Abstract
This chapter compares three forecasts of short-term oil prices using two compumetric methods and naive random walk. Forecasting oil prices remains an important empirical issue. Compumetric methods use model specifications generated by computers with limited human intervention. Users are responsible only for selecting the appropriate set of explanatory variables. The compumetric methods employed here are genetic programming and artificial neural networks. The variable to forecast is monthly US imports FOB oil prices. The resulting equations identified those variables most responsible for price changes. In the one-step-ahead model GP produced, variables, such as world crude production, the exchange rate, US crude production, OECD consumption, production and oil stocks, as well as OPEC production play a significant role along with lagged price changes. In the three-step-ahead model, similar variables including WP and US crude production, changes in OECD, and lagged prices rather than changes in prices play a significant role in determining future prices of crude. Each method is used to forecast one and three months ahead. The results suggest that neural networks deliver better predictions.
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
