A proposal to make the float factor of the Mexican stock market index more efficient: A mean reversion model for relative flotation

This paper develops two methodological approaches for calculating the float factor of the stock market index (IPC) of the Mexican stock market (BMV). The first one assumes that the free float factor is approximated by the normalization of the daily market rotation of each title, while the second one uses the relative daily market rotation. In both cases, the float factor is simulated by a variable with mean reversion similar to the one proposed in Cox, Ingersoll and Ross' model (1985). The methodologies we have established here remedy the inability to repli-cate the IPC because of the elements of confidentiality that the float factor currently has. Moreover, the proposed flotation factor can be updated instantly given the nature of the developed models. The paper shows that the weights of the firms in the ipc, calculated by using the formulated methodologies, are substantially different from those provided by the BMV when applying its flotation factor, which contains discretionary components of information difficult to verify.