Abstract
This paper explores the creation of a predictive model to forecast cocoa futures prices using a neural network. The model combines historical cocoa price data and real-time weather information to estimate the end-of-month price for cocoa futures. The goal is to develop a data-driven trading strategy that capitalizes on short-term price movements in the volatile cocoa market. Key inputs include weather data from the National Oceanic and Atmospheric Administration (NOAA) and cocoa futures prices from the Intercontinental Exchange (ICE). The model was tested and validated through a backtesting process to assess its effectiveness in predicting price trends and guiding trades. This research demonstrates the potential of using machine learning to improve trading strategies in commodity markets like cocoa.