Top Economic Forecaster Awards Methodology

Awards Methodology

Methodology

The Met the why particular Analyst Forecast Awards are yearly awards that recognize the analysts whose forecasts for the period are the most accurate. We assess the accuracy of the forecasts submitted to our Consensus Forecast survey by ranking forecasters according to the average of their discounted forecast errors. Errors are defined as the absolute difference between an individual forecast and the actual result of the macroeconomic indicator/price of the commodity.

To identify the best forecasters, we use forecasts submitted to our monthly survey over the course of 24 months. For example, in assessing the forecast accuracy for 2018 GDP, we take into consideration forecasts collected in the January 2017–December 2018 period. The outturn we use corresponds to the first official estimate, the release of which varies according to the frequency of the indicator.  

The absolute error (absolute difference between the individual forecast and the outturn of the indicator) is then discounted so that an error at the beginning of our forecasting cycle (January 2017) has less weight than an error at the end of the cycle (December 2018). This ensures that the forecasters are punished less for errors made at the beginning of the period when less information is available and hence forecasting is more difficult. Finally, panelists are ranked according to the average of the discounted absolute errors over the last 24 months. The highest ranked analyst is the one with the lowest average discounted error.   

To be considered for the award, the panelist must have submitted a minimum number of forecasts over the course of the period. We calculate forecast accuracy rankings for six main macroeconomic indicators (GDP, Fiscal Balance, Inflation, Interest Rate, Exchange Rate and Current Account) in 92 countries.

In addition to giving awards for each specific macroeconomic indicator mentioned above, we also provide a general ranking of the forecasts for the country, which is estimated by calculating weighted average ranking across all the indicators with GDP and inflation having slightly greater weight than the remaining indicators.

A note on the methodology for commodities:

To identify the best forecasters of commodity prices, we use price forecasts submitted to our monthly survey over the course of 24 months compared to the resulting average-of-year commodity price.  

The absolute error (absolute difference between the individual forecast and the outturn of the price) is then discounted so that an error at the beginning of our forecasting cycle (January 2017) has less weight than an error at the end of the cycle (December 2018). This ensures that the forecasters are punished less for errors made at the beginning of the period when less information is available and hence forecasting is more difficult. Finally, panelists are ranked according to the average of the discounted absolute errors over the last 24 months. The highest ranked analyst is the one with the lowest average discounted error.  

FocusEconomics Video

Improve your economic forecasting. This 1-minute video shows you how.

Play Video

Search form