David Kuenzel, along with Theo S.Eicher (University of Washington), ChrisPapageorgiou (International Monetary Fund), and Charis Christofides (International Monetary Fund), authored a new paper published in the International Journal of Forecasting entitled “Forecasts in Times of Crises.” In the paper, Kuenzel and his coauthors examine the International Monetary Fund’s (IMF) forecast accuracy of 29 key macroeconomic variables for countries in times of economic crises. In general, forecasts of the IMF add substantial informational value as they outperform naive forecast approaches. However, the papers also documents that there is room for improvement: two-thirds of the examined macroeconomic variables are forecast inefficiently, and six variables (growth of nominal GDP, public investment, private investment, the current account, net transfers, and government expenditures) exhibit significant forecast biases. These forecast biases and inefficiencies are mostly driven by low-income countries, perhaps reflecting larger shocks and lower data quality. Most importantly, errors in private consumption growth forecasts are the main driver of GDP growth forecast errors. The results can help to shed light on which macroeconomic variables require further attention by the IMF in designing future forecast models.
