Towards a fully RL-based Market Simulator

TL;DR

This paper introduces a novel RL-based market simulation framework where liquidity providers and takers learn concurrently, enabling realistic modeling of complex market behaviors for financial analysis.

Contribution

It proposes a fully RL-driven market simulator with shared policies for different agent types, advancing the realism and flexibility of financial market modeling.

Findings

Agents learn generalized policies through Deep RL

The simulator replicates complex market dynamics

Supports studying various financial scenarios

Abstract

We present a new financial framework where two families of RL-based agents representing the Liquidity Providers and Liquidity Takers learn simultaneously to satisfy their objective. Thanks to a parametrized reward formulation and the use of Deep RL, each group learns a shared policy able to generalize and interpolate over a wide range of behaviors. This is a step towards a fully RL-based market simulator replicating complex market conditions particularly suited to study the dynamics of the financial market under various scenarios.