Development of a forecasting methodology for cost indices: An application to the Greek construction industry

Forecasting is the process of estimating or predicting the future. The principle aim is to provide reliable estimates of future business that will assist management in making accurate decisions and sound plan for the future. Essentially, there are only two forms of approach to forecasting: Subjective, which incorporates the judgement of 'experts' (persons who work closely with the industry or product they are set to forecast). Objective, which apply a scientific process to the analysis of previous data using a statistical approach to create 'a fit' to the historic data then applying the 'model' to predict future occurrences. Forecasting can be considered to perform most effectively as a two-stage process with trends extracted from historical data and subjectively analysed by an expert predictor. There have been several attempts to see whether analytical mapping can indeed outperform the judgement of a so-called expert and vice versa. However, numerous publications have shown that combination methods are undoubtedly the most comprehensible. The prime objective of this paper is to introduce a forecasting methodology, which has been developed during the course of this research. The methodology combines analytical process of forecasting with judgmental input to tune and adjust forecasting figures. The methodology is being applied to cost indices in the Greek Construction Industry. It is concluded that such methodology can improve forecasts and provide managers with a vital tool to analyse historical data. The methodology is encapsulated in a form of a computer program.