A test of the effects of changing information asymmetry in a capital market

Abstract

Predictions by Lev (1988) about the observable effects of information asymmetry in capital markets are tested in the New Zealand sharemarket of late 1987 and early 1988. It is argued that the crash caused information asymmetries in that the extent of losses suffered by many companies were not public information. The crash should therefore have increased spreads, reduced volumes, and led to fewer quotes being available, while a subsequent special disclosure of the extent of losses should have had opposite effects. Both the median and the 5%-trimmed mean of spreads (adjusted for price changes) were found to increase at the time of the crash, and exhibited a short-lived decrease at the time of the announcement. In contrast, volumes rose for a few days after the crash, but then fell below pre-crash levels, and did not appear to change at the announcement date. The proportion of days on which both buy and sell quotes were available fell on both occasions. This evidence does not appear to support the idea that information asymmetry was important in this situation. It is noted that dealership models of a capital market may not give a good description of a market which lacks a dealership system. The autocorrelations suggest that spreads are generated by a nonstationary process.