Predicciones de modelos econométricos y redes neuronales: el caso de la acción de SURAMINV

the purpose of this paper is to construct statistical, econometric and artificial intelligence models that permit market behavior forecasts of the suraminv stock. evidence was found in favor of using econometric and artificial intelligence models constructed form main components that enable the daily behavior of the suraminv stock contrasting with the market's weak efficiency theory. the paper goes further than others that relate to the same topic in the sense that instead of looking for a good in sample forecast, it looks for out of sample results, controlling this way snooping data and therefore providing information that can be used in negotiation strategies.