Abstract
Understanding and efficiently estimating construction material prices is fundamental to the cost formation of an effective budget for public works. Early inefficiencies in evaluation can quickly develop to impact project risk and significantly impact on project programme and procurement management processes throughout the subsequent construction period. The market prices of materials are subject to regular surveys which give useful benchmark data, however there are time-lags in the production and release of survey data and impacts of unforeseen international (e.g. oil price fluctuations) or local events which can cause significant deviations from tender or budget forecasts and expectations.
In this paper, we evaluate actual construction material price data accumulated over the last five years, and examine the deviation of the moving average as a potential modelling tool for the prediction of prices within the near future. We validate the model within a relative average error using a multiple regression technique where price prediction is a purpose function that accounts for the current price, market tendency, and the deviation of the moving average. Finally we consider and discuss the likely effectiveness of our proposed price prediction model in reducing the likelihood of tender failures and other impacts on public works projects.
