New Classification of Developed and Emerging Countries Based on the Effects of Subprime Crises: Kohonen Map Method

This document provides a new classification of 42 countries following their assignments by the subprime crisis. We classify countries based on the fluctuations of economic indicators (GDP, interest rate, inflation, monetary aggregate, unemployment and exchange rate) and according of strategies implemented to deal with adverse effects of the crises. We employ Kohonen Self-Organizing Map (SOM) which is an unsupervised neural network algorithm that projects high dimensional data. Results of SOM show the optimal map represented by 56 micro-classes (cells or neurons) and five areas of grouping countries thus defining five principal different economic situations. Classification was done by level of development, evolutions of economic indicators, and by period. Each group of countries classified in the same cell implies that these countries have suffered similar effects of the crisis on economic indicators or they have applied a same strategy to fight against the crisis. Also, the results showed that changes in GDP, exchange rate, monetary aggregate M2, and inflation are the main characteristics of classification. We define the period of the global shock as running from 2008 to early 2010, which can explain why we find stronger results than earlier papers such as Obstfeld et al. (2009, 2010) and Rose and Spiegel (2009, 2010, 2011), Frankel and Saravelos (2011, 2012), which use 2008 and 2009 data.