Signalling by underwriter retention rate in the IPO market

Abstract

Underwriters in Taiwan have to purchase 10–25% of shares offered in initial public offerings (IPOs) for their own accounts. We present a signalling model showing that the underwriter retention rate can serve as a signal of firm value to investors because underwriters are investors as well. This mechanism can reduce information asymmetry between issuers and investors. The model shows that when underwriters retain more proportion of IPO shares, in equilibrium, the initial return is greater and the subscription success rate is lower. We further test the propositions using 616 IPO firms in Taiwan for the period 1998–2004. Overall, the empirical results support the propositions developed from the signalling model.

Acknowledgements

We wish to thank seminar participants at 2004 International Symposium on Knowledge-based Economy and Global Management held in Tainan, Taiwan and an anonymous referee for helpful comments. Hsuan-Chi Chen acknowledges the support from Yuan Ze University. Hsiu-Chuan Yeh acknowledges the support from the National Science Council of Taiwan under the grant number of NSC 92-2416-H-231-001.

Notes

¹ We exclude IPO firms with regulated retention rate of 50% because these IPO firms are special and risky technology firms and regulated by the government on the ground of profitability.

² Ritter (1984) points out that the other two possible explanations for the positive relation between firm value and insider holdings are the agency hypothesis and the wealth effect hypothesis.

³ The TSFC pricing formula is: P = EPS × (P/E) × 0.4 + DIV × 0.2 + BOOK × 0.2 + FDIV × 0.2, where P is the formula price; EPS is the past 3-year average earning per share of the IPO firm; P/E is the past 3-year average price–earning ratio of three comparable firms; DIV is the past 3-year average dividend per share of the IPO firm divided by the average of the past 3-year dividend yields of three comparable firms; BOOK is the IPO firm's book value per share; and FDIV is the IPO firm's 1-year forecast dividend divided by 1-year prime rate.

⁴ See Liu et al . (2001) and Chen et al . (2003) for more detail about IPO auctions in Taiwan.

⁵ The specific multiple is subject to change. Before 30 September 1999, the multiple was 1.5.

⁶ In this model, we assume that the value of the IPO is unknown at the underwriting stage and that the offer price is set primarily based on the negotiation between the issuer and the underwriter. Although we do not explicitly model the bargaining process, it can be seen later on that the offer price is associated with the value of the IPO through the participation constraint of the underwriter, which is V(α_L) = OP.

⁷ To keep the model simple, we assume that the underwriter underwrites one IPO at a time and holds only the risk-free asset for the rest of its wealth.

⁸ This notion is similarly used in the model of Leland and Pyle (1977).

⁹ As explained in Leland and Pyle (1977), such expected utility is possible whenever indifference curves are linear in mean and variance. An example is obtained when the utility function is a negative exponential function and returns are normally distributed.

¹⁰ In Equation 4, we implicitly assume V(α) ≥ OP and replace S(α) with L ₀ + L ₁[V(α) − OP], which will be verified to be true later in the Appendix.

¹¹ Although the regulation of underwriter retention rates started in December 1999, underwriters were allowed to retain 10–25% of IPO shares listed in the OTC market before December 1999. Therefore, we include those OTC IPOs with available underwriter retention rates in our sample.

¹² We thank the referee to point out the exploration for the possibility of underwriters’ immediate selling in the aftermarket and for the determinants of the underwriter retention rate in addition to the signalling effect.