Deep Stock Predictions

TL;DR

This paper enhances stock price prediction using LSTM neural networks by customizing loss functions, optimizing window and prediction lengths, and incorporating analyst calls, leading to improved trading strategy performance.

Contribution

It introduces a novel approach combining customized loss functions, data-driven parameter selection, and technical indicators in LSTM models for better stock forecasting.

Findings

LSTM with customized loss outperforms ARIMA in training.

Adding analyst calls improves prediction accuracy for some datasets.

Optimized window and prediction lengths enhance model performance.

Abstract

Forecasting stock prices can be interpreted as a time series prediction problem, for which Long Short Term Memory (LSTM) neural networks are often used due to their architecture specifically built to solve such problems. In this paper, we consider the design of a trading strategy that performs portfolio optimization using the LSTM stock price prediction for four different companies. We then customize the loss function used to train the LSTM to increase the profit earned. Moreover, we propose a data driven approach for optimal selection of window length and multi-step prediction length, and consider the addition of analyst calls as technical indicators to a multi-stack Bidirectional LSTM strengthened by the addition of Attention units. We find the LSTM model with the customized loss function to have an improved performance in the training bot over a regressive baseline such as ARIMA, while the addition of analyst call does improve the performance for certain datasets.