AETERNUS: A Six-Module Synthetic Market Pipeline with Controlled Validation on Real Futures Data

Key Points

• The aim is to develop a synthetic market pipeline for simulating trading dynamics and validating its performance using real data.
• Implemented a limit order book simulation to replicate market behavior.
• Applied volatility modeling and correlation compression for enhanced data processing.
• Utilized network topology for organizing trading agents within the market.
• Employed transformer forecasting for predicting price movements.
• Validated the framework using real output from a physics engine.
• Successfully simulated market scenarios that reflect real-world trading dynamics.
• Achieved accurate forecasting of futures prices based on transformed data inputs.
• Validated the multiagent approach, enhancing decision-making in trading simulations.