Influência das taxas de juros internacionais sobre volume de operações de crédito externo no Brasil : aplicação do método de regressão não-paramétrica com estimadores de núcleo

AUTOR(ES)
FONTE

IBICT - Instituto Brasileiro de Informação em Ciência e Tecnologia

DATA DE PUBLICAÇÃO

05/02/2009

RESUMO

The influence of interest on the amount of credit is an important topic in economic. In fact, it has attracted interest of several authors who have made use of parametric statistical models (Topel, 1988; Vivacqua, 2007, among others). This dissertation presented to the analysis of the influence of international interest rates on the volume of foreign credit operations in Brazil by means of regression models with non-parametric estimators of core. The motivation for the use of non-parametric methods is linked to the fact that these methods can adjust the regression curves in a vector space wider than the classical regression method (Härdle et al, 2004). By using regression between series observed over time, this work also found a long-term dependence of the variables in the context of time series. The ADF test (Dickey and Fuller, 1979) and the estimators of long memory parameter of the ARFIMA model (Reisen, 1994) were applied to analyze the stationarity of waste from regressions. The results of this analysis showed that the variables are co-integrated fractional order, ie, have long-term dependency. The trajectory of the curve of non-parametric regression showed empirically that, in the interest rate corresponding to 1% and 2%, the rate of increase does not necessarily imply a reduction in volume of external credit operation.

ASSUNTO(S)

regressão não-paramétrica estimador de núcleo bandwidth nonparametric regression bandwidth economia kernel functions

Documentos Relacionados