include a time-series prediction model based on GPT architecture in PyTorch. This will only be developed once the core interface is stable.
For now, CoinBot only supports the Coinbase Advanced Trade API to keep the scope focused. Alpaca (for stock trading) may return later as a module if interest and resources allow.
This tool is a component of a larger effort to:
- Build a private treasury for long-term funding of open-source and experimental projects.
- Develop emotionless, strategy-driven models that operate consistently regardless of market sentiment.
- Integrate real-time and historical data to train models on trend-following, momentum, and mean-reversion patterns.
- Eventually output tax-ready documents such as Form 8949 (CSV format) for gains/loss tracking.
git clone https://github.com/teleprint-me/coinbot coinbot
cd coinbot
python -m venv .venv
source .venv/bin/activate
pip install -r requirements.txt
Optional (for model training, later):
pip install torch --index-url https://download.pytorch.org/whl/cpu
Create a .env
file:
touch .env
vim .env
Add your keys:
COINBASE_API_KEY='your-api-key'
COINBASE_API_SECRET='your-api-secret'
python -m coinbot.coinbase
Licensed under the GNU Affero General Public License (AGPL):
- You are free to use, modify, and share under AGPL terms.
- Redistribution requires keeping this license.
- See https://www.gnu.org/licenses/ for details.