The Determinants of Issue Cycles for Initial Public Offerings

Abstract

This paper identifies the determinants of market-wide and industry-specific security issue cycles using an autoregressive conditional duration model. We examine the business conditions, investor sentiment, and time-varying asymmetric information hypotheses and show that issue activity in different industries is consistent with different explanations. We find that the business conditions and sentiment hypotheses explain issue activity by manufacturing firms; while issue activity by financial institutions is partly explained by the sentiment hypothesis. On the other hand, none of these explanations are capable of explaining issue activity in the business services industry. Surprisingly, when all of these industries are pooled to examine market-wide activity, we find that none of these hypotheses are significantly related to issue activity. One explanation is that market-wide aggregation washes out much of the industry-specific information because issue activity is not perfectly correlated across industries. Using this observation, we then consider whether technological innovations are important determinants of industry-specific issue activity. We test for industry contagion by examining the periods before and after the Netscape initial public offering. We find evidence of an increase in the correlation of issue activity in related industries, which is consistent with the technological innovations hypothesis.

Keywords: Issue Cycles, Initial Public Offerings, IPO,autoregressive conditional duration, ACD, contagion

JEL Classification: C22, G00, G32