Volatilities and Correlations of the largest Banks of Europe and America with Financial Conditions Indexes

Modelling covariances across financial asset returns is important for financial management. This research models the influence of financial conditions on the European and American largest banks volatilities and correlations. Using daily stock returns of both EU and US largest banks during the period 2001 to 2016 and proxying financial conditions by the European and American Bloomberg Financial Conditions Indexes, I find that a block structure in the (c)DCC models is needed to separate the influences of the EU and US FCI on the respective correlations. I also find that incorporating EU and US financial conditions indexes has a significant affect on the respective variance processes. Specifically, variances go up when financial conditions get worse. Another contribution is using a data sampling technique to incorporate different frequency data in the models. The Log-Garch-Midas-X and Spline-Garch-X are the most preferred variance processes. I forecast Value-at-Risk using different variance and correlation processes. Various statistical tests and performance measures are considered to obtain the statistically preferred processes.