Brazilian international and inter-state trade flows: an exploratory analysis using the gravity model

Recent efforts toward a world with freer trade, like WTO/GATT or regional Preferential Trade Agreements(PTAs), were put in doubt after McCallum's(1995) finding of a large border effect between US and Canadian provinces. Since then, there has been a great amount of research on this topic employing the gravity equation. This dissertation has two goals. The first goal is to review comprehensively the recent literature about the gravity equation, including its usages, econometric specifications, and the efforts to provide it with microeconomic foundations. The second goal is the estimation of the Brazilian border effect (or 'home-bias trade puzzle') using inter-state and international trade flow data. It is used a pooled cross-section Tobit model. The lowest border effect estimated was 15, which implies that Brazilian states trade among themselves 15 times more than they trade with foreign countries. Further research using industry disaggregated data is needed to qualify the estimated border effect with respect to which part of that effect can be attributed to actual trade costs and which part is the outcome of the endogenous location problem of the firm.

O mercado de câmbio brasileiro pela ótica da microestrutura

O objetivo deste trabalho é investigar se a relação entre a taxa de câmbio spot e o fluxo de ordens deriva do fato do fluxo agregar informações a respeito dos fundamentos econômicos dispersos na economia. Para efetuar este teste foi utilizada uma base de dados que engloba todas as transações dos segmentos comercial e financeiro no mercado cambial primário brasileiro entre janeiro de 1999 e maio de 2008. Mostramos que o fluxo de ordens foi razoavelmente capaz de explicar variações nas expectativas de inflação, relação que não se manteve robusta para escolha de outros fundamentos, como PIB e Produção Industrial.