Abstract
The impact of contagion between countries and international banks continue to grow in importance as localized shocks are more consistently translated into global crisis. By employing the Dees, di Mauro, Pesaran, and Smith – Global Vector Autoregression methodology, we are capable of capturing system wide volatility fluctuations. Twenty-two individual country models are constructed, utilizing stock price data from 11 large countries, including the US, UK, Germany, and the emerging markets Brazil, India, China, and South Africa and 11 international banks, chosen from the largest banks in each country, to measure the contagion effect across three distinct periods: the pre-crisis, crisis, and post-crisis. We formulate a global solution to our individual country models and construct a series of time-varying weight matrices for each period. The resulting impact elasticities created by the Global Vector Autoregression indicate an increase in co-movement of banks during the crisis period, but country results are less conclusive. HSBC Holdings, the international bank chosen for the UK, demonstrates a surprisingly high level of reactivity to changes in its foreign counterparts, indicative of risk. Finally, we create generalized impulse response functions for systemic shock simulations in order to monitor spillover potential in the global market. These results concur with those shown in our impact elasticity analysis. We conclude that the contagion effect was strongly present during the crisis period for the banking sector. Using the generalized impulse response functions, each country and bank is ranked based on its vulnerability to a negative shock in US Stock Prices. We show that Brazil, India, China and European Banks were particularly susceptible to a shock in the US during the crisis period.