AutoTrader: Multi-Agent Trading System

Copyright (C) 2024-2025 Chris R. (iAmGiG) | Licensed under AGPL-3.0 | See NOTICE

Powered by AgentEdge - AI-assisted trading with human oversight

---

Disclaimer

THIS SOFTWARE IS FOR EDUCATIONAL AND RESEARCH PURPOSES ONLY.

• Not Financial Advice: This system does not provide financial, investment, or trading advice
• Use at Your Own Risk: Trading involves substantial risk of loss
• No Warranties: Provided "as-is" without guarantees of accuracy or profitability
• Past Performance: Walk-forward validated results do not guarantee future performance

---

Overview

AutoTrader is a research-driven trading platform featuring a multi-agent AI architecture powered by Microsoft AutoGen. The system implements walk-forward validated strategies with rigorous statistical testing and human oversight for paper/live trading via Alpaca Markets.

Core Philosophy: Validated momentum + regime filtering + human decision making > curve-fit technical indicators

Key Research Findings (Jan 2026)

Recent rigorous validation experiments with corrected methodology:

• GEX Regime Filtering: Research contradiction identified - MACD+RSI and academic TSMOM behave differently in GEX regimes. Phase 3B on hold pending resolution. See project status.
• Transaction Cost Robustness: MACD+RSI shows 44% pass rate vs TSMOM's 19% despite 10x higher turnover (#519)
• MACD Parameter Stability: No robustly profitable MACD configs found out-of-sample. Best OOS Sharpe: -0.223 (#518)
• Simulation Fidelity Gap: Path-dependent backtesting engine created to address "wick risk" (#528)

Strategy Validation Results

Issue	Strategy	Finding	Status
#516	TSMOM+GEX Hybrid	Median improvement: -2.9% (worse than baseline)	CLOSED
#518	MACD Stability	OOS Sharpe: -0.223 (least unprofitable)	CLOSED
#519	Transaction Costs	MACD+RSI: 44% pass rate, TSMOM: 19%	CLOSED

Methodology: Walk-forward validation, turnover-proportional transaction costs, look-ahead bias protection, median reporting for outlier robustness.

Note: Past performance does not guarantee future results. Research is ongoing with focus on simulation fidelity improvements.

Production Status

Component	Status
VoterAgent	Production Ready
CLI Trade Assistant	Complete
Alpaca Integration	Operational
Position Management	Complete
Trading Cycle	Complete
SQLite Cache	Complete (90%+ hit rate)
LLM Intent Routing	Complete
ScannerAgent	In Development
RiskAgent	In Development
ExecutorAgent	In Development

---

Core Agents

VoterAgent (Production Ready)

Location: src/autogen_agents/voter_agent.py

• MACD Signal Generation: Optimized 13/34/8 Fibonacci parameters
• RSI Momentum Analysis: 14-period with 30/70 thresholds
• Consensus Voting: Strong signals when both indicators agree
• Validated Performance: 0.856 Sharpe ratio over 2024-2025

LLM-Based Intent Classification

• Natural Language Commands: GPT-4o-mini parses trade requests
• Context-Aware Routing: Distinguishes "any open orders?" from ticker "ANY"
• Scalable Design: No hardcoded keyword patterns for ticker disambiguation

Agents in Development

• ScannerAgent: Market opportunity identification
• RiskAgent: Position sizing and risk management
• ExecutorAgent: Trade execution via Alpaca API
• TradingOrchestrator: Multi-agent coordination

---

Installation

# Python 3.12+ required
conda create -n AutoTrader python=3.12
conda activate AutoTrader
pip install -e .

Configuration

Create config/config.yaml with API credentials:

POLYGON_IO: "your_key"           # Required: Market data
ALPHA_VANTAGE_KEY: "your_key"    # Required: Fallback data
ALPACA_PAPER_API_KEY: "your_key" # Required: Paper trading
ALPACA_PAPER_SECRET: "your_key"  # Required: Paper trading
ALPACA_ENDPOINT: "https://paper-api.alpaca.markets/v2"

# Optional - VoterAgent uses pure math, no LLM required
OPEN_AI_KEY: "sk-..."

---

Quick Start

Interactive CLI

python main.py

# Example session:
> buy 10 AAPL              # Execute trade
> show portfolio           # View positions
> check my alerts          # Position alerts
> /schedule                # Scheduler management
> /help                    # All commands

Daemon Mode

python main.py --daemon

# Runs twice daily:
# - Morning: 9:20 AM ET (reconciliation)
# - Evening: 3:50 PM ET (review)

---

CLI Commands

Trading

Command	Description
buy SYMBOL QTY	Place buy order
sell SYMBOL QTY	Place sell order
cancel ORDER_ID	Cancel order

Information

Command	Description
show portfolio	Portfolio summary
show positions	Open positions with P&L
show orders	Order history
show account	Account details

Configuration

Command	Description
show timeframe	Current timeframe
set timeframe 1d	Change timeframe
show config-file	View YAML config
show watchlist	Scanner symbols

Workflow

Command	Description
morning-routine	Morning scan and analysis
evening-summary	End-of-day report
monitor	Watch positions for exits
forward-test start NAME	Start validation test

---

Project Structure

AutoTrader-AgentEdge/
├── src/
│   ├── autogen_agents/     # AI agents (AutoGen framework)
│   │   ├── voter_agent.py  # Production MACD+RSI voting
│   │   ├── base_agent.py   # Base agent class
│   │   └── ...             # Other agents (in development)
│   ├── trading/            # Trading infrastructure
│   │   ├── broker/         # Alpaca integration
│   │   ├── orders/         # Order management
│   │   ├── positions/      # Position tracking
│   │   └── scheduling/     # Daily routines
│   ├── data_sources/       # Market data
│   └── cli/                # CLI tools
├── config/                 # API credentials (local only)
├── config_defaults/        # Default YAML configs
├── docs/                   # Documentation
├── reports/                # Trading reports
└── tests/                  # Test suite

---

Documentation

Topic	Location
Architecture	docs/02_architecture/
Development	docs/04_development/
Design Decisions	docs/05_decisions/
Features	docs/06_features/
Testing	docs/07_testing/

---

Development

Testing

# Unit tests
python -m pytest tests/ -v

# Code quality
ruff check src/
black --check src/

# VoterAgent validation
python -c "from src.autogen_agents.voter_agent import VoterAgent; print('OK')"

Contributing

1. Pick an issue from GitHub Issues
2. Create a feature branch
3. Make changes following docs/05_decisions/
4. Submit PR to development branch

---

License

AGPL-3.0 - See LICENSE file for details.