All Rights Reserved
AccessEcon LLC 2006, 2008.
Powered by MinhViet JSC

 
Marcelo Brutti Righi and Paulo Sergio Ceretta
 
''Extreme values dependence of risk in Latin American markets''
( 2011, Vol. 31 No.4 )
 
 
This paper aims to determine which extreme value copula is best suited to the bivariate relationships between shocks of U.S market with Brazilian, Argentine and Mexican markets. We used prices of S&P500, Ibovespa, Merval and IPC from January, 3, 2009 to December, 31, 2010, totaling 483 observations. We estimated Gumbel, Galambos, Husler Reiss and Student's t (extreme-value - TEV) copulas. Results allow concluding that there is a strong dependence on the tails of the joint probability distribution of these markets. Nevertheless, the Gumbel copula for Brazil, Husler-Reiss copula for Argentina and TEV copula for Mexico had the best fit. Thus, its important use an extreme value properly diversify the risk in a portfolio that consist of assets in these countries, especially in turbulence periods.
 
 
Keywords: Risk, Extreme values, copulas, Transmission, Emerging markets
JEL:
C4 - Econometric and Statistical Methods: Special Topics
 
Manuscript Received : Jun 22 2011 Manuscript Accepted : Oct 14 2011

  This abstract has been downloaded 1884 times                The Full PDF of this paper has been downloaded 166541 times