Forecasting Commodity Prices Using Long Short-Term Memory Neural Networks

TL;DR

This study explores the use of LSTM neural networks for commodity price forecasting, finding that combining machine learning and traditional models improves accuracy for cotton prices.

Contribution

It demonstrates the effectiveness of forecast averaging between LSTM and ARIMA models, highlighting a novel approach to enhance commodity price predictions.

Findings

Averaging forecasts from LSTM and ARIMA improves accuracy for cotton.

Machine learning models fit data well but do not outperform classical methods alone.

Forecast averaging does not improve oil price predictions.

Abstract

This paper applies a recurrent neural network (RNN) method to forecast cotton and oil prices. We show how these new tools from machine learning, particularly Long-Short Term Memory (LSTM) models, complement traditional methods. Our results show that machine learning methods fit reasonably well the data but do not outperform systematically classical methods such as Autoregressive Integrated Moving Average (ARIMA) models in terms of out of sample forecasts. However, averaging the forecasts from the two type of models provide better results compared to either method. Compared to the ARIMA and the LSTM, the Root Mean Squared Error (RMSE) of the average forecast was 0.21 and 21.49 percent lower respectively for cotton. For oil, the forecast averaging does not provide improvements in terms of RMSE. We suggest using a forecast averaging method and extending our analysis to a wide range of commodity prices.