Dynamic and Context-Dependent Stock Price Prediction Using Attention Modules and News Sentiment

TL;DR

This paper introduces the $oldsymbol{oldsymbol{ ext{α}}_{t}}$-RIM, a dynamic, attention-based neural network that models non-stationary stock prices and news sentiment to improve prediction accuracy in financial time series.

Contribution

The paper proposes the $oldsymbol{ ext{α}}_{t}$-RIM architecture, combining attention mechanisms and modular recurrent structures to better capture causal and dynamic relationships in financial data.

Findings

$oldsymbol{ ext{α}}_{t}$-RIM outperforms LSTM in stock price prediction.

The model effectively captures causal links between news sentiment and stock prices.

Improves generalization to unseen data in financial time series.

Abstract

The growth of machine-readable data in finance, such as alternative data, requires new modeling techniques that can handle non-stationary and non-parametric data. Due to the underlying causal dependence and the size and complexity of the data, we propose a new modeling approach for financial time series data, the $\alpha_{t}$-RIM (recurrent independent mechanism). This architecture makes use of key-value attention to integrate top-down and bottom-up information in a context-dependent and dynamic way. To model the data in such a dynamic manner, the $\alpha_{t}$-RIM utilizes an exponentially smoothed recurrent neural network, which can model non-stationary times series data, combined with a modular and independent recurrent structure. We apply our approach to the closing prices of three selected stocks of the S\&P 500 universe as well as their news sentiment score. The results suggest that the $\alpha_{t}$-RIM is capable of reflecting the causal structure between stock prices and news sentiment, as well as the seasonality and trends. Consequently, this modeling approach markedly improves the generalization performance, that is, the prediction of unseen data, and outperforms state-of-the-art networks such as long short-term memory models.