Ch. 5: Banks in Capital Markets 205
lead managed issues during the time period. However,Song (2004)is able to make a
related comparison in the bond underwriting market. Using a sample of 2,345 corporate
bond issues from 1991 to 1996,Song (2004)examines the clienteles and bond pricing
associated with three different syndicate structures: (i) commercial bank-lead syndi-
cates; (ii) syndicates with only investment banks; and, (iii) hybrid syndicates where an
investment bank leads the issue and commercial banks are co-managers. Song uses an
endogenous switching model with six equations: three selection equations, which cap-
ture the likelihood of choosing a given syndicate structure over the other options, and
three yield equations, one for each of the three syndicates.^17 The results of her model
indicate that commercial banks are more likely to co-manage an issue rather than serve
as lead manager when the purpose of the issue is to refinance bank debt and the issuer
has more loans from the commercial bank underwriters. Since these issues are more
likely to be prone to conflicts of interest, the results are consistent with the view that
acting as a co-manager allows commercial banks to mitigate perceptions that they will
exploit conflicts of interest. However, bond yields are similar when commercial banks
are lead managers as opposed to co-managers. This suggests that co-managing does not
improve the certification ability of commercial banks.
3.4. Conflicts of interest from equity holdings: Evidence from venture capital
Much of the focus so far has been on the trade-off between the private information
from lending allowing banks to be better certifiers of firm value with the potential
for conflicts of interest from misusing the information. In this section, we again ex-
plore this trade-off, but examine some of the different effects that can occur when
underwriters are equity holders in the firm. The evidence from venture capital can
provide insight into the potential consequences of allowing banks to hold equity in
firms.
Some authors maintain that allowing banks to hold equity claims helps increase a
financial intermediary’s credibility in certifying the firm’s value (see e.g.Leland and
Pyle, 1977), which provides a formal analysis of how equity holdings in the firm can
provide a signal of firm value). However, asPuri’s (1999)model points out, the hori-
zon for which equity is held is critical to this certification. If the bank can retire its
financial claim through the proceeds of the equity issuance, then holding equity can
hurt the credibility of the bank more than holding debt. There is some empirical evi-
dence on the impact of equity holdings on the certification ability of the underwriter,
derived from comparing IPOs where the underwriter has gained an equity stake through
venture capital investments and other IPOs where the underwriter does not have an
equity claim. Both papers that we survey do not find evidence of conflicts of inter-
est.
(^17) In another application of this method,Fang (2005)studies the relation between investment bank reputa-
tion and the prices of underwritten bonds and uses separate pricing equations for high and low reputation
underwriters.Fang (2005)finds that more reputable underwriters obtain lower yields and charge higher fees.