Gary H. Stern - President, 1985-2009
Published September 1, 2006 | September 2006 issue
Editor's note: This column is based on remarks presented at the Independent Community Bankers of Minnesota convention on Aug. 4, 2006, in Alexandria, Minn.
Community banks care deeply about how the government regulates them, their competitors and their business partners and about reforms to those rules. In this column, I discuss four such regulatory issues: (1) recently enacted deposit insurance reform, (2) proposed reforms to the operations and regulation of Fannie Mae and Freddie Mac, (3) the application by Wal-Mart to charter an Industrial Loan Company (ILC) and (4) general efforts to reduce the cost of banking regulation.
Superficially, the only link among these four topics is their status as important, high-profile issues of the moment. But as I will describe, they are all connected through the government's safety net policy for financial institutions and the debate about the proper size of that safety net.
The link to safety net policy helps to explain why a central banker—who believes that safety nets are necessary but should be kept as small as feasible—chooses to address these issues. Let me remind you that these are my own views, and not necessarily those of the Federal Reserve, on these matters.
The central bank connection and aversion to safety net expansion has a motivation worth mentioning up front. Central banks have a responsibility to try to maximize standards of living, which we do primarily through maintaining price stability.
But in addition to setting monetary policy, we are expected to provide stability in the financial sector. We carry out that responsibility both during and after a financial crisis, for example, in our discount window lending and through our prudential supervisory and regulatory efforts to prevent instability in the first place.
In discharging our responsibilities for financial stability, we also attempt to improve standards of living by ensuring that the financial system allocates society's scarce resources effectively. Unfortunately, safety nets can interfere with this objective. For example, during the thrift and banking crises of the late 1980s to early 1990s, we witnessed examples of real estate development, which, absent incentives associated with the safety net, would never have occurred. These developments represented a clear example of significant resource misallocation.
Against this background, it is not too surprising that I will, first, express concern about the potential for some aspects of deposit insurance reform to increase the size of the safety net; second, support restrictions on Fannie Mae and Freddie Mac that limit and better control the safety net support they receive; third, raise doubts about increasing the mixing of banking and commerce that already is occurring via the ILC charter; and, finally, suggest that policymakers tie more general regulatory reform to reductions in safety net support.
Recently enacted deposit insurance reforms were intended to address two issues. The first is that inflation-adjusted deposit insurance coverage has decreased over time. The second is that premiums charged to banks are not linked to the risks of loss that banks pose to the insurance system (and these premiums have the potential to move up and down in short order). Put another way, for some the safety net had grown too small (that is, too little coverage), and for others it had grown too big (for example, not charging enough for coverage).
On the first point, I've consistently argued that increases in deposit insurance coverage are unwise. There are two primary reasons why I continue to think this is the case. First, the available data suggest that few households—even among those with the highest incomes—have anywhere near $100,000 in deposits at banks. To be sure, I have no doubt that most banks have loyal customers with funds exceeding the deposit insurance coverage limit, but the best, broadest surveys of households suggest that they are relatively few and far between.
Second, more coverage means more protection for bank creditors and, therefore, for banks. As a result, banks face more muted market forces that would otherwise help to restrain risk-taking. With more government protection, bankers pay lower rates on deposits and attract more deposits than they otherwise would, and thus they have incentive, particularly if their bank creeps closer to insolvency, to continue with activities even when the risk is excessive. To be clear, there is nothing venal in responding to prices and incentives, and I fully understand the desire of many bankers to increase the coverage offered to depositors. But, bottom line, I believe the societal costs of increasing coverage are greater than the benefits of doing so.
I also have concerns that deposit insurance reform didn't go far enough in reforming the pricing of deposit insurance. Now, this is a tricky topic, with much detail to it, but the crux of the issue is straightforward. I think the Federal Deposit Insurance Corp. should charge premiums based on the riskiness of banks. The reforms recently enacted, and the FDIC's proposed rules to implement the legal changes, are good steps in that direction. However, proposed initial premiums—particularly for the banks in the weakest condition—appear at first blush to not fully reflect the risk of the bank. Moreover, Congress continued a link—albeit a less stringent one—between aggregate premium collections and the reserves held by the FDIC. But unlike a private insurer, there need not be a link between reserves and FDIC premiums. After all, even if the reserves were negative, that is, if the FDIC were to become insolvent, the FDIC would still make good on its insurance commitment.1 In this sense, reserves are a distraction from making sure the premiums charged are risk-based, a policy that would limit the deposit insurance subsidy.
For the past several years, Fannie and Freddie, and to a lesser extent the Federal Home Loan Banks, have been center stage in policy and legislative debates. A couple of years ago, the debate centered on whether the supervisor of the housing-related government-sponsored enterprises would have the full range of powers of bank supervisors and receivers. But with the accounting scandals at the GSEs, the debate has more recently focused on whether and how to restrict the mortgage assets they hold on their books (that is, "on balance sheet").
In my view, limitations on GSE mortgage holdings are a necessary, but probably not a sufficient, reform, because this step addresses one, but not both, of the fundamental flaws of the housing GSEs, which primarily relate to the safety net the firms enjoy. The first current problem is that policymakers have virtually no direct control over the amount of safety net subsidy the GSEs receive. The second flaw is that the primary tool the GSEs use to achieve their main mission of increasing homeownership does little in practice to encourage or permit renters to become owners.
Let me briefly elaborate. The amount of subsidy that the GSEs receive from their implied federal guarantee depends on the amount of risk they take and the amount of subsidized funding they raise. Currently, the government only indirectly controls GSE risk-taking through supervision, which has become more effective of late but which does not have a compelling record. The government also has exercised virtually no direct control over the amount of funding raised by GSEs (the Treasury is currently reviewing whether it will begin exercising more control over such funding). This means that the GSEs control the size of the subsidy they get, which is bad policy, to put it mildly.2
The GSEs also largely determine who gets the benefits of the subsidy. The GSEs have housing goals that target certain groups (for example, central city residents). However, the Department of Housing and Urban Development and many independent analysts have concluded that the GSEs do no more to support housing for such groups than do the non-GSE firms that fund mortgages. This may change as the goals are increased.
My second point is based on the Federal Reserve Bank of Minneapolis' 2001 Annual Report essay.3 That essay notes that the GSEs' primary tool to increase homeownership is to engender lower mortgage interest rates than would otherwise occur. The estimates put the rate reduction stemming from the activities of the GSEs at around 25 basis points or maybe even less. At the same time, the literature on homeownership suggests that reductions in rates on the order of 200 basis points are needed to get more than a very small change in homeownership. In other words, the principal tool used by the GSEs to achieve their main mission is largely ineffective. So society is providing the safety net but not getting much of what was promised (more homeownership) in return.
Restrictions on the portfolios of the GSEs would go some way to limiting the safety net subsidy they receive, but such restrictions would not put that subsidy under the explicit and direct control of government. Moreover, such restrictions would not lead to more homeownership. So if that reform is not sufficient, let me suggest a few options.
One option that addresses both flaws is to consider the variety of steps that might fundamentally reduce the implied government backing of the GSEs. But as long as Congress adamantly supports the housing mission of the GSEs, eliminating implied government backing poses a challenge. A potentially superior option is ensuring that a much greater share of the GSEs' activities benefit people who face difficulty in becoming homeowners on their own. For example, the GSEs could be required to provide more of their support for prospective homeowners through direct down payment assistance or cash grants. This could lead to a real targeting of federal support to these households.
Before discussing safety nets, banking and Wal-Mart, I should briefly reiterate my concerns about the safety net for the creditors of the largest banks, as these concerns shape my views on the ILC issue. As I have noted before, I worry that some of these creditors believe they will be bailed out in the event of institutional insolvency because their bank is viewed as "too big to fail."4 In particular, these creditors believe that policymakers will not let a large bank collapse, fearing that the failure of the bank will "spill over" and lead other banks to fail, capital markets to freeze up and the rest of the economy to suffer. No policymaker wants to stand by in the face of potentially large-scale systemic risk and roll the dice. There are constructive actions that can be taken both to change the too-big-to-fail expectations of creditors of large banks and to reduce systemic risk. Those are, however, subjects for another day.
But because I believe we should strive to limit protection for creditors of large banks, additional combinations of banks and nonbanking firms that potentially expand the safety net to cover activities not currently in scope are particularly troubling. As a matter of public policy, it would be far preferable from my perspective to take steps to limit the existing safety net. To be clear, I am not speaking from a legal perspective. Under current law, it may be hard to prevent Wal-Mart from having an ILC when other nonbanks already do.5 This issue, of course, is the one currently before the FDIC.
Some have opposed the Wal-Mart ILC for a different set of reasons. In particular, they worry that Wal-Mart will out-compete small banks in the same way that the retailer has driven higher cost competition out of other markets. But consumers benefit from fair competition, so I'm not particularly sympathetic to this argument. Moreover, community banks have to date been quite resilient in the face of competition, and some community banks already face competition from large, multibranch banking organizations that likely benefit from whatever scale Wal-Mart has.
Congress is debating legislation that will make a number of changes to regulatory policies and practices with the aim of reducing the regulatory burden on banks. The details of this legislation have been discussed in numerous testimonies by governors of the Federal Reserve Board, most recently Vice Chairman Don Kohn in March.6
Instead of focusing on the details of reform, I would like to put forward a different strategy for reducing regulatory burden en masse. I have previously described the central bank's concern about systemic risk. Managing such risk is one of the underlying reasons for bank supervision in the first place. Yet significant supervisory resources are allocated to examining and monitoring smaller banks, even though they don't pose systemic risk. This suggests there may be room for reallocating supervisory resources and focus from the smaller to the larger institutions. I have made this point in public forums in the past, and some bankers and banking organizations have taken such comments to bolster arguments for reducing regulation/supervision of small banks.
This reaction prompted me to consider conditions under which reallocation of supervisory resources might be appropriate. I see three options for achieving that goal:
Many of the significant issues identified by community banks have at their root the size and scope of the federal safety net. Understandably, the beneficiaries of the safety net would like it to be larger. As a central banker, I would prefer that we carefully limit these supports because, other things equal, they encourage excessive risk-taking and inefficient use of society's resources.
1 See Ron Feldman, "When Should the FDIC Act Like a Private Insurance Company?" The Region (September 1998), 43-50.
2 See Ron Feldman, "Uncertainty in Federal Intervention," The Region (September 1996), 5-13. In a letter (Aug. 25, 2000) to Congressman Richard Baker, Chairman Alan Greenspan stated, "The GSE subsidy is unusual in that its size is determined by market perceptions, not by legislation. Indeed, the prospectuses of the debentures issued by GSEs explicitly state that they are not backed by the full faith and credit of the United States government. Accordingly, the extent to which the subsidy is exploited is determined by the extent to which GSEs choose to issue debt and mortgage-backed securities, not by legislation."
3 See Ron J. Feldman, "Mortgage Rates, Homeownership Rates, and Government-Sponsored Enterprises," 2001 Annual Report, Federal Reserve Bank of Minneapolis.
4 See Gary Stern and Ron Feldman, Too Big To Fail: The Hazards of Bank Bailouts (Brookings Institution Press, 2004).(Excerpts from Too Big To Fail: The Hazards of Bank Bailouts in The Region, December 2003.)
5 For views of the Board of Governors on this issue, see Scott G. Alvarez, "Industrial Loan Companies," testimony before the Subcommittee on Financial Institutions and Consumer Credit, Committee on Financial Services, U.S. House of Representatives (July 12, 2006).