Previsão de volatilidade da taxa de câmbio dólar/real por meio de modelagem GARCH

Abstract

A much sought after topic in scientific and practical terms in economic science is the possibility of specific estimation of distinct variables. And especially when specifically estimating the variation of financial series of derivatives or foreign currencies, the interest and complexity are generally even greater. Thus, given the target financial asset, the possibility of anticipating its future price is one of the great recurring unknowns in the financial market. In this research we will work with time series methodologies, and to avoid circumventing this situation, the objective of this research is to apply the GARCH methodology to an exchange rate series in the relationship between the US Dollar and the Brazilian currency Real, to verify the possibility of predicting the volatility of the series. observed. The results found found that when we ran the GARCH model for the series of exchange rates observed in this work, we found that the projected and realized volatility had an average difference of 0.33%, comparing daily ten days of projected volatility with ten days of volatility , in fact, accomplished.

Keywords: GARCH model, forecast, exchange rate.