Date of Award
Doctor of Philosophy (PhD)
Very little theory has been developed on the effect of marketability on the distribution of returns. As a result, this study is an empirical exploration of that relationship, without any strong, preconceived hypotheses. Since the distributions were likely to be non-normal, but otherwise indeterminate as to form, a test had to be developed to determine the existence of differences. It was decided that the first four moments and the studentized range could effectively capture the characteristics of a distribution and provide a set of measures to serve as variables in discriminant analyses. Marketability was defined as shares traded divided by shares outstanding. Four samples were selected representing differing degrees of marketability, but homogeneous in all other respects. These samples were then rigorously tested on the basis of both daily and monthly holding periods. The results of these tests indicated that a daily holding period is too short to reveal any reliable results. The analysis of daily returns produced results that conflicted with any logical risk-return relationship and that were inconsistent with the results of the tests conducted on the monthly holding period sample. The tests using the monthly holding period sample did indicate a significant relationship between marketability and the characteristics of ex post market generated return distributions. Further testing was conducted on random portfolios generated from the samples, verifying the prior results and indicating a strong non-diversifiable component in the relationship. Differences in marketability did not influence the speed of diversification.
Mcmanis, Bruce Loren, "A Study of the Effects of Marketability on the Distribution of Market Generated Returns." (1981). LSU Historical Dissertations and Theses. 3647.