Abstract
Private forecasters' inflation forecasts are one of the important indicators examined by the Bank of Israel in its conduct of monetary policy. An assessment of the quality of those forecasts and finding the combination of forecasts that provides maximum information will help to improve the reliability of the data used in the decision-making process. This study reviews the methods usually employed to assess and grade inflation forecasts, and applies them to those of the forecasters in Israel. It also examines whether different combinations of forecasts yield better results than those of any individual forecaster.
In the paper the forecasters are graded by quality, grouped according to the horizon of the forecast. The correlation of the grades between the groups was very low, and in some cases even negative indicating that forecasters specialize in different horizons. The study did not find any forecaster whose forecasts scored a high grade in all the horizons examined. It was also found that certain combinations of forecasters predict the CPIs more successfully than using the simple average of the forecasts, i.e., they reduced the mean squared deviation, but these forecasts were not significantly better than the simple average. The author therefore recommends continued use of the simple average of all the forecasters as an indicator of future inflation.