An Examination of the Accuracy of Economic Country Forecasts produced by the Economist Intelligence Unit (EIU)

• Main Author: Falconer, Nick
• Other Authors: Van Walbeek, Cornelis
• Format: Dissertation
• Language: English
• Published: University of Cape Town 2019
• Subjects: Economics
• Online Access: http://hdl.handle.net/11427/29631

This paper investigates the accuracy of Country-Level Forecasts for six variables: Real GDP Growth, Inflation (Consumer Price Index), Short-Term Interbank Rate, Budget Deficit (as % of GDP), Current Account Balance (as % of GDP) and the Unemployment rate. The data used in this analysis is produced by the Economist Intelligence Unit (EIU) and covers the period 1996-2013. The two forecast horizons of primary interest are Current-Year (t) and One-Year-Ahead (t+1) forecasts. A range of statistical methods are used to ascertain the accuracy of the EIU forecasts in 103 countries for most of the variables, while the accuracy of the Unemployment rate forecasts used data for 68 countries. Generally, the results confirm this paper’s four primary hypotheses: (1) EIU forecasts are more accurate than No-Change-Forecasts; (2) EIU forecasts for higher income countries have smaller errors than forecasts for lower income countries; (3) The accuracy of EIU forecasts decreases as forecast horizon increases; and (4) The accuracy of EIU forecasts was relatively weaker during the period of the Global Financial Crisis (2008-2009) than in other periods.