Can leading indicators assess country vulnerability? Evidence from the 2008-09 global financial crisis

• 成果类型: Article
• 署名作者: Frankel, Jeffrey; Saravelos, George
• 署名单位: Harvard University; Deutsche Bank
• 刊物名称: JOURNAL OF INTERNATIONAL ECONOMICS
• ISSN/ISSBN: 0022-1996
• DOI: 10.1016/j.jinteco.2011.12.009
• 发表日期: 2012
• 页码: 216-231
• 关键词: crisis; Early warning; Emerging markets; leading indicators; reserves; 2008

摘要

We investigate whether leading indicators can help explain the cross-country incidence of the 2008-09 financial crisis. Rather than looking for indicators with specific relevance to the recent crisis, the selection of variables is driven by an extensive review of more than eighty papers from the previous literature on early warning indicators. Our motivation is to address suspicions that indicators found to be useful predictors in one round of crises are typically not useful to predict the next round. The review suggests that central bank reserves and past movements in the real exchange rate were the two leading indicators that had proven the most useful in explaining crisis incidence across different countries and episodes in the past. For the 2008-09 crisis, we use six different variables to measure crisis incidence: drops in GDP and industrial production, currency depreciation, stock market performance, reserve losses, and participation in an IMF program. We find that the level of reserves in 2007 appears as a consistent and statistically significant leading indicator of who got hit by the 2008-09 crisis, in line with the conclusions of the pre-2008 literature. In addition to reserves, recent real appreciation is a statistically significant predictor of devaluation and of a measure of exchange market pressure during the current crisis. We define the period of the global financial shock as running from late 2008 to early 2009, which probably explains why we find stronger results than earlier papers such as Obstfeld et al. (2009, 2010) and Rose and Spiegel (2009a,b, 2010, 2011) which use annual data. (C) 2012 Elsevier B.V. All rights reserved.