Ensaios sobre investimento, abertura comercial e complexidade econômica

Abstract

This work focuses on studying the effect of foreign direct investment, trade openness, economic complexity and public investment on gross fixed capital formation. In the first chapter, the complementarity or substitution relationship between foreign direct investment and gross capital formation, also taking into account trade openness and economic complexity, through a data panel with information from 55 countries, between 1983-2021, and PVAR models. The evidence points to positive effects of shocks on trade openness and foreign direct investment on gross capital formation. Shocks in economic complexity, however, seem to have a negative impact on gross capital formation. In the second chapter, the occurrence of non-linear effects of trade openness, foreign direct investment and economic complexity on domestic investment considering different stages of development. To this end, a panel with data from 73 countries, between 1995 and 2019, is used, and a dynamic model for panel data in the presence of regressors and endogenous threshold estimated by the generalized moments method in first difference (DF- GMM), proposed by Seo e Shin (2016). The evidence conrms the occurrence of asymmetries in the relationships investigated. Foreign direct investment, trade openness and economic complexity report positive impacts in countries with HDI<0.80. On the other hand, in economies with very high development, HDI>0.80, such effects are negative; In other words, economic openness and complexity attract domestic investment in the least developed regime and replace it in economies with very high HDI. In the third chapter, finally, we seek to revisit the discussion surrounding the impact of public investment on the gross formation of fixed capital in the Brazilian economy, in order to verify whether there is a complementarity effect (crowding-in) or replacement (crowding-out), considering the level of public debt. To this end, monthly information is used, between the years 2007 and 2023, and DOLS models along the lines proposed by Stock e Watson (1993). The results seem to indicate a complementary relationship between the two variables, however, this relationship reverts to a crowding-out in the period after the change in debt level, from February 2015, the month in which a structural break in Brazil’s public debt.