A Response to Critics of Market Discipline
Top of the Ninth
Gary H. Stern
- President, 1985-2009
Published September 1, 1999 | September 1999 issue
Society relies on bank supervision and regulation to assess the risks that banks take and to ensure that such risks do not get "too large." Over the last several years I have argued that we would benefit from more market discipline and, in particular, from incorporating market information into the supervisory process. We have proposed a plan to achieve that goal. (See Summary of the work of the Minneapolis Fed in this area.) Other policymakers have also advocated more market discipline.
However, any change to increase the role of markets in bank regulation, no matter how modest, raises strong opinions and concerns. Indeed, new questions arise and the intensity of older criticisms increases as discussion turns to implementation. Thus, it seems like a good time to respond to some of the questions that I have received while advocating a larger role for market discipline. What follows are my responses to questions that have recently been posed to me.
It seems a little out of character to hear a bank regulator advocating market discipline, since we already have supervision and regulation to monitor banks' behavior. What does market discipline add to the mix?
I have to explain what I mean by market discipline. What you want to do is put some group of creditorslarge depositors or holders of other bank obligationsunequivocally at risk of losing some of their money if their bank fails. These creditors will then have more incentive than at present to pay attention to the caliber of banking institutions with which they do business. Given that incentive, you will start to get pricing of risk in the market. What does that mean? That means that those providing large sums of money to a bank-by buying the bank's debt, for example-will demand a higher return from a bank perceived as riskier. In practice, this risk of loss could mean that a holder of bank subordinated debt does not receive full repayment or that we allow uninsured depositors to lose some amount, say 10 percent of their funds, when their large bank fails. The point here isn't to argue the specifics of the method for putting funds at risk, but to reiterate the case for its necessity.
This type of market discipline would add several favorable attributes to the current regulatory mix. Most importantly, it would give regulators the market's take on how much risk a bank is taking. We rely on market signalsprices and so forthfor resource allocation all the time in our economy. And much to our advantage, I might add. In this sense, it seems almost irresponsible not to rely on information available from the market. Not only do market signals provide a good sense for which bank is taking more risk than another, the use of these signals could lead regulators to act more quickly than they might otherwise. Market signals also provide regulators with a sense of how much risk taking is efficient. I'm not sure how regulators can determine the right amount of risk without the help of markets, which probably helps to explain why regulators go from worrying about being "too lax" to fearing creation of a credit crunch.
I should note that occasionally other types of reform fall under the heading of increased market discipline. Some confuse market discipline with a mandate that regulators look at the results of the models that banks use to measure their risk and incorporate those model results in their exams. This may be a positive step, but it does not tell regulators what outside creditors think of the bank's risk taking. Others claim that having banks disclose more information will produce the favorable results I mentioned. While more disclosure also may be helpful, we first need to give creditors a reason to care about the financial condition of a bank.
This last point raises an issue to emphasize about market discipline: Once you change the incentives you are going to change the way participants behave. I think that's true not just of bank creditors, who clearly would have more incentive to pay attention, but banks will find it worthwhile to provide more information and, indeed, to market the safety and soundness of their institution. After all, banks that hold assets that are difficult to analyze would have to pay a higher cost to raise money. It follows, then, that these banks would either provide more information, charge borrowers more money when the bank holds their opaque loans, shift to more transparent assets, such as Treasuries, or some combination. Contrast this situation with what we have currently. It's really interesting, if you stop to think about it, no bank today says, "Leave your money with us because we only make low-risk loans or hold low-risk securities."
What makes you think we can really get the kind of information that will allow for better assessment and management of bank risk taking?
If you don't follow the economics of banking too closely, this question may seem somewhat odd. As I said, our economy routinely relies on markets to judge risk and allocate resources, and worldwide trends over the last 15 years suggest that this market-based system has grown in popularity But, in fact, this question actually cuts to the heart of the debate. Banks are a little different from other firms because the loans they hold are not easy for an outsider to assess. This fact, in part, helps justify the current regulatory regime. Even with this difficulty, however, we have plenty of evidence indicating that you would expect to see pricing distinctions between banks based on their riskiness.
It is worth repeating that I am not relying on some abstract theory, but instead can point to a number of studies showing that markets can be effective in assessing risk. Given the existing body of research, I'm not sure what additional evidence could be provided to policymakers without actually moving to a pilot project of some type where we try to incrementally increase the amount of market discipline in the banking system and in banking regulation. I guess I would like to hear what specific facts we need to acquire before we have enough evidence to move to a system of more market discipline. Unfortunately, and while I'm sure they are a minority, I think there are some people who oppose the various plans to increase market discipline and information because they don't trust that markets can work. I'm not sure that they would ever support a change in the current system, no matter what evidence others or I could provide.
You've convinced me that market discipline is important, so why only go part way? Why not introduce even more market discipline and get out of the business of regulating banks?
First of all, I view market discipline as a complement, or supplement, to what we're already doing in terms of safety and soundness examinations. We have a decent amount of evidence that regulators are able to serve a useful role in obtaining information about certain attributes of banks, such as asset quality. So I have never claimed that market discipline is a substitute for supervision, but that doesn't mean regulators couldn't use a little help. As I've said before, there's no excuse for not taking advantage of market information. Indeed, critics of market-based reforms are really saying that such information has little or nothing to add. This strikes me as a difficult case to make.
But you can go too far. There are legitimate concerns that if you relied exclusively on market discipline you would introduce more instability in the financial sectorand particularly in the banking sectorthan is desirable. So, what you're trying to do is strike a balance and incorporate both supervision and market discipline. And our plan and others like it try to achieve this balance by limiting who would suffer losses and how much they could suffer. As a result, I find the concern that increasing market discipline will create a system besieged by banking panics to be a straw man. Credible proposals will only put the largest creditors at risk and take steps, such as capping their losses, to address the potential for instability. In fact, it is the current system that lacks a balance between government protection and banking system stability. So, while I think it's fair to say that there are some things that may limit your enthusiasm for relying exclusively on market discipline, we're nowhere near that extreme right now, nor are we advocating going to that extreme.
I'm a community banker, and every time I hear the words "market discipline" I get nervous because it usually means that my large competitors get an advantage. I don't fear the market, but I do fear rules that favor big banks. How can I be sure that your proposal won't give big banks an advantage?
One way to be sure is to think about the current system where big banks, under the too-big-to-fail practices, actually do have an advantage. Nobody pretends that small community banks are covered by any too-big-to-fail doctrine. So small banks, from a number of perspectives, are subject to a reasonable amount of market discipline.
In fact, just look at the record following the passage of FDICIA. One of the things that we're certain did work about FDICIA is that small banks now do not always have their uninsured depositors protected. If you have over $100,000 at a small bank in Minnesota, for example, you need to know that the failure of your bank could expose you to risk of loss (the FDIC did not protect uninsured deposits at about 60 percent of failures between 1992 and 1997). The same depositor at a large institution has little reason to believe right now, under the current regime, that they're going to lose money. So what we're talking about and what most reforms try to achieve is leveling the playing field. Small banks should favor these plans.
I should also add that I don't find concerns about a slippery slope to be very compelling. In particular, I don't see how putting the largest depositors at risk at the largest banks creates an environment where the smallest depositors at the smallest banks will lose government protection. Our proposal does not advocate any change in the $100,000 limit.
Whenever we make more rules we add more costs. In an era where we are trying to reduce regulatory burden, why would you support plans that could lead to these higher costs?
I think we need to be careful in thinking about the costs of the current system and a future one under a plan like ours. The current system results in too much risk taking on the part of banks, which has a cost in terms of a misallocation of resources. And, as we saw in the late 1980s and early 1990s, cleaning up those problems can be very expensive to the taxpayer.
Why do those problems occur? Well, they occur in large measure because we've decided to shield banks, particularly the largest ones, through government programs like deposit insurance so that they do not have to pay the true costs for funds that markets would charge.
We're only suggesting that they pay costs closer to what the market would otherwise charge. This seems much more fair than imposing the costs of large bank risk taking on society as a whole, as occurs through the current system. And even under our plan, banks control the funding costs they face by determining how much risk they want to bear. This suggests that the cost of market discipline proposals has very little in common with the costs of command and control type banking regulations.
The good health of the banking system means that we have not tested FDICIA, which was passed in 1991 to address, in part, the concerns you have expressed; why not see if this bill will work?
FDICIA did a lot things that should prove helpful over time. But it did not do much, from my perspective, to increase market discipline for the largest banks, and it clearly did not require regulators to make use of market information, even though the evidence suggests that market signals can help determine whether Bank A is riskier than Bank B. So, as I've suggested before, a supporter of FDICIA is really saying that market signals don't have anything to add, that such market information is not helpful. Otherwise, why should we wait? Moreover, we're not talking about gutting the law and starting from scratch, but simply adding to FDICIA by taking advantage of market information. The spirit of our proposal goes hand in hand with the intent of FDICIA and should only serve to further its resolve.
Finally, we should undertake significant reform when the system is healthy and stable. You can't do this kind of thing when you're in the midst of a banking crisis, because that's not the time to change the rules of the game and the incentives; you would only add to the concerns and to the crisis atmosphere.
If some form of the financial modernization bill passes and banks continue to evolve into an all-purpose financial service firm, will your reforms still apply?
I think from a broad perspective the more that large banks become more like nonbank providers of financial services-like Merrill Lynch or Goldman Sachs-the more compelling the argument for reining in government protection becomes. After all, we don't provide this kind of safety net for these large nonbank firms. Rather, we have a system where these firms bear the cost of their actions. If we're not careful we will find ourselves in the rather strange situation of protecting a broader range of activities at the same time that larger banks lose those attributes that led to the protection in the first place.
The Minneapolis Fed on Increased Market Discipline
The Minneapolis Fed has argued for an increased role for market participants and information in deposit insurance and banking regulation over the last 15 years. In 1996, we renewed our calls for reform, arguing that changes to banking regulation should occur when the economy and banks are healthy and before banks are granted expanded business powers ("An Opportune Time for Deposit Insurance Reform," The Region, December 1996).
We were particularly concerned about the influence of the current regulatory regime on the risk taking of the largest banks. Mergers are creating more and larger banks that could fall under the so-called too-big-to-fail designation. The problem is that uninsured depositors and other legally unprotected creditors believe the government will bail them out if the large bank falters. As a result, these creditors do not incorporate the full riskiness of the banks' activities into the interest rate they charge. Banks get the wrong signal from the market and end up taking on more risk than they would otherwise. This policy ultimately leaves the taxpayer on the hook.
Congress tried to fix the TBTF problem in the Federal Deposit Insurance Corp. Improvement Act of 1991, but we believe it simply codified existing, flawed practices. Bank risk taking would be better addressed by putting uninsured creditors at risk of loss. As such, we proposed that Congress amend FDICIA to prohibit the full protection of uninsured depositors and other creditors in TBTF cases ("Fixing FDICIA: A Plan to Address the Too-Big-to-Fail Problem", 1997 Annual Report essay, The Region, March 1998).
Once creditors of these large banks have been credibly put at risk, the market will generate assessments of risk that will prove useful for regulators. We therefore believe these market signals should be incorporated into the deposit insurance and supervisory system. Banks that markets judged to pose higher risks would pay higher deposit insurance premiums or face more regulatory scrutiny, for example ("Managing Moral Hazard with Market Signals: How Regulation Should Change With Banking," The Region, June 1999).