Forecasting is the act of making predictions. It is a critical element in decision making and serves the purpose of reducing the risk in decision making to reduce unexpected cost. The time series models are the best techniques to use when calculating fuel prices. According to Beasley (2010), time series can be defined as a collection of data that is recorded over a certain time frame which can be weekly, monthly, and so on. A time series model has four components; the trend, seasonal variation, the cyclical variation, and the irregular variation. The trend includes sales, stock prices, as well as other economic patterns. Cyclical variation is the business cycle that comprises of periods of boom followed by recession periods and then recovery. Also, production and sales fluctuate with seasons and finally, irregular variation entails the episodic and residual variations which are unpredictable. Therefore, since oil prices are constantly fluctuating and have characteristics that satisfy the four components, the time series model is appropriate in calculating oil prices through the use of moving averages.
A causal model comprises of a set of mathematical equations and or of a graph that lays down the hypothesized causal structure pictorially. Causal models that are predictive usually relate dependent variable to independent variables (Martin, 2011). Independent variables are those that can be manipulated and an example is force. By manipulating force applied on an object, the velocity of the object can be measured. Therefore, types of independent variables in causal modeling can include interest rates, sales volumes, exchange rates and demand. For example, an oil company that plans to expand its network of modern self-service stations, the independent variable can be the traffic flow which is represented by the number of cars using a service station per hour. As such, causal models can also be used in calculating oil prices.
Beasley, J. E. (2010) Forecasting. Retrieved on 06 Mar. 2014 from http://people.brunel.ac.uk/~mastjjb/jeb/or/forecast.html
Martin, S. (2011) Predictive Causal Models. Retrieved on 06 Mar. 2014 from http://www.cs.arizona.edu/people/rts/ergalics/predictivecausal.html