The proposed model effectively takes into account the mutual influence of the elements of the dynamic series that influence the increase of competitiveness of small agricultural formations, that is, the influence of various economic parameters on each other when they simultaneously manifest themselves. In this case, the forecasting operator is actually trained on the statistical material of the past. From this point of view, the prediction model introduced by us, is a neural network.