The 2008 financial crisis prompted a renewed interest among economists to construct an “early warning system” but a new study concludes that out of 65 potential causes of a global economic meltdown, few factors would have predicted the severity of the crisis across a multitude of countries.

The forthcoming paper for the National Bureau of Economic Research, “Cross-Country Causes and Consequences of the 2008 Crisis: Early Warning,” is co-authored by Professor Andrew Rose, the Bernard T. Rocca, Jr. Chair in International Business & Trade at the University of California, Berkeley’s Haas School of Business, and Mark M. Spiegel of the Federal Reserve bank of San Francisco.

“If the causes of the crises differ across countries, there is little hope of finding a common statistical model to predict them,” they said. Furthermore, the study’s findings present a challenge to the International Monetary Fund’s (IMF) current agenda to develop effective early warning models.

Rose and Spiegel think that any good early warning system must do two things: predict which countries will be most affected, i.e. the cross-sectional component, and also, when the crisis will occur, or the timing factor. Using a Multiple-Indicator Multiple-Cause model (MIMIC, Goldberger 1972), their study focuses on the countries most dramatically affected by the crisis as well as less affected countries in order to serve as control data.

The severity of the crisis in each country was measured by four variables: 1) the 2008 real gross domestic product (GDP) growth; 2) the percentage change in a broad measure of the national stock market over 2008; 3) the 2008 percentage change in the Special Drawing Right (SDR) exchange rate; and 4) the change in a country’s credit-worthiness rating from Institutional Investor, a business news and research publisher. The IMF defines the SDR as “an international reserve asset, created by the IMF in 1969 to supplement its member countries' official reserves. Its value is based on a basket of four key international currencies, and SDRs can be exchanged for freely usable currencies.”

“Our research found that size, from an economic perspective, had no significant impact on the incidence of crises across countries,” says Rose, “while income has a significantly negative impact; richer counties experienced more severe crises.”

The next step was to identify a list of characteristics that may have impacted a country’s economic performance during the crisis. Collected using data from 2006 and earlier, the characteristics included financial policies and conditions; domestic macroeconomic policies; and appreciation in equity and real estate markets.

In the United States, for example, many fault inflated real estate values and the subprime credit market for causing the American economy’s failure and subsequent recession. The results found while an economic collapse occurred in countries where the real estate bubble burst, other countries such as Germany and Japan, had an equally serious crisis despite the fact that both countries did not experience significant decline in housing prices.

Furthermore, Rose points out that the bubble in real estate prices was mirrored by an increase in stock prices. Notably, equity appreciation emerged as the only variable that could robustly help predict crisis severity. The sample countries’ geography and proximity to strong neighboring economies were also considered in the study but did little to consistently predict a future meltdown.

The paper concludes, “Overall, our results suggest that measurable pre-existing conditions across countries had little common impact on the relative severity of these countries’ crisis experience. “

Rose cites three possible reasons for the predictive failure. He says first, the causes of the 2008 crisis may have differed across countries. Second, the crisis might have been the result of a “truly global shock” unrelated to the characteristics considered in the study. Or finally, the shock may be of a national origin that then spreads to other countries. Rose concludes, “Politicians seek a magic instrument to prevent another serious economic crisis but our findings show there is no mechanical easy way to predict major downturns in the future.”

See the full paper: http://www.nber.org/papers/w15358

More on Andrew Rose: http://www2.haas.berkeley.edu/Faculty/rose_andrew.aspx