METODOLOGÍA SISTÉMICA PARA SOLUCIONARPROBLEMAS ECONÓMICO – FINANCIEROS: CASO MEXDER

Abstract

This research proposes a model to predict the future trend of the Mexican Derivatives Market (MexDer), by using Artificial Neural Networks (ANN’s) The main contribution is to demonstrate that the proposed model based on artificial intelligence is able to obtain good approximations in prognosis. Also, the work contains a proposed methodology for the application of neural networks to financial time series. The variables evaluated at the moment are four, since these currencies Dollar and Euro and rates as interest the TIIE and CETES. The networks most convenient to use in this type of problems are recurrent networks for time series, multilayer perceptrón and Self-organized maps, because these are variables to predict mid and long term in time. This research used the multilayer perceptrón. The input data for each network model takes into account: the current date, current inflation, and the date of prediction. The ANN’s are be trained with historical data from the last 10 years to 2008. The variable data are be compared with actual data from the current year (2009).