Résumé en anglais

When firms use alliance relationships, important decisions they must make concern: (1) How the alliance should develop in order to accomplish the parent firms’ objectives; (2) How they can manage relational risks in the alliance relationship. In order to find an answer to these problems, researches have shown the importance of commitment and cooperation. For instance, cooperative behaviour has been analysed through the lens of the prisoner’s dilemma in which each player‘s decisions depend on the other's moves. In its simplest form, this framework involves two players, A and B who can either co-operate or not cooperate with each other. In the case, the game would be played only once; the non-cooperative strategy leads to individual strategy optimization. When results depend on the “shadow of the future”, iteration improves the prospects for cooperation by encouraging strategies of reciprocity. In this perspective, performance results from a tit-for-tat strategy (Axelrod, 1984). However, the prisoner’s dilemma game is not a very realistic way to model strategic alliances and it does not cover all the angles for successful strategic behaviour (Camerer, 1991). Firm can also commit unilaterally. These unilateral investments determine alliances’ success. Khanna, Gulati, and Norhia (1998) demonstrate the advantage of not co-operating in repeated games as increasing over time. They suggest that unilateral commitments can emerge as one way to influence the outcomes and are more favourable to the success of the relationships. Firms face uncertain futures in alliance relationship, and the investment opportunities they face are, to an important degree, a function of their prior investment commitments (Doz, 1996). Thus, the real options framework appears to precisely fit firms’ strategic challenges by linking current actions to uncertain futures. Consequently, unilateral commitments can be seen as mechanisms that help a firm manage risks in alliance relationships in a proactive manner.