Cross-Country Causes and Consequences of the 2008 Crisis: Early Warning Andrew K. Rose and Mark M. Spiegel* Comments Welcome Revised as of: October 5, 2011 Abstract This paper models the causes of the 2008 financial crisis together with its manifestations, using a Multiple Indicator Multiple Cause (MIMIC) model. Our analysis is conducted on a cross-section of 107 countries; we focus on national causes and consequences of the crisis, ignoring cross-country “contagion” effects. Our model of the incidence of the crisis combines 2008 changes in real GDP, the stock market, country credit ratings, and the exchange rate. We explore the linkages between these manifestations of the crisis and a number of its possible causes from 2006 and earlier. We include over sixty potential causes of the crisis, covering such categories as: financial system policies and conditions; asset price appreciation in real estate and equity markets; international imbalances and foreign reserve adequacy; macroeconomic policies; and institutional and geographic features. Despite the fact that we use a wide number of possible causes in a flexible statistical framework, we are unable to link most of the commonly-cited causes of the crisis to its incidence across countries. This negative finding in the cross-section makes us skeptical of the accuracy of “early warning” systems of potential crises, which must also predict their timing. Keywords: empirical; data; cross-section; crisis; credit; stock; country; model; international; MIMIC model; early warning models JEL Classification Numbers: E65, F30 Andrew K. Rose Mark M. Spiegel Haas School of Business FRB San Francisco University of California 101 Market St. Berkeley, CA USA 94720-1900 San Francisco CA 94105 Tel: (510) 642-6609 Tel: (415) 974-3241 Fax: (510) 642-4700 Fax: (415) 974-2168 E-mail: [email protected] E-mail: [email protected] * Rose is B.T. Rocca Jr. Professor of International Trade and Economic Analysis and Policy in the Haas School of Business at the University of California, Berkeley, NBER Research Associate and CEPR Research Fellow. Spiegel is Vice President, Economic Research, Federal Reserve Bank of San Francisco. We thank Haibin Zhu for sharing his real estate data. Helpful comments were received from: an anonymous referee, Joshua Aizenman, David Cook, Mike Dooley, Marcel Fratzscher, Sophia Rabe-Hesketh, Kadee Russ, and seminar participants at the BIS and the APEA. Rose thanks the Federal Reserve Bank of San Francisco for hospitality during the course of this research. Christopher Candelaria provided excellent research assistance. The views expressed below do not represent those of the Federal Reserve Bank of San Francisco or the Board of Governors of the Federal Reserve System. A current version of this paper, key output, and the main STATA data set used in the paper are available at http://faculty.haas.berkeley.edu/arose. This paper began to circulate in June 2009, on the basis of data collected mostly by March 2009. It was subsequently criticized explicitly or implicitly by a number of authors, including among others: Blanchard, et al (2010); Claessens et al (2010); Giannone et al (2011); and Lane, and Milesi-Ferretti (2011). Our response, using an updated data set, is provided in Rose and Spiegel (2011).1 “we agree … that the FSB [Financial Stability Board] should collaborate with the IMF to provide early warning of macroeconomic and financial risks and the actions needed to address them” - Final Communiqué G-20 Summit April 2, 20091 “Any early warning system to detect impending dangers to the world economy must find a way of bringing together the scatter of international and national macrofinancial expertise. We at the Fund have already begun intensifying our early warning capabilities and will be strengthening our collaboration with others involved in this area.” - Dominique Strauss-Kahn2 I: Motivation The 2008 global financial crisis is notable for a number of reasons, including most obviously its severity and speed. The international span of the crisis has also been remarkable; essentially all the industrialized countries have been affected, as well as a large number of developing and emerging economies. In this paper we seek to deepen our understanding of the international breadth of the crisis; we are particularly interested in modeling the causes of the crisis, and why its severity differs across countries. We are interested in understanding the causes of 2008 the crisis both out of intrinsic interest, and to investigate the feasibility of modeling financial crises like this empirically. Economists do not have a particularly good track record at predicting the timing of crises, which is one of the objectives of an early warning system.3 Historically however, the profession has had some success at modeling the incidence of crises across firms, banks, and/or countries.4 That is, we find cross-sectional analysis easier than time-series analysis. In this paper, we attempt to model empirically the cross-country incidence of the financial crisis of 2008. Ours is an exploratory approach; we view it as a first step toward creating an international early-warning system, which necessarily includes both time-series and cross-sectional elements. Our objectives are: a) to determine whether the data patterns can be fitted within sample; and b) to provide2 preliminary evidence on which causes of the financial crisis seem to predict its ex post incidence across countries. We conduct a non-structural exercise, using a “MIMIC” (Multiple-Indicator Multiple Cause) model, which we apply to a cross-sectional data set of 107 countries. Our MIMIC specification explicitly acknowledges that the severity of a financial crisis is a continuous, rather than a discrete phenomenon, and one that can only be observed with error. It treats the severity of the financial crisis as a latent variable, observed only imperfectly in terms of such 2008 manifestations as equity market collapses, exchange rate depreciations, recessionary growth, and declines in the perceptions of a country’s creditworthiness. The MIMIC methodology (described in more detail below) simultaneously links these “indicators” of a financial crisis with potential “causes” of the crisis. In the process, we obtain estimates of the severity of each country’s crisis