This paper presents a nonlinear forecasting method for financial time series taking into account calendar effects, improving the quality of the forecasts and leading to the development of profitable trading strategies (excluding taxes, transaction and other costs