USE OF ERRORS OF PREDICTION IN IMPROVING FORECAST ACCURACY: AN APPLICATION TO WOOL IN AUSTRALIA

Over the past 20 years, the RAE has produced commodity forecasts in BAE Trends in Australian Agricultural Commodities, Farm Costs and Farm Incomes and other commodity outlook publications. There have been several studies on the accuracy of the forecasts, either comparing RAE forecasts with naive forecasts or making simple calculations of biases and standard deviations. Such approaches are not sufficiently informative in that they do not give any indication of what has caused the inaccuracies, nor of the timing of significant changes in prediction accuracy. In this analysis, an attempt is made to improve upon this situation using some tools of time series analysis.