Published in The Review of Economics and Statistics, Volume 66, Issue 4, November 1, 1984, pages 678-682. Copyright © 1984 MIT Press. The definitive version is available at http://www.jstor.org/stable/1935994.
NOTE: At the time of publication, the author Michael L. Marlow was not yet affiliated with Cal Poly.
It is argued that the estimation techniques used by previous researchers to study rivalry in financial markets are inappropriate. The assumptions of both ordinary least-squares and Tobi analysis are violated when these techniques are used to analyze mobility and turnover data. To overcome the difficulties in the previous studies, we suggest a non-linear model (which is closely related to the Poisson model). This model is designed for describing frequency data and is not subject to the criticisms to which ordinary least-square and Tobut are subject.