Modelling financial volume curves with hierarchical Poisson processes

Creighton Heaukulani,Abhinav Pandey,Lancelot F. James
2024-06-01
Abstract:Modeling the trading volume curves of financial instruments throughout the day is of key interest in financial trading applications. Predictions of these so-called volume profiles guide trade execution strategies, for example, a common strategy is to trade a desired quantity across many orders in line with the expected volume curve throughout the day so as not to impact the price of the instrument. The volume curves (for each day) are naturally grouped by stock and can be further gathered into higher-level groupings, such as by industry. In order to model such admixtures of volume curves, we introduce a hierarchical Poisson process model for the intensity functions of admixtures of inhomogenous Poisson processes, which represent the trading times of the stock throughout the day. The model is based on the hierarchical Dirichlet process, and an efficient Markov Chain Monte Carlo (MCMC) algorithm is derived following the slice sampling framework for Bayesian nonparametric mixture models. We demonstrate the method on datasets of different stocks from the Trade and Quote repository maintained by Wharton Research Data Services, including the most liquid stock on the NASDAQ stock exchange, Apple, demonstrating the scalability of the approach.
Statistical Finance,Machine Learning
What problem does this paper attempt to address?
The paper is primarily dedicated to addressing the issue of modeling intraday trading volume curves in financial trading. Specifically, the research aims to predict these so-called "volume curves," which are crucial for guiding trade execution strategies. For example, a common strategy is to distribute the required buy and sell quantities according to the expected intraday volume curve to avoid impacting the stock price. To achieve this goal, the authors propose a hierarchical Poisson process model based on the Hierarchical Dirichlet Process (HDP) to describe the intensity functions of different stocks' intraday trading times. This model can capture the commonalities between different stocks as well as the similarities of a single stock across different trading days. Additionally, it allows for an infinite number of mixture components through Bayesian nonparametric methods. The paper also introduces how to use the Markov Chain Monte Carlo (MCMC) algorithm for posterior inference and demonstrates the model's effectiveness and scalability on different stock datasets from the Trade and Quote data warehouse of Wharton Research Data Services. In summary, the main contribution of this paper lies in proposing a novel hierarchical Poisson process model that can effectively simulate the intraday trading volume curves of financial instruments and validating the practical value of the model through empirical studies.