A Forecasting Model for Futures Prices Based on Time Series Analysis: Dairy Commodities Data (303243)

Keywords: Stationary and Non-stationary Process, ARIMA with Regressors, Agricultural Economics

The price of dairy futures is crucial for dairy farmers who rely on hedging strategies to remain profitable during contractions in the commodities market. Several long-standing heuristics and predictions from both public and private sources are available to aid in the hedging process for Class III (hard and cream cheese production quality) and Class IV (non-fat dry quality) milk. We analyze publicly available data for efficacy in predicting future prices for various commodities using univariate autoregressive integrated moving average (ARIMA) models for stationary and non-stationary processes. In the analysis we use explanatory variables such as related commodities, financial products, and fiscal indices. The costs and benefits of ARIMA model are substantiated with advanced analytics. This poster will provide a review of the models used with and without exogenous regressors such as back-shifted prices through a broadly applicable commodity based example. Tables and graphs are used to demonstrate the results of the analysis.