EFFECTS OF CHANGES IN THE LEVEL OF PUBLIC DISCLOSURE ON THE ACQUISITION OF PRIVATE INFORMATION: AN EXPERIMENTAL MARKETS INVESTIGATION.

Abstract

This study reports the results of experimental laboratory markets designed to test two propositions set forth by Verrecchia 1982 . The first proposition addressed the change in the level of private information acquisition given a change in the level of public information in a competitive market. The second proposition considered the amount of informedness in the market given an increase in the level of public information and the resultant change in private information. The development of these propositions was motivated by the ambiguous results produced from the market-based accounting research investigating the impact on market price of mandated accounting disclosures. A limitation of the market based research is the inability to control for changes in the level of private information acquisition given a change in the level of public information which may explain the ambiguous results. A laboratory markets method was used to test the propositions because of the control provided by this research method. The market mechanism employed was a version of the PLATO computerized double-auction mechanism described by Smith, Suchanek, and Williams 1985 . This trading mechanism allows traders to communicate bids and offers and to form contracts to buy and sell assets in a computerized market which provides a high degree of control. In addition to the market for assets, a posted offer market for private information was used to allow traders to acquire private information. The results show significant decreases in private information acquisition in markets with higher levels of public information. Thus, public and private information appear to be substitute goods in this experimental setting. The results also indicates that the variance of contract prices around the true dividend value is significantly greater in markets with lower levels of public information. This remains true when controlling for possible confounding variables including market day, the contract number, the dollar value of private information, and the number of informed traders that executed the contract.