Abstract
This paper addresses the problem of testing financial models in the presence of market microstructure effects. The moment restrictions implied by the financial and market microstructure models are jointly tested using Hansen’s (1982) GMM approach. To illustrate the methodology, I consider the random walk model in combination with the bid-ask price effect model of Blume and Stambaugh (1983). Within this sufficiently simple framework, I obtain closed-form expressions for the estimators, standard errors of the estimators, and the test statistic, which affords an opportunity to examine the precision of the estimators and the power of the test as the return interval increases. I show that apparent rejections of the random walk model cannot be sustained when tests of the model are adjusted for market microstructure effects, and I discuss other applications of the methodology.
Original language | English |
---|---|
Pages (from-to) | 519-540 |
Number of pages | 22 |
Journal | Journal of Financial and Quantitative Analysis |
Volume | 29 |
Issue number | 4 |
DOIs | |
Publication status | Published - 1994 |
Externally published | Yes |