Federal Reserve Bank of San Francisco

FRBSF Economic Letter

1998-32 | October 23, 1998

More Economic Letters

Risk and Return of Banks’ Section 20 Securities Affiliates

Simon Kwan

Western Banking Quarterly is a review of banking developments in the Twelfth Federal Reserve District, and includes FRBSF’s Regional Banking Tables. It is published in the Economic Letter on the fourth Friday of January, April, July, and October.

In recent years, the push to allow greater affiliation among banks and other financial firms has intensified. Through regulatory measures, banking organizations have made inroads into both securities and insurance activities. The legislative proposals under consideration in this session of the Congress would take things even further by tearing down virtually all remaining barriers to financial integration. While the legislative outcome for this year remains uncertain, the momentum is such that further financial integration is inevitable.

With full financial integration on the horizon, it is important to understand what these affiliations might mean for the profitability and risk of banking organizations, and thus the stability of the banking sector. This Economic Letter considers the implications of affiliations between commercial banking and securities activities. The evidence is drawn from a recent study (Kwan 1998) that examines the relationship between the performance of commercial banks and their so-called Section 20 securities affiliates.

Section 20 subsidiaries: some background

Sections 16, 20, 21, and 32 of the Glass-Steagall Act provide the legal basis separating commercial banking from investment banking since 1933. In particular, Section 20 prohibits Federal Reserve member banks from being affiliated with any organization that is engaged principally in securities underwriting or dealing, except municipal general obligation bonds, U.S. government bonds, private placement of commercial papers, and real estate bonds, which collectively are called “bank eligible securities.” However, since the term “engaged principally” was not defined in the act, the courts and the regulators have had to determine the meaning of these terms in enforcing the law. Beginning in 1987, the Federal Reserve authorized bank holding companies to establish securities subsidiaries to engage in limited underwriting and dealing in bank-ineligible securities. Today, these securities subsidiaries can underwrite and deal in a variety of securities, including corporate bonds and equities, as long as the revenues from conducting bank-ineligible securities activities do not exceed 25% of total revenues. Since the ineligible securities activities were authorized by the Fed under Section 20 of the Glass-Steagall Act, these securities affiliates are commonly referred to as Section 20 subsidiaries.

A sample of 23 domestic banking organizations with Section 20 subsidiaries indicates that such firms tend to be large: their average banking assets over the period 1990 to 1997 was $59 billion. Section 20 subsidiaries that are primary dealers of government securities are affiliated with even larger banks, which had average total banking assets of $94 billion. The volume of securities activities conducted by Section 20 subsidiaries that are primary dealers is much higher than nonprimary dealers, with total trading assets averaging $20 billion for primary dealers and $428 million for nonprimary dealers.

Return relationship between banking and securities activities

Kwan (1998) addresses two questions: (1) What are the profitability and risk of securities activities relative to banking activities? and (2) What are the potential diversification benefits of securities activities to a banking organization? To answer the first question, the study examines the mean and variance of the return on securities activities and compares them to those of banking activities, with the mean return measuring profitability and the variance of the return measuring risk. To provide insights into the second question, the study examines the return correlation between banking and securities activities, taking into consideration the stand-alone risk of each type of activity.

The average quarterly return on equity (ROE) was found to be slightly higher for bank subsidiaries than securities subsidiaries, although the difference was not statistically significant. However, the average standard deviation of ROE was much smaller for bank subsidiaries than securities subsidiaries, and the difference is highly significant. It is also worth noting that the ROE volatility was much higher among nonprimary dealers than primary dealers. The findings suggest that while Section 20 securities subsidiaries were no more profitable than bank subsidiaries, securities subsidiaries were much riskier than their bank affiliates.

However, the results also point to potential diversification benefits, since the return correlations between the bank subsidiaries and their Section 20 affiliates are close to zero. This suggests that the combination of a Section 20 subsidiary and a bank subsidiary can improve the risk and return tradeoff of the banking organization. In other words, if the bank and securities subsidiary operate independently, the firms can increase their returns only by proportionally increasing their risk exposure; with a bank and a securities subsidiary combined, however, the banking organization can achieve a higher return with the same level of risk or the same return with lower risk. Although the results indicate that banking organizations can reduce their risk exposure by engaging in the right amount of securities activities, too much securities activity can raise their overall risk due to the high stand-alone risk of Section 20 subsidiaries.

Since certain kinds of securities activities are bank permissible and are performed by banks rather than their securities affiliates, this may confound the analysis of banking vis-a-vis securities activities by examining activities at the subsidiary level. For robustness, the analysis also was performed at the activity level. In essence, the securities activities conducted by banks were subtracted from the bank subsidiaries and combined with the securities activities conducted by their Section 20 affiliates. This method is more precise in classifying banking versus securities activities and allows desegregating securities activities into trading and underwriting. Securities trading refers to buying and selling of securities, whereas securities underwriting refers to the distribution of securities from issuers to investors. However, the amount of capital and indirect expenses in the bank subsidiaries that is allocated to securities activities is unknown, so only the gross return on assets (ROA) can be computed.

Looking at ROA, the findings indicate that securities trading has a significantly higher return and higher risk than banking activities, especially for securities subsidiaries that are not primary dealers. Securities underwriting is found to have similar risk and return profiles to banking activities for primary dealers of government securities. For nonprimary dealers, securities underwriting appears to be less profitable than commercial banking. Regarding the portfolio diversification implications of securities activities, the trading activities by primary dealers are found to have potential diversification benefits for banking organizations; this is not the case for nonprimary dealers. On the other hand, securities underwriting is found to provide diversification benefits regardless of whether the securities subsidiaries are primary dealers or not.


Using micro data from banks’ Section 20 securities affiliates that were authorized by the Federal Reserve to engage in bank-ineligible securities activities, Kwan (1998) studied the effects of securities activities on banking organizations’ risk and profitability. Section 20 subsidiaries were found to be riskier, but not necessary more profitable, than their bank affiliates. The low return correlations between bank subsidiaries and Section 20 affiliates suggest that securities affiliates can provide diversification benefits to banking organizations.

Within the class of securities activities, securities trading is found to be more profitable and riskier than banking activities. Securities underwriting is found to have similar return profiles to banking activities for primary dealers, but it is less profitable than banking activities for nonprimary dealers. Trading activities by primary dealers seem to provide diversification benefits to banking organizations, while trading activities by nonprimary dealers do not. Securities underwriting by both primary and nonprimary dealers is found to be able to diversify banking risk.

Simon Kwan
Senior Economist


Kwan, S.H. 1998. “Securities Activities by Commercial Banking Firms’ Section 20 Subsidiaries: Risk, Return, and Diversification Benefits.” Federal Reserve Bank of San Francisco Working Paper 98-10.

Opinions expressed in FRBSF Economic Letter do not necessarily reflect the views of the management of the Federal Reserve Bank of San Francisco or of the Board of Governors of the Federal Reserve System. This publication is edited by Sam Zuckerman and Anita Todd. Permission to reprint must be obtained in writing.

Please send editorial comments and requests for reprint permission to
Research Library
Attn: Research publications, MS 1140
Federal Reserve Bank of San Francisco
P.O. Box 7702
San Francisco, CA 94120