The Efficient Market Hypothesis, the Gaussian Assumption, and the Investment Management Industry

The purpose of this paper is to disclose how the Gaussian form of the concept of market efficiency is at the origin of the contemporary professional debate on passive index-linked management which continues on despite the growing popularity of indexing among investment management practitionners in Europe. This particular Gaussian form entered the investment management industry in the 1970's and carries strong assumptions about the behavior of returns and the structure of the information set. We argue that this ill-defined debate on indexing is due to a confusion between efficiency and Gaussian efficiency. The originality of the paper resides in the point of view choosen as the « main thread ». Instead of focusing on informational issues, now better understood since the seminal paper of Grossman and Stiglitz (1980), we concentrate on the probabilistic aspects included in the testable applications of the concept, so as to connect Fama's statement of 1970 to Bachelier's work (Theory of Speculation) of 1900. We establish the link between Bachelier's dissertation and portfolio management applications of market efficiency. We argue that understanding the precise characteristics of the link associating the informational efficiency concept itself with the underlying probabilistic hypothesis leads to a better approach to the problems facing the investment management industry and allows us to understand the professional impact of the Gaussian form of efficiency. The issues of non normality of empirical distributions (fat tails problem) and concentration of performance are linked with the efficiency paradigm so introducing a rationale for the the new and emerging concept of model risk, which today appears relevant to the investment management industry.