Propensity Score Matching and the New Issues Puzzle

Abstract

The characteristic matching approach consistently provides evidence of long-term underperformance following seasoned equity offerings (SEOs). This paper discusses the limitations of standard matching procedures, and applies propensity score matching, i.e., matching on the probability of equity offering, on a sample of SEOs from 1970 to 1997. Using standard industry, size, and book-to-market matching procedures, I replicate the significant negative abnormal returns documented in prior studies. However, when applying propensity score matching on the same sample, I show there is no evidence of negative long-term abnormal returns. Because propensity score matching enables one to find more adequate control firms, I conclude that sufficient matching can explain away the underperformance of the equity issuers.

Keywords: Propensity score matching, Long-term abnormal returns, New issues puzzle

JEL Classification: C51, G14, G2