02245cas a2200277 a 4500 AALejournalIMF021644 230101c9999 xx r poo 0 0eng d 18.00 USD 9781513573588 1018-5941 BD-DhAAL BD-DhAAL Hellwig, Klaus-Peter. Predicting Fiscal Crises : A Machine Learning Approach / Klaus-Peter Hellwig. Washington, D.C. : International Monetary Fund, 2021. 1 online resource (66 pages) IMF Working Papers <strong>Off-Campus Access:</strong> No User ID or Password Required <strong>On-Campus Access:</strong> No User ID or Password Required Electronic access restricted to authorized BRAC University faculty, staff and students In this paper I assess the ability of econometric and machine learning techniques to predict fiscal crises out of sample. I show that the econometric approaches used in many policy applications cannot outperform a simple heuristic rule of thumb. Machine learning techniques (elastic net, random forest, gradient boosted trees) deliver significant improvements in accuracy. Performance of machine learning techniques improves further, particularly for developing countries, when I expand the set of potential predictors and make use of algorithmic selection techniques instead of relying on a small set of variables deemed important by the literature. There is considerable agreement across learning algorithms in the set of selected predictors: Results confirm the importance of external sector stock and flow variables found in the literature but also point to demographics and the quality of governance as important predictors of fiscal crises. Fiscal variables appear to have less predictive value, and public debt matters only to the extent that it is owed to external creditors. Mode of access: Internet Foreign Exchange imf Informal Economy imf Underground Econom imf IMF Working Papers; Working Paper ; No. 2021/150 Full text available on IMF http://elibrary.imf.org/view/journals/001/2021/150/001.2021.issue-150-en.xml IMF e-Library