Pre-Trade Transparency on Equity Markets: The Effects of Hidden Order Procedure in the Paris-Bourse

Dominique Vacheron and Gilles Duteil
Centre d'Economie et des Techniques du Financement, Université d'Aix-Marseille

Abstract

A key issue in the Market 2000 study is whether transparency -real time dissemination of trade and quote information- enhances or diminishes market quality, and thereby benefits or harms market participants. The U.S. experience in the past three decades has shown that high levels of transparency have resulted in fair, liquid, and efficient markets, and thus have benefitted market participants. Evidence from the Paris Bourse suggests on the contrary that block traders gain from a lesser degree of transparency, indicating the possibility of an optinal degree of transparency depending on the market structure and/or the strategic behavior of market participants. The findings also reveal an asymetric effect of transparency on purchases and sales.

The issue of transparency was examined using a sample of orders and transactions from the Paris Bourse from January 3 to January 28,1994. From a regulatory perspective, the results of this research are relevant to policy makers who are concerned with the issue of market quality.

The effect of tranparency on the market quality measured as premium is examined based on 30-minute intervals. The methodology involves a linear programming model to form seven portfolios that are common in attributes, but as different as possible in the level of transparency. The portfolios are controlled for firm and time effects as well as for other effects found in the previous literature. Use of this procedure avoids serious problems common in most research designs using a general linear model. Further, this technique allows us to control the mean and higher moments of additional common attributes across portfolios. The major conclusions are: (1) the market quality is enhanced with a lesser degree of transparency; (2) relative to sellers, buyers tend to pay lower premiums for lower levels of transparency; and (3) an optimal degree of transparency is achieved since less tranparent portfolios are generally associated with the lowest premiums.

Since the debate on market tranparency is relatively new, it will be of continuing interest.