Gaussian Process Based Trading Strategy Identification
39 Pages Posted: 4 May 2012 Last revised: 4 Dec 2012
Date Written: November 23, 2012
The advent of electronic financial markets and associated technologies has dramatically improved the trading functions that are available to market participants in terms of speed, capacity and sophistication. Advanced data feed and audit trail information from market operators also make the full observation of market participants’ behavior possible. The primary objective of this study is to model algorithmic trading behavior using Bayesian inference under the framework of inverse reinforcement learning (IRL). We model trader behavior as a Gaussian process in the reward space. With incomplete observations of different market participants, we aim to recover the optimal policies and the corresponding reward functions to explain trader behaviors under different circumstances. We show that algorithmic trading behavior can be accurately identified using the Gaussian Process Inverse Reinforcement Learning (GPIRL) algorithm developed by Qiao and Beling (2011), and that this algorithm is superior to the linear features maximization approach. Real market data experiments using the GPIRL model consistently give more than 95% trader identification accuracy using a classification method based on support vector machines (SVM). We also show that there is a clear connection between the existing summary statisticbased trader classification proposed by Kirilenko etc. (2011) and our behavior-based classification. To address the potential change in trading behavior over time, we propose a score-based classification approach to address variations of algorithmic trading behavior under different market conditions. We further argue that because our behavior-based identification is a better reflection of traders’ actions and value propositions under different market conditions than the summary statistic-based method, it is therefore more informative and robust than the summary statistic-based approach, and is well suited for discovering new behavior patterns of market participants.
Keywords: Inverse Reinforcement Learning, Gaussian Process, High Frequency Trading, Algorithmic Trading, Behavioral Finance, Markov Decision Process
JEL Classification: C11, C14, C13
Suggested Citation: Suggested Citation