Uma avaliação acerca da falha empírica do teorema da paridade descoberta da taxa de juros entre o Real e o Dólar

Abstract

ResumoEste artigo testa a validade do teorema da paridade descoberta de juros para os dados da economia brasileira no período de 2000 a 2014. Nossos resultados corroboram a não validade empírica, conhecida na literatura como de UIP Failure ou Forward Premium Puzzle. O coeficiente do diferencial de juros estimado por um modelo GARCH apresenta sinal negativo, contradizendo parte dos testes da UIP em economias periféricas. Já os modelos markovianos de mudança de regime apontaram para dois padrões bem delimitados: nos períodos de menor volatilidade o coeficiente assume valor negativo e, nos meses de alta volatilidade, valor positivo. Os resultados encontrados são coerentes com a abordagem da taxa de juros exógena em uma economia aberta.Palavras-chave: Paridade descoberta de juros; Abordagem da taxa de juros exógena; UIP failure; Teoremas de taxa de juros; Taxa de câmbio. Abstract An evaluation of the uncovered interest parity failure for the Brazilian economyThe present paper tests the validity of the uncovered interest parity (UIP) theorem using data from the Brazilian economy from 2000 to 2014. Our results endorse the empirical failure of this theorem, known in the literature as UIP Failure or Forward Premium Puzzle. The estimated coefficient of the interest rate differential in a GARCH model is negative, contradicting most of the UIP tests for emerging economies. However, using Markov switching models we found two well-defined patterns: during periods of lower exchange rate volatility the estimated coefficient becomes negative, while in high volatility periods the estimated coefficient is positive. These results are consistent with the exogenous interest rate approach in an open economy.

Distribution and Conflict Inflation in Brazil under Inflation Targeting, 1999–2014

“…So the positive interest rate differentials often lead in Brazil to a process of continuous nominal exchange appreciation. Empirical evidence of this effect in Brazil (Cieplinski, Braga, and Summa 2015) shows that it is a relation between interest rate differentials and the change (not the level) of the nominal exchange rate. It has also been found that the relation between interest rate and changes in the exchange rate itself changes over time.…”