We develop an early-warning model of sovereign debt crises. A country is defined to be in a debt crisis if it is classified as being in default by Standard & Poor's, or if it has access to nonconcessional IMF financing in excess of 100 percent of quota. By means of logit and binary recursive tree analysis, we identify macroeconomic variables reflecting solvency and liquidity factors that predict a debt-crisis episode one year in advance. The logit model predicts 74 percent of all crises entries while sending few false alarms, and the recursive tree 89 percent while sending more false alarms.

This paper identifies leading indicators of fiscal crises based on a large sample of countries at different stages of development over 1970-2015. Our results are robust to different methodologies and sample periods. Previous literature on early warning sistems (EWS) for fiscal crises is scarce and based on small samples of advanced and emerging markets, raising doubts about the robustness of the results. Using a larger sample, our analysis shows that both nonfiscal (external and internal imbalances) and fiscal variables help predict crises among advanced and emerging economies. Our models performed well in out-of-sample forecasting and in predicting the most recent crises, a weakness of EWS in general. We also build EWS for low income countries, which had been overlooked in the literature.

We develop an early-warning model of sovereign debt crises. A country is defined to be in a debt crisis if it is classified as being in default by Standard amp; Poor's, or if it has access to nonconcessional IMF financing in excess of 100 percent of quota. By means of logit and binary recursive tree analysis, we identify macroeconomic variables reflecting solvency and liquidity factors that predict a debt-crisis episode one year in advance. The logit model predicts 74 percent of all crises entries while sending few false alarms, and the recursive tree 89 percent while sending more false alarms.

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.

A key objective of fiscal policy is to maintain the sustainability of public finances and avoid crises. Remarkably, there is very limited analysis on fiscal crises. This paper presents a new database of fiscal crises covering different country groups, including low-income developing countries (LIDCs) that have been mostly ignored in the past. Countries faced on average two crises since 1970, with the highest frequency in LIDCs and lowest in advanced economies. The data sheds some light on policies and economic dynamics around crises. LIDCs, which are usually seen as more vulnerable to shocks, appear to suffer the least in crisis periods. Surprisingly, advanced economies face greater turbulence (growth declines sharply in the first two years of the crisis), with half of them experiencing economic contractions. Fiscal policy is usually procyclical as countries curtail expenditure growth when economic activity weakens. We also find that the decline in economic growth is magnified if accompanied by a financial crisis.

Vol. 54, No. 2 includes three notable contributions from the Seventh Jacques Polak Annual Research Conference (ARC) hosted by the IMF in November 2006. Its lead paper, by Olivier Blanchard of Harvard University, is the 2006 Mundell-Fleming Lecture (delivered at the ARC), which analyzes current-account deficits in the advanced economies. Other papers in this issue look at the relationship between international financial integration and the real economy. Other papers discuss whether (or not): i) the next capital account crisis can be predicted; ii) accepted definitions of debt crises are adequate; iii) the Doha Round of trade talks (if they are ever successfully completed) will lead to preference erosion; and finally iv) there is room for political opportunism in countries deciding between money-based or exchange-rate-based stabilization programs.

In this paper, we consider incidences of arrears to the IMF, focusing on protracted arrears cases and attempt to identify determinants of their occurrence. We use narrative and formal methods. In addition, we analyze determinants of the duration of arrears. We find that previous arrears, reserves coverage, and institutional quality are among the main determinants of arrears. In addition, we identify a role for political developments, including civil unrest, which make arrears more likely to arise and to last longer. We conclude that improved macroeconomic conditions and turnaround of political fortunes would help to clear the currently remaining protracted arrears cases.

This book fills the normative gap arising from the absence of a multilateral mechanism for sovereign debt restructuring.

In response to the global crisis, the Fund overhauled its lending toolkit and boosted its resources, strengthening its ability to pre-empt financial crises. This paper—with the companion paper on Mapping Cross-Border Financial Linkages—takes another look at the recent global crisis in the context of a broader review of past systemic crises to (i) assess whether rising linkages across countries is a source of latent systemic instability and (ii) ascertain whether the global financial safety net (GFSN) is adequate to contain crisis and contagion risks arising from such systemic instability. This paper develops a new methodology to identify systemic crises and reviews associated policy responses from a global, rather than country-level, perspective