Abstract: The Asian financial crisis and Greek debt crisis showed that small countries can lead to a severe crisis. Apart from Thailand and Greece, those countries which can trigger a regional financial crisis or international financial crisis should not be ignored. Based on complex network theory, a global macroeconomic network containing 148 countries and regions using international trade data of 2007 is constructed. Moreover, the susceptible infected recovered (S.I.R) epidemic model with a changeable probability of infection is utilized to simulate the financial crisis transmitting through network. By this method, the number of countries and the aggregate economic volume that different countries will affect in different crisis levels can be obtained according to which countries are classified in six categories using dynamic clustering method. By calculating the Spearman correlation coefficients between contagiousness and GDP, import volume and index of network topology comprising degree, closeness centrality, betweenness centrality and clustering coefficients, the major influence factors that can explain the difference of contagiousness are explored. The results show that there is region feature in the difference of contagiousness and many countries with lower GDP can lead to a severe crisis. Countries located in category 1 to category 4 in the global macroeconomic network are potential infectious source. Countries with higher GDP and import will lead to negative spillover when a small crisis occurs, while countries with more trade partners and in the center of the network have the potential to lead to a severe crisis. Then the financial crisis transmitting through network of 2009 is simulated and the results pass the robust test. In this paper, the conclusions make for discovering countries which should be concerned and building a global early warning system to watch for future financial crisis.

Key words: complex network, international financial crisis, potential infectious source, S.I.R model, contagiousness