We examine the relation between stock price performance and the identity of the investors buying the shares in private placements of equity. We find that although the shareholders not participating in the placement experience post-issue negative long-term abnormal returns, the participating investors purchase the shares at a discount and earn normal returns. For the non-participating investors, both announcement and long-term abnormal returns are significantly higher when the shares are placed with affiliated than only with unaffiliated investors. Additionally, when we exclude financially-distressed firms, we find insignificant announcement returns followed by negative long-term abnormal returns in placements to unaffiliated investors. On the other hand, consistent with affiliated investors having a certification effect, we find positive announcement returns and normal long-term returns following placements to affiliated investors. Thus, the disparity found in private placements between the positive announcement period and the negative post-issue long-term abnormal returns disappears when we control for financial distress and participating investor identity.

We develop and test a theory explaining the equilibrium matching of issuers and underwriters. We assume that issuers and underwriters associate by mutual choice, and that underwriter ability and issuer quality are complementary. Our model implies that matching is positive assortative, and that matches are based on firms' and underwriters' relative characteristics at the time of issuance. The model predicts that the market share of top underwriters and their average issue quality varies inversely with issuance volume. Various cross‐sectional patterns in underwriting spreads are consistent with equilibrium matching. We find strong empirical confirmation of our theory.