Estimating value at risk and optimal hedge ratio in Latin markets: a copula-based GARCH approach

In this paper we use a copula-based GARCH model to estimate conditional variances and covariances of the bivariate relationships between U.S. market with Brazilian, Argentinean and Mexican markets. To that we used daily prices of S&P500, Ibovespa, Merval and IPC from January 2009 to December 2010, totaling 483 observations. The results allows to conclude that both the volatility of Latin markets, such as its dependence with the U.S. decreased in the period, resulting in lower estimates for the VaR and Hedge, compared with those based on the unconditional variance and covariance, emphasizing that after theeffects of the 2007/2008 U.S. crisis, these Latin markets can again be considered as options for international diversification for investors with assets of the U.S. market in their portfolio.