It is a truism of Washington that there are no final victories. Even when things seem settled once and for all, some new event will offer an opportunity to retrade an old deal.

So it is with Gramm-Leach-Bliley and its repeal of some of the principal elements of Glass-Steagall. The corporate scandals of this past spring and summer have given hope to the would-be regulators that they might be able to reclaim some lost territory. After all, they argue, if it weren't for Gramm-Leach-Bliley, Citigroup Inc., J.P. Morgan Chase & Co., and other banks would not have gotten themselves into trouble by lending to Enron or WorldCom in the hope of picking up securities underwriting fees. This is not quite true, of course. The Fed had authorized affiliations between banks and securities firms under its so-called Section 20 rules well before the repeal of Section 20 by the Gramm-Leach-Bliley Act. But it is nevertheless true that if Glass-Steagall had never been modified--first by the courts through statutory interpretation, then by the Fed complying with the courts' interpretations, and finally by Congress in ratifying what was already occurring in the market--various banks would not have been able to make improvident loans in pursuit of underwriting business.

It might occur to someone at this point that the loans made to Enron and WorldCom were not the only improvident loans ever made by banks. Perhaps it is mischievous to suggest this, but banks have always had the ability to make improvident loans on their own, without the temptation associated with securities underwriting fees. Surely the LDC debt crisis in the early 1980s and the failure of 1,600 banks in the late 1980s have not been entirely forgotten. There is little evidence that allowing banks to be affiliated with securities firms has suddenly introduced the evil of improvidence into otherwise pristine territory.

Still, the proponents of regulation do have a point: If we can eliminate all the various ways in which banks might get into trouble by trying to earn fees outside of the lending business, we might be able substantially to reduce the phenomenon in which banks make loans to companies that subsequently fail.

Let's consider this possibility, starting with leasing affiliates.

For years, it has been assumed that banks can be associated with companies that lease, say, airplanes, because leasing is simply another way of financing. In a leasing transaction, an affiliate of the bank holds the equity in an asset--let's use an airplane as an example--that is purchased from the manufacturer with a loan from the bank. The craft is then leased to an airline, which pays a lease fee that is supposed to be sufficient to repay the bank's loan, and when the lease term is up the airplane should have sufficient residual value to allow the bank's affiliate to recoup its equity investment.

In the course of structuring this transaction, there are many fees due to the lessor company, and thus to an affiliate of the bank. Here is a source of temptation we can immediately eliminate, and in this way enhance the safety and soundness of the banks previously engaged in this business. Now that the airlines are having so much financial difficulty, banks will be happy to be quit of this business anyway.

Then there is financial advice, especially on mergers and acquisitions. The fees here can be very large, and the temptation for banks correspondingly great. It can happen that a bank, in order to get an advisory position for a merger or acquisition, will make an improvident loan. Clearly, if we want to prevent the making of loans like this, banks should be prohibited from advising corporations on financial matters. That's OK; there will be plenty of other advisers when M&A activity finally comes back.

It is also necessary to recognize that fees are not the only source of temptation for banks. Long-term relationships with customers are also avidly sought, and can give rise to loans on terms that are more favorable than they should be. Sometimes, in an effort to establish a banking relationship with a potential customer, banks will make loans that--in hindsight--they and their supervisors will regard as excessively liberal. We might need regulations that will stop banks from attempting to form new relationships in this way. The old relationships ought to be good enough.

Of course, some might think that this is carrying things too far. But if it is, there must be some distinction between seeking securities underwriting fees and seeking other kinds of revenue.

The old arguments in favor of retaining Glass-Steagall were based on the claim that securities underwriting was uniquely risky, and that permitting banks to affiliate with securities firms would create unacceptable risks for the bank. This argument was shown to be fallacious when it became clear that underwriting securities was in fact less risky than making a long-term bet on a company's financial health, which is what banks normally do when they make a loan.

But the new argument for the reinstatement of Glass-Steagall relies on a new principle. The issue is no longer the riskiness of the bank-affiliated business, such as securities. Now it is the propensity of that business to lead otherwise sensible bankers into perdition.

Until someone can show that the securities business has a greater propensity to produce this outcome than, say, leasing, financial advice, and competitive lending itself, we should ignore these arguments for reestablishing the restrictions of Glass-Steagall.