Trading System

Algorithmic trading system for the Argentine market (CEDEARs). Combines data ingestion, technical analysis, strategy backtesting, signal generation, and portfolio management.

Quick Start

Prerequisites

• Docker and Docker Compose
• Python 3.12+
• Poetry

1. Clone and setup environment

git clone https://github.com/MauroPerna/ar-trading-bot
cd ar-trading-bot

# Copy environment file
cp .env.example .env

The default .env values work out of the box with docker-compose.

2. Start the database

docker-compose up -d

This starts:

• PostgreSQL 16 on port 5432
• Redis 7 on port 6379

3. Install dependencies

poetry install

4. Active the virtual environment

# if the plugin is not yet installed
poetry self add poetry-plugin-shell

poetry shell

5. Run database migrations (Optional)

poetry run alembic upgrade head

6. Start the application

poetry run python main.py

The API will be available at http://localhost:8000

6. Verify it's working

# Health check
curl http://localhost:8000/health

# API docs
open http://localhost:8000/docs

Project Structure

src/
  application/          # FastAPI app, DI container, lifecycle
  domain/
    etl/                # Data extraction, transformation, enrichment
    signals/            # Signal generation and aggregation
    strategies/         # Backtesting engine and strategy selection
    portfolio/          # Portfolio optimization (Markowitz, HRP)
    trading/            # Trade execution (FakeBroker for dev)
  infrastructure/
    broker/             # Broker clients (IOL, FakeBroker)
    data/               # Data providers (yfinance)
    database/           # SQLAlchemy models and repositories
    scheduler/          # APScheduler jobs
    config/             # Settings and configuration

Architecture

Signals Engine

Transforms OHLCV data into actionable signals:

• Technical indicators (pandas-ta): RSI, MACD, Bollinger Bands, etc.
• Custom analyzers: Breakout, Fair Value Gap, Trend
• Interpreters: Momentum, Structure, Volume, Volatility, Risk

Strategy Selection

Backtesting engine powered by backtesting.py:

1. Fetches historical data from yfinance
2. Runs signal engine
3. Evaluates multiple strategies
4. Ranks by Sharpe ratio, drawdown, win rate
5. Persists best strategy per symbol

Portfolio Optimization

Two optimizers available:

• Markowitz: Mean-variance optimization (max Sharpe)
• HRP: Hierarchical Risk Parity

Scheduled Jobs

• StrategyPerTickerJob: Monthly strategy re-evaluation
• GenerateSignalsJob: Hourly signal generation
• PortfolioWeightsJob: Monthly portfolio rebalancing

Development

Run tests

poetry run pytest

Environment modes

• development: Auto-creates default portfolio, enables hot reload
• production: Requires manual setup, no auto-seeding

Configuration

Key environment variables:

Variable	Description	Default
DB_URL	PostgreSQL connection string	-
ENVIRONMENT	development/production	development
LOG_LEVEL	DEBUG/INFO/WARNING/ERROR	INFO
SCHEDULER_TIMEZONE	Timezone for jobs	America/Argentina/Buenos_Aires
IOL_USERNAME	IOL broker username	-
IOL_PASSWORD	IOL broker password	-

Stopping the application

# Stop the API (Ctrl+C in the terminal)

# Stop Docker services
docker-compose down

# Stop and remove volumes (clean slate)
docker-compose down -v