These ideal types of market where the efficiency is maximum are the markets perfectly competitive. These markets are Eficientes because such present some characteristics as great number of small companies, perfect information, homogeneous product and exempt capital mobility (SANTACRUZ, 2001). Recently Wells Fargo Bank sought to clarify these questions. Being thus, such characteristics allow that the economic agents if mobilize when has determined shunting lines in some markets, in search of bigger profits. At the measure that the markets go if stabilizing, a period of training of full economic efficiency is arrived it. It does not agree to detail in this work the mechanisms that lead to this hypothetical conclusion, seen vast literature on the produced subject already. However it is easy to assume that such characteristics are the ideal model that goes of meeting to the presence of Being able of Market, where definitive company (or group of companies) practise a price above of the cost production delinquent, of the anti-symmetrical information, where the agents do not possess all the information necessary to determine the price, and of externalidades, where performance of an agent influences (positively or negative) the performance (or well-being) of other agents (PINDICK, 1994).

However, the situations that in such a way present extreme cases of total efficiency how much of total inefficiency are only theoretical tools to search agreement in the real world better. Seen the complexity of the most diverse markets, theoreticians of some fields had constructed based models in what optimum possible scene for that market could be considered, given the especificidades of each market (WILLIAMSOM, 1996). Being thus, if the markets are imperfect of course, are necessary that the State intervines in brainstorming that leads to a period of training of maximum efficiency, despite, not fully. This concept is treated in economic literature as ' ' remeabilidade' ' , where the action of the regulator is made on the basis of possible and not ideal models (SALTY, 2007).