Traditionally, univariate models in time series are studied, such as the ARIMA model, and _x000D_ from this model there is an extension to multivariate VAR models and VEC models that are _x000D_ stationary and non-stationary. For non-stationary multivariate models, it is required to prove _x000D_ that the null hypothesis is non-stationary by means of the cointegration tests, with the objective _x000D_ of determining that the cointegration tests based on the VAR and VEC models are applicable in _x000D_ non-stationary multivariate models. The research that presents a study of applied, explanatory _x000D_ approach, studying the VAR and VEC models through the methodologies of the Dicker-Fuller _x000D_ Augmented Test and the method of the first differences, as well as the information criteria, _x000D_ including an example to be able to prove that the series of the models comply with the theor...