Analysis of co-movements between the capital markets in Brazil and the United States

Abstract

In this article copula theory is used to analyze the co-movements between the Brazilian and American capital markets. To formulate an effective asset allocation strategy, it is important to understand extreme events – both positive (booms) and negative (crashes) – and their effects on markets. The market indexes used are the Ibovespa and the S&P 500, covering the period from March 2001 to April 2007. We tested the adherence to the log-returns of the main copulas found in the financial literature, using the following criteria: log-likelihood, Akaike information criterion and Bayesian information criterion. The results show that the symmetrized Joe-Clayton copula is most suitable to model the dependence structure between the log-returns of the Ibovespa and the S&P500. This work differs from some previous ones (e.g., Mendes & Moretti, 2005 and Canela & Collazo, 2005) because we take into account the modeling of dynamic copulas, as introduced by Patton (2006). Finally, from the tail-dependence indexes over time, it can be concluded that the occurrence of crashes in the American market tends to affect the Brazilian market more than does the occurrence of booms.