The mimetic nature of the behavior of financial actors.

Summary The emergence of the notion of mimicry has not been easy in management. This is due to the fact that classical economic theory is based on the individual, one and rational. As Dupuy (1983) notes: "economics presents itself as the negation of the crowd". It is precisely by rejecting the ignorance of the crowd that the economic agent gains access to rationality. In the Valrasian theory, exchanges take place at equilibrium and this equilibrium is the result of individual and independent confrontations. Under these conditions, it is easy to understand the difficulty of understanding the notion of group in a managerial context. If Keynes is one of the first theorists to have described imitation as economically rational behavior, Orlean (1986) offers a demonstration of this assertion and identifies two important concepts in Keynesian thought. The first notion is universal in scope. It describes the case of a system in which not all participants have the same information. If an agent is in a state of total uncertainty, resorting to imitation can only improve his performance, since he is plagiarizing the decision of an actor who is necessarily better informed than he is. The second notion describes the case of a system where no actor is informed. In this situation, imitation remains rational. In this case, agents seek to reduce their risk in the face of competition. Imitating one's neighbor is tantamount to making a decision that is in the common direction. This attitude is perfectly applicable to markets, which penalize being wrong against everyone much more than they improve the fact of being the only one to be right. The study of mimetic phenomena is of definite interest, insofar as they offer answers to certain anomalies detected in the markets. The most classical questions, such as the violation of the principle of the random walk of prices, or the creation of bubbles, take on a new impetus with the notion of mimetism. Here, our objective is certainly not that of a global modeling of the notion of mimicry. The knowledge on human behavior remains today too partial to hope for such a result. We have four main goals. First, we hope to provide a synthesis of existing theoretical work.