Análises dos testes de cointegração e de correção de erro dos preços do café e do cacau no mercado internacional de futuros e opções

Abstract

In this paper is examined the properties of individuals time series of the prices of the coffee and cocoa in the international market, by means of owed tests, regard to stationarity and tests for unit roots. The distinction between longrun and short-run characteristics in time series has attracted much attention in the last decades. Long-run characteristics in economic and financial date are usually associated with non-stationarity time series and called trends, whereas short-run fluctuations are stationary time series and are called cycles. Economic and financial time series can be viewed as combinations of these components of trend and cycles. Moreover, the existence of common factors among two or more time series may have such that the combination of these time series demonstrates no features which the individual time series possess. There could be a common trend shared by two time series. If there is no further trend which exists in only one time series, then it is that these two time series are cointegrated. This kind of common factor analysis can be extended and applied to stationary time series as well, leading to idea of common cycles.