Investigação da memória de longo prazo na taxa de câmbio no Brasil

Abstract

Sumário: 1. Introdução; 2. Metodologia; 3. Dados; 4. Resultados empíricos; 5. Conclusão.Neste trabalho, é medida a evolução da memória de longo prazo da taxa de câmbio diária, Real contra Dólar dos Estados Unidos, no período de 1995 a 2004. Essa medição é realizada por meio da análise R/S clássica, com janela móvel de dados. O trabalho focaliza o abandono do regime de câmbio administrado em favor do de câmbio flutuante, ocorrido em 1999, identificando antipersistência da taxa de câmbio durante a vigência do primeiro regime e memória longa a partir do início da vigência do segundo regime. Mostra também evidência de memória longa para as volatilidades dos retornos das taxas analisadas. This paper presents measures of long-range dependence in daily exchange

Analysis of Grassland Degradation in Zona da Mata, MG, Brazil, Based on NDVI Time Series Data with the Integration of Phenological Metrics

“…Simultaneously, processes may persist over long periods because of traditional production systems and long-term cultivation and practices with low efficiency on farms. The persistence of temporal phenomena was estimated by the R/S analysis and Hurst exponents-the memory effect that has been extensively studied in several fields of study [26,29,30]. An ES S was built to obtain solutions for action planning and to diagnose several processes involving decision-making from a knowledge base of the study area.…”

“…R/S analysis and the Hurst exponent have been structured to determine memory in time series, where H < 0.5 indicates anti-persistence or noncorrelation, with the phenomena tending to revert in the future; H = 0.5 indicates random motion, also called Brownian motion, with values not reaching or defining a trend; and H > 0.5 indicates persistence or memory, with past values affecting future results [21]. In economics, when applied to price series, investments, and capital markets, the Hurst exponent is used to detect memory effects and trends [26,[44][45][46]. It has also been used in academic, environmental, and remote sensing applications [47].…”

“…Hurst exponents were estimated for the matrices arranged into time series using the principles established in the equations below and were numbered from 6 to 10 [26]. The calculations for each set of pixels in time were performed by binary division.…”

“…The estimate of the slope metric from the NDVI data, integrated annually, has been combined with the Hurst exponent using rescaled range (R/S) analysis and is used to create classes of vegetation degradation based on vegetation changes and long-term trends [24]. R/S analysis and the Hurst exponent have been used for trend analysis in different fields of study, such as economics and computation [25][26][27][28][29]. Hurst exponents have also been estimated for a region in Tibet, China, and showed future trends in vegetation dynamics with persistence for a large portion of the forests and lower persistence (nonpersistence) for pastures, both of which are affected by topography and altitude [30].…”

Investigando a hipótese da paridade do poder de compra: um enfoque não linear

“…No caso do Brasil, muitos autores já analisaram a evolução da taxa cambial, dentre eles destacam-se Badani e Hidalgo (2005), Lima e Tabak (2008), Albuquerque e Portugal (2006), Sousa Junior (2010), Souza et al (2006) com diferentes metodologias e objetivos. Para um survey dos testes de raiz unitária para a taxa de câmbio real no Brasil pode-se consultar Freixo e Barbosa (2004) e especialmente Kannebley-Jr. (2003).…”