Abstract
The need to understand and measure market maker bid/ask spreads is crucial in evaluating the merits of competing market structures and security designs. Prior studies of bid/ask spreads suffer from several forms of misspecification, including inadvertent and erroneous use of weighted least squares regression. This study develops a simple, parsimonious model of the determinants of spread, and then tests it empirically on a sample of NASDAQ stocks. The model performs well and avoids the distortions of prior work. The study demonstrates the importance of proper model specification in providing meaningful inference regarding the determinants of spread.