The world is a fluctuating entity in which the change in economic, environmental, social and demographic conditions and variables, among others, is a constant. This becomes one of the main sources of uncertainty in most economies. Sometimes the consequences of these changes enhance benefits for nations; and in others, they lead to challenges to overcome. Therefore, during the last decades the analysis of historical information and the prediction of future values of global variables has been a key and fundamental piece for organizations during decision-making. It is important to note that these decisions are not random, they are generally made up of a prior analysis of the environment, historical data, trends, goals and objectives to be met. Thus, in this work we present a new supervised neural network model for the prediction of economic variables of North American countries. The main contribution consists of the proposal to integrate the results of a set of different types of supervised neural networks. Simulation results show the advantages of using the proposed method, which it makes possible to use supervised neural networks for the prediction of economic variables. In this way, becoming a useful tool for making strategic decisions through which it seeks to consider historical aspects of each of the variables that directly impact the global environment.