Profitability of Using Forecasting Techniques in the Commodities Market

Abstract

Box and Jenkins' Autoregressive Integrated Moving Average (ARIMA) forecasts for commodity prices one year into the future are compared to the futures market for accuracy. The ARIMA forecasts were nearly as accurate as the futures prices for predicting commodity prices. On the average, the futures market's Mean Absolute Percentage Error (MAPE) was approximately one percent less than that of the ARIMA models. By incorporating the ARIMA forecasts with the futures prices, it was concluded that a more profitable strategy for purchasing commodities could be obtained .This thesis showed that an average percentage reduction in purchasing costs of approximately twenty percent resulted when using the policy of buying commodities through futures only when the futures price was less than the ARIMA forecast price.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Dec 01, 1985
Accession Number
ADA165144

Entities

People

  • Dale M. Nees

Organizations

  • Naval Postgraduate School

Tags

DTIC Thesaurus Topics

  • Accuracy
  • Autocorrelation
  • Business Administration
  • Commerce
  • Commodities
  • Data Science
  • Efficiency
  • Information Science
  • Literature Surveys
  • Mainframe Computers
  • Random Walk
  • Schools
  • Security
  • Statistical Analysis
  • Time Series Analysis
  • United States
  • United States Naval Academy

Readers

  • Computational Modeling and Simulation
  • Economics
  • Statistical inference.