Should Glass-Steagall Return?

 |  About: SPDR S&P Bank ETF (KBE)
by: William Armstrong

The separation of financial functions among commercial banks, investment banks and insurance companies that was the hallmark of the Glass-Steagall Act worked hand-in-hand with deposit insurance to produce a financial regulatory framework that served the U.S. well for many years. There are now some well-founded calls for its return.


A few days ago senators John McCain, Russ Feingold and Maria Cantwell introduced a deceptively simple one-page bill that would have the effect of reintroducing the separation among banking activities, insurance underwriting and securities business that existed from 1933 to 1999 when the Glass-Steagall (GS) Act was in effect. (GS was replaced as the main piece of U.S. banking legislation by the overly permissive Graham-Leach-Bliley Act of 1999.)

When considering the desirability of this proposal (or similar ones that may come along later) it is useful to return to first principals and consider the public policy underpinnings of GS’s separation of functions. It is easy to forget how much common sense and logic were embodied in this law.

At its most fundamental, the idea behind the separation of functions is to isolate the checking and savings deposits of the general public from the risks inherent in more aggressive forms of financial intermediation like securities underwriting and trading, and insurance underwriting. It sounds like a pretty good idea on the face of it, especially now that we are living with the damage caused by the absence of the kind of caution that GS enforced.

One key question to analyze when considering the kind of legislation the senators are proposing is whether the separation of functions can be pulled off without impeding the flow of capital to productive enterprise. The long and successful history we had with GS suggests that this is no problem.


Most countries provide a government guaranty on bank deposits up to a certain limit. This enables the general public to have confidence in the banking system, even though few will have the technical skills to measure a bank’s soundness. In the U.S. the FDIC analyzes the soundness of a bank that wishes to take advantage of its guaranty, and if it passes the test the FDIC gives the bank a license to take guaranteed deposits. Even though the banks are assessed a fee for this guaranty, almost all find it a good deal and join the FDIC system. In the U.S. the guaranty is provided for deposits of up to $250,000 per person. Since most individuals have less than this on deposit, the guaranty effectively applies to most of the deposits in the U.S. banking system. (Most experts think this is probably more than is necessary and that we could go back to the $100,000 limit that existed until recently without any ill effects.)

This license to take deposits that are guaranteed by the federal government represents a tremendous advantage to banks, enabling them to finance themselves at a very low cost. It is natural that this advantage should come with obligations that are designed to protect the enormous risk that the deposit guaranty represents to the national treasury.[1] These obligations are set out in regulations that limit the kinds of investments that banks can make in loans and securities. GS-type legislation provides a simple and useful framework for these regulations by prohibiting banks from most securities underwriting and trading, and insurance underwriting.

From a public policy point of view, this picture hangs together quite nicely. The FDIC guaranty helps to provide households with a safe place to save while giving banks a low cost of funds. Part of this low cost is shared with borrowers. The restrictions on investments enable banks to safely gear themselves up to a level where it is possible to turn a low lending margin into a high return on equity.

Savers earn a rate of return on their savings that over time tends to be about equal to the rate of inflation, so they are able to preserve the real value of their savings without exposing themselves to any risk. Plus they get the benefit of payment services thrown in, often for free. Families are able to finance 80% of the cost of a home over 30 years at a rate of 5-6% p.a. and sound businesses have access to working capital loans at reasonable rates of interest. And shareholders can earn 15-20% on their capital, more than sufficient to attract enough capital to the banking sector to create the desired level of competition. If regulation and supervision are done responsibly (as it was during the Glass-Steagall period) the government guaranty is hardly ever called.

There is not much wrong with this picture.

[1] Banks are assessed a charge to fund the guaranty but the government picks up the tab if the guaranty fund is insufficient to repay all the deposits of failed institutions, as tends to be the case during a financial crisis.

Disclosure: Long JPM, GFS, NLY