Aplicação dos modelos da classe ARCH na previsão de volatilidade de taxas de juros das operações sobre os activos dos Bancos Comerciais no mercado Moçambicano

Abstract

Financial time series, such as those of interest rates, stock prices, exchange rates inflation rates, etc., often exhibit the phenomenon of volatility clustering, i.e, periods in which their prices show large oscillations for an extended period of time, followed by periods in which there is relative calm, suggesting that the variance of time series time series vary over time. The main scope of this work focuses on the determination of a quantitative model consistent and representative model for forecasting interest rate volatility of operations on commercial banks' assets through statistical and econometric procedures for univariate time series. The ARCH, GARCH and EGARCH models were compared to analyze the interest rate volatility in the period from January 2001 to July 2014. The application of the selection criteria suggested that the GARCH (1,1) model is the best fits the behaviour of the interest rate series within the period under study. Nevertheless, for forecasting purposes, that is, for subsequent periods, the best fit to the series of interest interest rates is the ARCH (1). Thus, from the estimated models, the best fit to the interest rates series volatility is the ARCH (1) model. interest rates is the ARCH model (1). (TRADUÇAO NOSSA)