On Modeling Investors in Simulated Markets

Explore the intricacies of modeling investors in simulated markets through this 27-minute conference talk by Svitlana Vyetrenko from JP Morgan AI Research. Delve into the importance of market simulation, comparing market replay with multi-agent simulation techniques. Examine the ABIDES (Agent Based Interactive Discrete Event Simulation) framework and its applications in trading and market mechanism design. Learn about the development of trading agents, including rule-based agents and those learned from historical data. Understand the market impact of algorithmic execution and the calibration of multi-agent simulators. Discover the use of reinforcement learning in creating trading agents and gain insights into modeling human traders. This talk, part of the "Workshop on Machine Learning for Investor Modelling" at the Fields Institute, offers valuable knowledge for those interested in financial modeling and market simulation techniques.

Outline
Why market simulation?
Market replay vs. multi-agent simulation
Example: Market Impact of Algorithmic Execution
ABIDES* - Agent Based Interactive Discrete Event Simulation
Applications to trading
Market Mechanism Design
How to design trading agents?
ABIDES Rule-Based Agents
Market Impact of Algorithmic Execution Explained
Agents learnt from historical data
Calibration of Multi-Agent Simulators
RL to Learn Trading Agents
Modeling Human Traders