New early warning system can flag global financial crises

Policymakers around the world may soon be able to effectively avert or lessen the impact of financial crisis, thanks to a new early warning system tested across a range of global regions, including India and China.
The study by experts at the University of Birmingham in the UK compared the new system against two recently developed methods using econometrics - the application of statistical methods to economic data.
All three systems - multinomial logit regression, dynamic signal extraction and the university's refined binary logit model - use the technique of logistic regression to analyse various indicators, such as a country's exposure to debt, foreign trade, domestic growth and government expenditure.

The new system was found to significantly outperform the other two methods, researchers said.
The study looked at how the three systems performed across a range of global regions, including China, India, Latin America, Africa and South-East Asia.
An early warning system based on the Birmingham model can be recommended to policymakers in those regions - particularly when they value avoiding negative market sentiments and damage to international reputation, which might possibly be triggered by false signals of sovereign debt problems, researchers said.

Existing prediction systems failed to forecast the global crash of 2008, which led to several governments bailing out their banks and European nations, such as Greece, Portugal, Ireland and Spain, being plunged into a sovereign debt crisis.

The new system takes into account the differences between developed and developing countries.
This has produced a more accurate set of indicators that warn of impending fiscal catastrophe.
"We have developed a new type of 'early warning system' that will provide more accurate predictions of sovereign debt crises and how long they are likely to last," said Frank Strobel, Senior Lecturer in Economics.
"Financial monitoring tools that can forewarn the build-up of major financial turmoil are increasingly important.
"Our system provides policymakers with time to take corrective action that would help avert, or at least mitigate, the damage associated with an approaching crisis," Strobel said.

Until recently, early warning systems focused on factors relating only to developing countries. Birmingham's development takes account of the causes and associated indicators of sovereign debt crises varying between different countries and regions.
"There is a clear difference between regions in how early warning signs behave. External factors such as openness to trade behave differently around crises in Asia and Latin America, whilst domestic macroeconomic conditions seem to play the major role in Africa," Strobel added.
"In developed countries, only the factors around exposure to debt appear to be potentially important signals of impending debt crises. A rise in the global lending rate increases the cost of servicing debt and magnifies the risk of sovereign defaults in general," said Strobel.

The study was published in the Journal of Financial Stability.