Analyzing Commodity Market Volatility and Price Forecasting: A GARCH and ARIMA Model Approach

Abstract

Commodity trade is a cornerstone of world financial markets, providing investment opportunities, risk management, and price discovery. As commodities are inherently volatile, understanding their price fluctuations and forecasting future trends is essential. This study examines the performance and volatility of four widely traded commodities in the United States - Gold, Silver, Wheat, and Crude Oil using the Generalized Autoregressive Conditional Heteroskedasticity (GARCH) model to measure volatility and the Autoregressive Integrated Moving Average (ARIMA) model to predict future price trends. The GARCH model effectively captures volatility clustering, a key characteristic of financial time series data, while ARIMA analyzes historical patterns for price prediction. Using a decade's worth of daily historical price data from secondary sources, this research provides a robust dataset for in-depth analysis. Additionally, this study highlights the need for advanced predictive models that enhance accuracy during market fluctuations. By analyzing GARCH and ARIMA applications in commodity trading, this research contributes to financial modeling and risk management literature, encouraging further exploration of alternative forecasting methods.