Date of Award

1990

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Finance

First Advisor

G. Geoffrey Booth

Abstract

This dissertation investigates the time-series properties of intradaily stock prices. It provides a model of the return generating process that is capable of incorporating not only such institutional constraints as the specialist's bid-ask spread, but also the presence of dependence in the conditional variance. It extends the literature by introducing a conditional error distribution, the power-exponential, that adequately accounts for not only leptokurtosis but also peakedness in the empirical distribution. Evidence is presented that suggests intradaily returns are best modelled as a mixture of distributions. Furthermore, it documents the inability of information proxies, such as volume or the number of trades, to account for the presence of autoregressive conditional heteroscedasticity in the data. And lastly, it examines the robustness of variance ratio statistics to test the null hypothesis of a random walk in the presence of higher order dependence.

Pages

169

Share

COinS