Manipulation of the Bitcoin market: an agent-based study

Area of Science:

• Computational Finance
• Market Microstructure Analysis
• Agent-Based Modeling

Background:

• Market manipulation poses systemic risks and erodes trust.
• Previous studies highlight the need for effective countermeasures against fraudulent trading.

Purpose of the Study:

• To design an agent-based model simulating Bitcoin market behavior during a 2017-2018 price manipulation event.
• To validate the model's ability to reproduce historical Bitcoin price, volume, and fraudulent agent activity.
• To analyze the impact of a fraudulent agent on market dynamics and price development.

Main Methods:

• Development of an agent-based model incorporating a limit order book and diverse trading strategies.
• Initialization of a fraudulent agent with empirical data to simulate market manipulation.
• Validation of simulation results against historical Bitcoin price, traded volume, and agent holdings.

Main Results:

• The agent-based model successfully reproduced Bitcoin market behavior during the specified period.
• Simulations explained observed price dips and volume anomalies attributed to the fraudulent trader.
• The fraudulent agent's presence was found critical for the late 2017 Bitcoin price increase.

Conclusions:

• The study validates the significant impact of market manipulation on cryptocurrency prices.
• Agent-based modeling provides a powerful tool for understanding and potentially mitigating illicit trading activities.
• Further research should explore the relationship between market liquidity and the efficiency of manipulation.