Forecasting Wheat Futures with Convolutional Neural Networks

Abstract

In this paper, we utilize a convolutional neural network to analyze aerial images of winter hard red wheat planted areas and cloud coverages over the planted areas as a proxy for future yield forecasts. We trained the model to forecast the futures price 20 days out and carry out recommendations for either a long or short position on wheat futures. Our method shows that achieving positive alpha within a short time window is possible if the algorithm and data choice are unique. However, the model’s performance can deteriorate quickly if the input data becomes more easily available as the strategy becomes crowded, as is the case with the aerial imagery we utilized in this paper.

Keywords: Machine Learning, Convolutional Neural Network, Futures Price Forecasting, Commodity Futures

JEL Classification: G11, G13, G17