RESUMEN
La relevancia que ha adquirido la formación de joint ventures contrasta con que la utilización de esta figura asociativa puede entrar en conflicto con las disposiciones comunitarias en materia de competencia. Para velar por su cumplimiento, la Comisión tiene encomendada la tarea de emitir un dictamen acerca de la adecuación a la normativa en vigor sobre las propuestas de creación de joint ventures que se presentan en el Espacio Económico Europeo. El presente trabajo pretende conocer cuáles son las variables más relevantes y en qué sentido inciden en su decisión y, al tiempo, elaborar un modelo predictivo que permita anticipar cuál será su decisión ante futuras operaciones de constitución de joint ventures. Para ello, se lleva a cabo un análisis discriminante que permite agrupar las joint ventures analizadas según la decisión adoptada por la Comisión (compatible, exenta e incompatible) y una red neuronal que contrasta y corrobora los resultados alcanzados acerca de la capacidad predictiva del modelo y de las variables empleadas. Del análisis realizado se sigue que son tres las características especialmente significativas que inciden en la decisión final de la Comisión: (a) El liderazgo que mantendrá la joint venture tras su creación, (b) las aportaciones realizadas por los copartícipes, y (c) el carácter cooperativo o concentrativo de la joint venture.
ABSTRACT
The acceleration in the process of globalization of the world-wide economy, among other reasons, has imposed an adaptation challenge to the companies, which they have confronted, often, resorting to diverse business associative formulas. In this context, joint venture has acquired an outstanding relevance, becoming a habitual figure of business world-wide. However, this figure can cause conflict with the communitarian dispositions in the matter of competition. This paper tries, firstly, to know which are the most excellent variables and in what sense affect its decision, and secondly, to elaborate a predictive model that allows anticipating which will be the decision of the Commission in each case. The predictive model has been elaborated by means of a discriminant analysis and an artificial neural network. The discriminant analysis allows to explain the membership of each case, joint ventures of the sample, to the pre-established groups: the three possible decisions of the Commission (compatible, exempted and incompatible). With the artificial neural network it has been possible to contrast, and to corroborate in its case, the results obtained with the discriminant analysis, about the predictive capacity of the model and the variables used. The analysis shows the existence of three characteristics contained in the agreements for the establishment of joint ventures that are specially excellent in the final decision of the Commission: (a) The leadership that will hold the joint venture created, (b) The contributions made by the partners and (c) The cooperative or concentrative nature of the joint venture.