Files
Abstract
In this paper, we aimed to investigate beef cattle markets integration, when
taking into account structural breaks in multivariate (auto)regressions, and transaction
costs, in the Southeast and Central-West regions in Brazil. Data set is monthly, ranging
from 1980 to 2012. Our methodological approach is built on two main blocks: first, time
series are divided into smaller samples sharing common characteristics, according to the
structural breaks test due to Qu and Perron (2007); second, momentum threshold model
(M-TAR, ENDERS; SIKLOS, 2001) is employed addressing to issues as cointegration parâmeand
asymmetrical adjustment. Our results show two significant breaks, one related to the hyperinflationary period,
and another to the 2007/2008 financial crisis. Moreover, those breaks are decisive to understand the time series
dynamics. For instance, in the hyperinflationary sample, transaction costs were higher, although markets were
integrated. Conversely, between 2007 and 2012, prices were not synchronized, while transaction costs were almost
non-significant. Thus, integration analyses that do not consider structural breaks may mislead some relevant
aspects of the data.