External Balance Assessment (EBA): Data and Estimates
Last Updated: July 2024
The IMF's Research Department has developed the External Balance Assessment (EBA) methodology for assessing current accounts and exchange rates. The EBA methodology, originally developed in 2013, was modified in 2015 by extending the sample used in estimation (from 2010 to 2013), allowing nonlinearities in the impact of demographics, and adding the REER level model to the EBA toolkit. A more comprehensive review of the methodology took place in 2018, with the changes focused on (i) further extending the sample used in estimation (from 2013 to 2016); (ii) improving the modeling of certain fundamentals—demographics, the measurement of external positions, and institutional quality—and macroeconomic policies—foreign exchange intervention and credit excesses; and (iii) developing complementary tools to inform the role that structural and trade policies could play in explaining current account imbalances. A further review of the methodology took place in 2022, with the changes focused on (i) extending the sample used in estimation (until 2019, and including three additional countries); (ii) improving the construction of the terms-of-trade and oil and gas reserves variables, (iii) including an IMF staff-produced measure of capital controls; and (iv) simplifying the regression models by only keeping current account and exchange rate determinants that are statistically robust. A description of the latest generation of the EBA models can be found in the IMF Working Paper No. 2023/047, issued on March 3, 2023.
The following are available for public use:
- EBA model results: Estimates from the 2024 EBA exercise (analysis of 2023 current accounts and real effective exchange rates). Estimates of earlier exercises can also be found (2022 EBA, 2021 EBA, 2020 EBA, 2019 EBA, 2018 EBA, 2017 EBA, 2016 EBA, 2015 EBA, 2014 EBA).
- Data inputs and codes: Data inputs and programs used in the analysis of 2023 current accounts and (index) real effective exchange rates and level real effective exchange rates (including coefficient estimates and graphical comparisons of actual and regression-implied values). Data inputs for earlier vintages can be made available upon request.
Disclaimer: A number of caveats apply to the EBA data inputs. The EBA exercise drew on data and estimates that were available ahead of publication, some of which may since have become outdated. Some data may have different definitions or coverage than IMF or IMF member country statistics, and some data definitions may not be fully comparable across countries. Some data, including for capital controls and foreign exchange intervention policies, are imperfect proxy measures. For that reason, the numerical values used in the exercise to indicate desirable policy settings may be benchmarks that do not always fully represent the views of IMF staff. The information and estimates available here should not be considered as representing the assessments or views of the IMF or its Executive Board.