Cassandra and the Sirens: Economic Forecasting in Emerging Economies

Economic forecasts in emerging economies share many of the characteristics of those for advanced economies. There is little sign of consistent biases but turning points are almost invariably missed, individual forecasts tend to converge with each other and become more accurate as the forecasting horizon reduces, the consensus forecasts for GDP and inflation are significantly better than simple alternative forecasting rules, it is rare for individual forecasters to consistently outperform the consensus, and the IMF and private sector forecasts tend to be of similar quality. A key difference, however, is that average forecasting errors are larger in emerging economies. This is because the task is inherently harder in emerging economies, given their greater exposure to external shocks and poorer data. A policy implication is that emerging economy central banks setting inflation targets should be less ambitious than central banks in advanced economies.