Deep Probabilistic Modelling of Price Movements for High-Frequency Trading

Abstract

In this paper we propose a deep recurrent architecture for the probabilistic modelling of high-frequency market prices, important for the risk management of automated trading systems. Our proposed architecture incorporates probabilistic mixture models into deep recurrent neural networks. The resulting deep mixture models simultaneously address several practical challenges important in the development of automated highfrequency trading strategies that were previously neglected in the literature: 1) probabilistic forecasting of the price movements; 2) single objective prediction of both the direction and size of the price movements. We train our models on high-frequency Bitcoin market data and evaluate them against benchmark models obtained from the literature. We show that our model outperforms the benchmark models in both a metric-based test and in a simulated trading scenario.

Learning Probability Distributions in Macroeconomics and Finance

“…Salinas et al (2020) build an autoregressive recurrent neural network, which learns mean and standard deviation for Gaussian, and mean and shape parameter for Negative binomial. Lim and Gorse (2020) classifies price movements for high-frequency trading via deep probabilistic modelling when optimizing parameters of different families of distribution. Although similarly to Salinas et al (2020), proposes to use a deep temporal convolutional neural network to estimate parameters of Gaussian distribution to model probabilistic forecast, and they further propose to use the same architecture for non-parametric estimation of quantile regression.…”