Resumen
The main goal this article was to find the best way of making forecast about future volatility using implicit or statistic forecast. The work is based on Telemar S.A. shares data from 21/09/1998 to 21/10/2002 and Telemar S.A. shares data from 2/10/2000 to 21/10/2002. The implicit volatility was obtained using back-out procedure from the Black-Scholes model. The statistics forecasts were obtained using weighted moving average models, GARCH, EGARCH and FIGARCH models. The Wald statistic shows that EGARCH and FIGARCH models are efficient and are not biased forecasts for Telemar S.A. absolute variation between t and t + 1. The volatility evaluation during the maturity time of an option, rejects the hypothesis that implicit volatility is the best forecast to future volatility and the Wald statistic show that FIGARCH model is an efficient and not biased forecast.