Event-Based Limit Order Book Simulation under a Neural Hawkes Process: Application in Market-Making
This paper proposes an event-driven LOB simulation with a Neural Hawkes process, modeling midprice dynamics for enhanced deep RL market-making.
What it examines
This paper designs an event-driven Limit Order Book model using a Neural Hawkes Process to capture 12 key LOB events. It simulates realistic high-frequency midprice dynamics, overcoming diffusion model limitations, and provides a foundation for improved algorithmic trading and market-making strategies.
What it concludes
The Neural Hawkes LOB model effectively simulates midprice dynamics for high-frequency trading, especially market-making. While it captures essential volatility features, further refinement is needed for finer details. This research applies to risk management, algorithmic trading, and market strategy optimization, with future work to enhance event differentiation.
Evidence objects
Researchers developed an innovative event-based simulation framework for Limit Order Books, employing a Neural Hawkes process integrated with a LSTM network to model twelve market events and capture trading interactions.
key_findings bullet 1 · key_findings · validation V0
Empirical analysis shows simulated order fills and price movements closely mimic actual market data with high volatility and fat tails, although subtle discrepancies in skewness and memory effects still remain.
key_findings bullet 2 · key_findings · validation V0
Researchers introduce new terminology and categorize market events through self-modulated multivariate Hawkes intensities. They leverage LOBSTER data and deep reinforcement learning for market-making, though dynamic details still require further refinement.
key_findings bullet 3 · key_findings · validation V0
This paper introduces an innovative event-driven simulation framework for Electronic Financial Markets by integrating a Neural Hawkes process with deep reinforcement learning to simulate market-making across $12$ LOB event types. Its approach and extension of diffusion models significantly enhance LOB modeling, offering a fresh perspective with practical high-frequency trading applications.
key_findings bullet 4 · key_findings · validation V0
Raw abstract and provenance
Abstract: In this paper, we propose an event-driven Limit Order Book (LOB) model that captures twelve of the most observed LOB events in exchange-based financial markets. To model these events, we propose using the state-of-the-art Neural Hawkes process, a more robust alternative to traditional Hawkes process models. More specifically, this model captures the dynamic relationships between different event ty… ▽ More In this paper, we propose an event-driven Limit Order Book (LOB) model that captures twelve of the most observed LOB events in exchange-based financial markets. To model these events, we propose using the state-of-the-art Neural Hawkes process, a more robust alternative to traditional Hawkes process models. More specifically, this model captures the dynamic relationships between different event types, particularly their long- and short-term interactions, using a Long Short-Term Memory neural network. Using this framework, we construct a midprice process that captures the event-driven behavior of the LOB by simulating high-frequency dynamics like how they appear in real financial markets. The empirical results show that our model captures many of the broader characteristics of the price fluctuations, particularly in terms of their overall volatility. We apply this LOB simulation model within a Deep Reinforcement Learning Market-Making framework, where the trading agent can now complete trade order fills in a manner that closely resembles real-market trade execution. Here, we also compare the results of the simulated model with those from real data, highlighting how the overall performance and the distribution of trade order fills closely align with the same analysis on real data. △ Less
Source row: 743 · abstract type: unknown