Comment

Abstract

Reinhart and Reinhart provide a systematic study of the episodes of influx of capital, namely a “capital flow bonanza.” The authors develop an algorithm, following the work of Kaminsky and Reinhart (1999), to date the incidence of bonanzas. This algorithm allows them not only to detect the smooth deterioration of the current account but also to analyze the macroeconomic developments surrounding the bonanzas. The study suggests a strong link between flows, global interest rates, and commodity prices. Using data from 181 countries from 1980 to 2007 and a core sample of 66 countries from 1960 to 2007, the authors show that (a) the path of the current account around bonanzas is V‐shaped, (b) bonanza periods are associated with a higher incidence of banking and currency crises in developing countries only, (c) bonanzas precede sovereign default episodes, and (d) fiscal policy is procylical around bonanzas. I think that this paper is an extremely valuable study for anyone who is interested in capital flows and sovereign debt. My main comments will be about the data issues, robustness, and the generalization of the results. Most of the analysis is done for 66 countries, during 1960–2007: 58 are middle‐ (emerging) and high‐ (industrialized) income and eight are low‐income (Africa) countries. Are there really data from 1960–2007 for all these countries? It would be nice to indicate which countries have how many years of data in appendix table A1. If a certain set of countries have more years of data than others, then these sets might be biasing the results. More important, most of the low‐income and some of the middle‐income countries are heavily indebted poor countries, which have received a lot of aid that shows up as capital inflows. Côte d’Ivoire, Bolivia, Honduras, and Mauritius are some examples. There are also some countries with debt forgiveness that will show up as capital outflows. I think that there needs to be a robustness exercise in which these countries are excluded. At the moment