Secondary share sales at IPO by insiders happen frequently and on a large scale. Current literature offers mixed explanations. For example, signaling theory (Leland & Pyle, 1977) suggests that secondary share sales at IPO by insiders signal poor quality of the IPO firm. The premise is that insiders have more private information about the firm than outsiders. Therefore, insider sales should be indicative of trouble in the firm. This implies that insiders' secondary share sales will be associated with poor pre- and post-IPO performance. Agency theory (Jensen & Meckling, 1976) suggests that insiders will lower their commitment to the firm after they sell part of their shares; after such sales, managerial and firm interests are more poorly aligned. This theory suggests that poor post-IPO performance is causally associated with insiders' secondary share sales at IPO. Finally, risk aversion may drive insiders to diversify their risk away from the focal firm by selling secondary shares at IPO. This would suggest that sales have nothing to do with firm quality or managerial commitment.

      Although the above theories provide different implications for practice, the mixed nature of their explanations prevent us from having a clear understanding of the phenomenon. Additionally, prior studies have been unable to tease them apart. To address this issue, this dissertation investigates the following related questions: what factors predict insiders' secondary share sales at IPO and how do such sales affect various firm performances. Only by looking at the antecedents and consequences of insiders' share sales at IPO, as well as finding exogenous variation that affects secondary share sales and is unrelated to the characteristics of the firm can we see if the sales are associated with firm quality or risk aversion or if insiders lower their commitment after sales. 

The answers to these questions are investigated in three essays. In Essay 1, I ask which CEOs sell shares at IPO and under what conditions? Using a sample of 651 U.S. software IPOs from 1990 to 2011, I find that when more of the CEO's wealth is tied up in the firm, they are more likely to sell. The effect is especially strong for CEO founders. Interestingly, when board members also engage in equity share sales at IPO, CEOs are more likely to sell. This latter result suggests weakened board oversight of the CEO. Using an instrumental variable approach, I tease apart cotemporaneous selling due to poor firm quality and selling that only occurs with the reduction of oversight.

In Essay 2, I ask when equity share sales at IPO influence IPO underpricing. Through an analysis of 633 IPOs in the U.S. computer software industry, I find that the equity share sales by outside directors (VCs and other institutional investors) are associated with upward offer price revision pre-IPO and lower IPO underpricing. The interpretation is that outside directors may be able to bargain for a higher offer price when they attempt to sell part of their equity shares at IPO. As such, the upward offer price revision pre-IPO results from outsiders' bargaining leads to lower IPO underpricing. These results are robust to a Heckman two-stage approach that addresses potential selection bias. 

In Essay 3, I examine whether insiders' secondary share sales at IPO impacts a variety of performance measures post-IPO and the contingencies under which any impact may vary. Through the analysis of 500 IPOs of the U.S. computer software industry, in general, I find that insiders' secondary share sales at IPO are not associated with sales or sales growth. Rather, they are only associated with slower R&D growth in the year post-IPO. This effect is less negative for large firms. The results are robust to an instrumental variable approach to address the potential endogeneity issues. 

Taken together, this dissertation finds that insiders' secondary share sales are not significantly associated with post-IPO firm performances, providing no support to signaling theory or agency theory. The findings are more consistent with risk aversion theory and imply that insiders' secondary share sales at IPO are not a significant negative signal and traditional wisdom may overreact to the sales.