Centre for Economic Policy Research

5.6
Contemporary universal banking in the United States
While we now see that most allegations made during the Congressional hearings
cannot be supported by accumulated evidence, the myths propounded in the
hearings ensured that the New Deal legislation imposed a long halt on 
the movement towards universal banking in the United States. Constrained by the
Glass-Steagall Act and vigilant regulators, commercial banks largely ignored 
the securities business.
69
Only in the 1960s did commercial and investment banks again attempt to enter
each other’s turf and pressure regulators to change some regulations. Commercial
banks sought to persuade the Federal Reserve to expand the permissible activities
for bank holding company subsidiaries. This opening remained tightly controlled,
however, with revenue generated from formerly ‘ineligible’ activities initially 
limited to 5% of a subsidiary’s revenue and with firewalls imposed between the
subsidiary and the bank. On this limited basis, member banks moved into 
underwriting commercial paper, municipal revenue bonds and mortgage and 
consumer receivables backed securities.
70
In 1987, the Federal Reserve enabled commercial banks to re-enter investment
banking by permitting them to set-up bank holding company subsidiaries under
Section 20 of the Glass-Steagall Act in order to underwrite corporate securities. By
limiting their gross revenue from underwriting to 10%, the Federal Reserve 
considered these subsidiaries not to be in violation of the Act that prohibited 
affiliation with any firm that was ‘principally engaged’ in investment banking.
These new entities could underwrite commercial paper and were allowed to
underwrite corporate debt in 1989 and equities in 1990. Section 20 subsidiaries
were established subject to the erection of firewalls to limit information, resource
flows, and financial linkages with the affiliated commercial banks. The revenue
limit was later raised to 25% in 1996, and some of the more restrictive firewalls
were removed. 
The re-emergence of universal banks has not resulted in any perceived exploita-
tion of conflicts of interest; in fact, the market appears to believe that commercial
banks’ subsidiaries provide valuable information. Gande et al. (1997) examined
the characteristics and pricing of securities underwritten by the top 20 
underwriters (four Section 20 subsidiaries and 16 investment banks) for 1993-95.
Like their pre-Glass-Steagall predecessors, they found that bank subsidiaries
underwrite relatively smaller issues compared to independent investment banks.
71
Over time, as banks gained experience, the average issue size has declined even
further. This fact points to the special role of commercial banks in the financial
system and their ability to assist the firms facing the greatest asymmetric 
information hurdles, increasing small firms’ access to the market. 
Since the repeal of Glass-Steagall, one new concern has arisen with the 
reappearance of universal banking in the United States. Newly organized 
68 Conflicts of Interest in the Financial Services Industry

---

universal banks have been increasing their share of the investment banking 
market, often by giving clients credit facilities, which investment banks 
traditionally did not provide, on favourable terms (Cairns et al. 2002). This 
development may reflect economies realized from the combination of commercial
and investment banking or a conflict of interest that favours investment banking
units at the expense of the commercial banking units in universal banks. The 
policy issue is that large cheap credits may increase the risk to commercial banks,
and potentially the safety net.
72
Although there may be concerns for the safety net, recent experience suggests
that issuing firms have gained from the entry of commercial banks. The 
investment banking subsidiaries of commercial banks have provided strong 
certification for investors. Gande et al. (1997) found that when the commercial
bank has a significant lending stake in the issuing firm, the yield was 27 basis
points lower for lower-rated issues compared to investment banks. If the issue is
used to refinance part of the commercial bank debt and the parent bank still holds
a stake, the yield was 42 basis points lower. In the case where the loans are 
completely refinanced, the market does not penalize the subsidiary underwriter
and there is no significant difference in yield between debt issued by subsidiaries
or independent investment banks. They concluded that the firewalls are not so
high as to prevent banks with subsidiaries from more effectively certifying 
securities. For the economy, Gande et al. (1999) believe that there are significant
benefits from commercial bank entry into corporate debt underwriting. Focusing
on the period 1985-96, they determined that underwriter spreads, ex-ante yields,
and market concentration has declined. Increased competition was most apparent
among the lower-rated smaller debt issues. While underwriting spreads for 
corporate bonds have declined, there is no trend for equities, where commercial
banks do not yet have a significant presence.
The presence of substantial firewalls may limit conflicts of interest that might
emerge if regulations were not drawn so tightly. Gompers and Lerner (1999) 
investigated whether conflicts of interest were exploited when underwriting
investment banks hold stakes in the issuing firms through venture capital 
subsidiaries. Their sample covered venture-backed IPOs from 1972 to 1992, a 
period over which this type of IPO increased substantially. In this analogous, but
unregulated case, there is no evidence to support the limitations imposed on 
combining commercial and investment banking. The market appears to be 
concerned with potential conflicts of interest and offers a lower price for the 
securities, even though IPOs underwritten by investment banks with stake-
holding subsidiaries had the same or higher five-year excess returns and fewer
failed compared to IPOs of unaffiliated investment banks. Like the universal banks
of the 1920s, the response of the investment banks with equity stakes is to 
underwrite less information sensitive issues, and investor discounting is 
mitigated by the reputation of the bank.
73