In Table we cannot reject the

In Table 4, we cannot reject the null Doxorubicin of the Strazicich tests (2003, 2004) at the 5% level of significance. The real effective exchange rate for these countries is not stationary in the presence of structural breaks. Another result in Table 4 is that with two structural breaks, the exchange rate of Brazil and Colombia experienced a structural break between August and October 1998. This structural break is associated with the capital outflow in emerging economies in the Asian crisis in 1997, the Russian crisis in August 1998. As a result, this outflow caused the overshooting of the exchange rate in Brazil in January 1999. As discussed in Section 3.2, the characteristic of non-linearity of series can reduce the power of linear unit root tests. Regarding the exchange rate, according to Cuestas and Regis (2013), the non-linearity of the series occurs for changes in the exchange rate regimes, periods of hyperinflation and deflation. To these possibilities, Kapetanios et al. (2003) add the technology change, which implies a permanent change in the terms of trade and relative prices between tradables and non-tradables. In this sense, with the results of the linearity test, in which the series of the effective real exchange rates in Mexico and Venezuela are configured as non-linear, we executed the non-linear unit root tests of Kapetanios et al. (2003) and Kruse (2011), reported in Table 5. It is observed in Table 5 that the hypothesis is valid only for Mexico. This result was confirmed by both the Kapetanios et al. (2003) and the Kruse (2011). The Kruse approach is the most suitable since it has higher power than the Kapetanios et al. (2003) test. In sum, of the seven Latin American countries in which the unit root tests were applied, only three (Chile, Mexico and Peru) found statistical support for the validation of PPC for the time period analyzed. These results are consistent with the work of Bahmani-Oskooee et al. (2008) and Su et al. (2014) that just as our work also use the real effective exchange rate. However, when considering the real exchange rate our results are consistent with Drine and Rault (2008), Su et al. (2011), as they conclude that the PPP hypothesis is valid only for some Latin American countries. In this context of real exchange rate the hypothesis of purchasing power parity is valid for all the countries of Latin America to Cheng et al. (2008), Divino et al. (2009) and He et al. (2014). The divergences identified in the results of these papers can be attributed to differences in the periods under review as well as the use of different approaches of exchange rates (RER or REER) as mentioned previously.
Conclusions Empirical studies show that the validity of the PPP hypothesis depends largely on the period of analysis, the type of data used and the economic characteristics of the countries. As emphasized in the literature (Alba and Park, 2003; Divino et al., 2009; He et al., 2014; Su et al., 2011), it is known that the Latin American countries share important similarities in their economic histories, such as the period of high inflation and processes of trade liberalization after the 1980s. Thus, these specific characteristics of Latin America would be some of the factors that contribute to the empirical verification of the PPP hypothesis. However, our results contrasted with this expectation to verify the validity of the PPC only for three of the seven countries analyzed.
Introduction The European Union (EU) is an important trade partner of Brazil, representing 22.5% of Brazilian total trade in 2012, according to MDIC (2014b). However, since 2014 Brazil no longer benefit from the EU\'s generalized scheme of preferences (GSP), loosing preferential tariff treatment granted to developing countries exports, as it was classified as an upper middle income country. As Brazil resumed talks on a possible EU–Mercosur agreement in 2013, opening the possibility of individual negotiations among Mercosur members, the formation of a preferential trade agreement could reverse the loss of this GSP preferential treatment with the European block.